Nanocellulose has attracted much attention in many fields due to its excellent properties and unique structure. Its preparation methods and applications have become the research hotspot in related areas. At present, there are many methods for preparing nanocellulose, but they still face great challenges. Ionic liquids (ILs) have shown great potential in the preparation of nanocellulose because of their excellent solubility and recyclability to lignocellulose. Therefore, the application progress of ionic liquids in the preparation of nanocellulose was reviewed, with the focus on the achievements of ionic liquids as a pretreatment method and as both solvent and catalyst in the preparation of nanocellulose, respectively. Besides, the recovery methods of ionic liquids were also summarized.

The spherical hydrogel adsorbent prepared by titration-gel method has a 3D inverted funnel-like micro-morphological structure, a wide pore size distribution, and a fast response mechanism to heavy metals, dyes and other pollutants in water. It has been widely used in water treatment process research. The one-step titration-gel method is really a facile way to fabricate the hydrogel adsorbents with observably spherical structure. Herein, the formation mechanism of the hydrogel beads which were developed by the one-step titration-gel method was discussed. And, the authors of this manuscript review the researches which published in recently that used the hydrogel beads to remove the different pollutants from waste water. The main difficulties and limitations, or the prospects of the spherical hydrogel beads used in waste water treatment, also, were stated. The authors of this manuscript hoped this work would provide a corresponding guideline for the researchers who are aiming at preparing the spherical hydrogel adsorbents for waste water treatments.

Biodegradable polymers/drug nanoparticles have outstanding advantages in the fields of the drug targeted delivery, effective component encapsulation and medical diagnosis and so on. The nanoparticles preparation technologies from supercritical fluid have been widely studied because of its green environmental protection, many kinds of preparation methods, easy to adjust product quality according to the requirements of use, easy subsequent separation and purification, and so on. In order to obtain polymer/drug nanoparticles to meet the requirements, supercritical fluid granulation technology is one of the effective methods. The characteristics and applications of biodegradable polymer materials were discussed. Supercritical fluid and its characteristics were briefly introduced. The technological characteristics, preparing methods, basic principle and research progress on the rapid expansion of supercritical solution (RESS), supercritical anti-solvent precipitation (SAS), supercritical CO₂ assisted atomization (SAA) and supercritical fluid emulsion extraction (SFEE) were mainly introduced. The development directions of polymer/drug nanoparticles prepared by supercritical fluid technology were prospected.

Black phosphorene is a new type of two-dimensional material with high carrier mobility, tunable direct band gap, unique anisotropic physical and chemical properties, leading to wide application prospects in energy storage, photoelectric, medicine, sensing and other fields. However, black phosphorene is easy to be degraded at the presence of moisture and oxygen, which limits its practical applications. Therefore, how to solve the problem of phosphorene stability has become the focus of current research. In order to let researchers who just entered the field have a comprehensive understanding of this problem, the unstable mechanisms of black phosphorene were first reviewed and then several typical strategies in recent years to stabilize black phosphorene were summarized, the progress of our research group in solving the stability of black phosphorene were also introduced. Finally, on the basis of the current progress, future research directions to improve the stability of black phosphene were suggested.

The microcystins (MCs) produced and released by cyanobacteria blooms are highly stable in the natural environment and are difficult to be effectively removed by traditional water treatment technologies, posing a serious threat to drinking water safety and ecosystems. Microbial biodegradation is an effective strategy where microbial metabolic potential can be harnessed forin-situ andex-situ detoxification of MCs. At present, many studies have shown that some bacteria isolated from water bodies and sediments are capable of degrading MCs, which as an efficient and environmentally friendly strategy for MCs removal. The critical enzyme identified in bacterial strains is referred to as MlrA (also known as microcystinase), play a major role in the water body bioremediation process, responsible for catalyzing the first step reaction in the MCs biodegradation pathway. In this paper, based on evolutionary relationships of MlrA, reviews the characteristics of enzyme activity, molecular structure and catalytic mechanism of this enzyme. The further research subjects of MlrA were also proposed. This review could provide a reference for bioremediation of contaminated water environment.

The microscopic nature of methane hydrate decomposition is simulated in NVE ensemble by SPC/E-UA, SPC/E-AA, TIP4P-UA, and TIP4P-AA molecular models. The equilibrium temperature and decomposition heat under different models are calculated and comparatively analyzed. The hydrate-liquid water interface is distinguished by calculating order parameter F₃, and the time-varying and space-varying hydrate-liquid boundary is monitored. The variations of potential energy at different zones are compared which indicate the decomposition process has significant non-equilibrium heat transfer characteristics. The spatial density distribution of hydrate water and methane is contrastively analyzed and the variation of the number of methane molecules that escape from the methane hydrate during decomposition is determined. It s found that the escape velocity decreases gradually with decomposition. The activation energies required for decomposition under different models are calculated and compared with the experimental results. By comparing the mean square displacements and diffusion coefficients in different decomposition regions, it s found that the decomposition exhibits significant non-equilibrium mass transfer characteristics. The study shows that the methane molecular model has little influence on decomposition, while the water molecular model has observable influence. The SPC/E water model can better reproduce the experimental values of equilibrium temperature, activation energy and decomposition heat in methane hydrate decomposition under NVE ensemble.

During the flight of the rocket, a low temperature helium gas of about 90 K is used to pressurize the kerosene tank at room temperature to allow kerosene to flow out to ensure the supply of engine fuel. In order to reduce the amount of helium, the cryogenic gas is designed to inject into the tank under the liquid level, in which way it can absorb the heat from the liquid kerosene in the tank while rising, and then enters the gas phase to pressurize the tank. However, this process may lead to two unfavorable results. First, the surface of the cryogenic helium gas line immersed in kerosene may freeze, and the ice may sink to the bottom to block the engine screen. Second, the helium gas may be carried by kerosene, thus two-phase flow will appear at the discharge port. Both of the circumstances have an adverse impact on the operation of the rocket engine, and therefore neither of them is allowed. In this paper, numerical simulation study on two-phase flow of the cryogenic helium gas injecting into the liquid kerosene is performed through two injection ways, including the central injection from a core and circumferential injection from three cores. The unsteady two-phase heat transfer process is modeled based on the Euler-Euler model. The numerical results are qualitatively compared and verified with the experimental observations on the ground. The phase distribution and the possibility of kerosene icing in the two injection structures during the discharge process are investigated. The results provide more insights into the experimental phenomena and also a reference for the scheme of kerosene discharge.

By means of perspex hopper platform built in the laboratory, the flow behaviors of powders were investigated under the action of various flow channel structures in hoppers without insert (No-In), with confined insert (Con-In) and with unconfined insert (Ucon-In). Discharge experiments of free-flowing glass beads and adhesive pulverized coal and polyvinyl chloride were carried out and a model was established to predict the solid flow rate. The effects of inserts contributing to increasing the flow rate were analyzed quantitatively and the flow characteristics of glass beads, pulverized coal and polyvinyl chloride were compared considering distinct flow channel structures. The research shows that the introduction of the modified fluid is beneficial to increase the flow rate of the silo. Con-In promotes the most obvious flow effect. For the weak coal powder, the flow rate of the lower flow rate reaches the maximum of 58%. Based on the evolution of shear flow zone, a correction factorF was proposed to modify the minimum energy theory equation, and the model for ideal hoppers was extended to take effect under actual situations. Furthermore, regarding Con-In, according to the structural characteristics of the flow channels and the force analysis on powders, the term relating to conical angle in the model was modified. In terms of Ucon-In, based on the competition mechanism, a specific flow sequence during discharge was proposed and the relationship between flow rate of inner layer and that of interlayer was obtained. Finally, a model to predict the solid flow rate in hopper with complicated flow channel structures was established. The model took into account the effects of powder properties, flow patterns and structural parameters, which can effectively predict the discharge rate of free-flowing and adhesive powders flowing through traditional hoppers (No-In) and hoppers with insert (including Con-In and Ucon-In) with the deviation less than 10%.

Flow patterns and pressure drop characteristics on high-viscosity oil and gas two-phase flow in upward pipe were experimentally studied with an indoor experiment apparatus. Seven flow patterns were observed in the experiment. Pressure fluctuation signals and pressure drop values under different operating conditions were obtained. The results show that, due to the influence of liquid viscosity, the transition boundary of most flow patterns in the upward tube shifts to the left side of the flow pattern diagram, and the larger the viscosity is, the greater the migration degree is.The data of flow patterns were compared with Barnea model, whose error was found to be increased at higher viscosity. According to the pressure drop data, due to the boost of adhesion at higher viscosity, the phenomenon that the pressure drop decreases with the increase of liquid viscosity may occur during the lower superficial velocities. Comparison verification of pressure drop with OLGA model and Zhang model was performed, whose calculation error under high-viscosity was found to be greater than that of low-viscosity. Finally, the Zhang model was modified with the closure relationships for high-viscosity oil and the updated results showed that the accuracy can be significantly improved.

There are many problems in the enhanced geothermal system used for dry hot rock thermal energy exploitation, such as high investment, high risk, working medium leakage, equipment corrosion, land subsidence,etc. Using super long gravity heat pipe to mine hot dry rock geothermal energy can effectively avoid these problems. In this paper, an experimental platform for super long gravity heat pipe is built. The suitable liquid filling capacity, stability of operation and heat transfer performance under different cooling water flow rates are studied experimentally. The possible causes are analyzed. The research shows that under the constant power, the suitable liquid filling amount of the heat pipe is about 40% of the total liquid filling amount of the evaporation section. During operation, compared with the conventional short heat pipe, the super long heat pipe exhibits strong oscillation, oscillation frequency and heating power and charging. The amount of liquid is closely related. At a constant heating power, as the flow rate of the cooling water increases, the power of the heat pipe increases first and then gradually becomes gentle. In addition, the heat transfer performance of the heat pipe under the extreme liquid filling amount is especially discussed. The research shows that under the extreme liquid filling amount, the gas column forms a certain height at the bottom of the heat pipe, and the heat cannot be transferred to the top of the heat pipe due to the continuous existence of the gas column. The experimental results preliminarily verify the feasibility about using the super long gravity heat pipe to mine hot dry rock geothermal energy and offer fundamental supports for the future practical applications.

The dynamic icing presents a porous structure. This microscopic pore structure is a key factor affecting its macroscopic physical properties. However, it is difficult to quantitatively characterize the three-dimensional pore structure with traditional methods. Aimed to describe the microcosmic characters objectively, a 3-D modeling approach of porous structure based on quantitative results of 2-D images is proposed in this paper. Firstly, the ratio of measuring scale between real and modeling size is studied and given, and its influence on resolution and modeling is also analyzed. Secondly, based on the porosity and diameter distribution, method to determine the amount and diameters of pores is presented. And the random approach of pores location is given. After these preparations, the 3-D porous structure of ice is formed by a matrix with element value 0 or 1. The experimental results demonstrate that the 2-D quantitative information of pores extracted from 3-D generated ice is highly similar to the real ice, which shows that the proposed approach is accurate and feasible. Through this paper, a new way to 3-D display, characters extraction and corresponding simulation of ice in microcosmic view can be provided.

A closed spray cooling test bench is built. The transient heat transfer process of spray cooling was experimentally studied, and the experimental curve describing the heat transfer process accurately was obtained. The influence of initial cooling temperature, heating power and medium type on transient heat transfer process is analyzed. The results are as follows: the trend of surface temperature can be divided into rapid decline, continuous increase and second decline. After experiencing the initial enhancement effect, if the initial surface temperature is less than the temperature T_fat Leiden frost point , the surface temperature decreases continuously and the thermal equilibrium is achieved in the nucleate boiling region; On the contrary, the surface temperature increases and the thermal equilibrium is achieved in the film boiling zone. The value of constant heating power determines the rate of surface temperature change. With the increase of constant heating power, the rate of surface temperature decreasing or rising accelerates. In addition, for different types of media, the higher nozzle inlet pressure and saturation temperature mean the higher temperature T_f at Leiden frost point.

The boiling heat transfer characteristics of R290 in horizontal micro-fin copper tubes with inner diameter of 4mm and 6mm were investigated experimentally. And the effects of mass flux(100—250 kg·m^-2·s^-1),saturation temperature(7—11℃),heat flux(13—24 kW·m^-2) ,tube type and vapor quality(0.1—0.9) on the boiling heat transfer coefficient and critical quality were analyzed .The results showed that the boiling heat transfer coefficient increased with the increase of mass flux or saturation temperature. The boiling heat transfer coefficient increased at first but then decreased as the increase of heat flux. And the results also showed that critical quality decreased with the increase of heat flux. Comparing to the smooth tube, micro-fin had a larger critical quality. With the increase of vapor quality accompanied by the heat absorbing boiling process of R290, the boiling heat transfer coefficient increased at first but then decreased. Six commonly used correlations of boiling heat transfer were used to predict the boiling heat transfer coefficient of R290. Comparing with the experimental results, the prediction accuracy of Fanget al. and Choiet al. was higher.

Based on the laws of mass conservation and energy conservation, the regeneration performance prediction model of IH-UARS is established and verified by experiments. Besides, experimental runs were conducted to verify the feasibility of the developed model. The results show that there exists an optimal range of internal heating, which makes the regeneration system present the most energy-efficient regeneration. The optimal heating power was found increasing logarithmically with the growth of the flow rate of the desiccant solution but is weakly affected by the flow rate of the scavenging airflow. Under the standard operating conditions of the present study, the optimal heat power for IH-UARS was around 275 to 350 W. Furthermore, it was also found that the internal heating seems to be more beneficial for regenerating the dilute desiccant solution with a higher initial mass fraction. For instance, by increasing the internal heating power to 800 W, the \mathrm{D}\mathrm{M}\mathrm{F}{\mathrm{I}}_{G_l} of the desiccant solution, with inlet concentration of 36% was better than the desiccant solution with inlet concentration of 24%, with the improvement of \mathrm{D}\mathrm{M}\mathrm{F}{\mathrm{I}}_{G_l} amplitude reached 37%. The study may help realize the optimal running of the internally-heated desiccant regeneration system.

The synthesis of benzoic acid using ozone combined with hydrogen peroxide in a rotating packed bed (RPB) was proposed in this study. Firstly, the process (RPB-O₃/H₂O₂) was compared with other ozonation processes, such as BR (bubbling reactor)-O₃/H₂O₂，RPB-O₃，BR-O₃, and RPB-O₃/H₂O₂ exhibited higher oxidation property for the synthesis of benzoic acid. Secondly, effects of different operating conditions including solvents, ozone gas concentration, molar ratio of hydrogen peroxide to toluene, high gravity factor, liquid flow rate, solvents and reaction time were investigated. The results revealed that the optimal operation conditions were solvent of acetonitrile, the ozone gas concentration of 80 mg·L^-1, molar ratio of hydrogen peroxide to toluene 0.15, high gravity factor of 40, liquid ?ow rate of 120 L·h^-1. Under the optimal conditions, the benzoic acid is obtained in 45% yield. The active radical generated in the reaction process was characterized by electron paramagnetic resonance (EPR), and ·OH was only found. Besides, the main intermediates of benzyl alcohol and benzaldehyde were identified by gas chromatography-mass spectrometry (GC-MS). Based on the results of ERP experiment and GC-MS characterization, the possible reaction process of benzoic acid synthesis by oxidation of toluene with ozone / hydrogen peroxide was explored.

Catalysts of La_1-xLi_xMnO₃(x=0, 0.05, 0.15, 0.25) perovskite-type complex metal oxide were prepared by solvothermal method. The catalysts were characterized by XRD, FT-IR, XPS, H₂-TPR and NO-TPD. The catalytic performance of the catalysts for eliminating nitric oxide (NO) and diesel particulate matter （soot） simultaneously was evaluated in a fixed micro reactor. The results showed that La_1-xLi_xMnO₃ catalysts had single perovskite structure and Mn species existed in the form of Mn³⁺ and Mn⁴⁺. When La³⁺ is partially substituted by Li⁺ in LaMnO₃, there were more Mn⁴⁺ and oxygen vacancies in the La_1-xLi_xMnO₃ catalysts, which can promote the adsorption and activation of oxygen species and nitrogen oxides, thus increase the redox properties and improve the catalytic activity. The La_0.85Li_0.15MnO₃ catalyst had better performance, where the doping amount of Li⁺ was 0.15, the temperature of the maximum combustion rate (T_max) and the ignition temperature (T_ig) was 463℃ and 327℃, respectively, the maximum conversion of NO (X_{\mathrm{N}\mathrm{O}}^{\mathrm{m}\mathrm{a}\mathrm{x}}) was 42.1%, reaction temperature (T_{\mathrm{N}\mathrm{O}}^{\mathrm{m}\mathrm{a}\mathrm{x}}) was 383℃.

Nano-Fe₃O₄ particles are first synthesized under ultrasonic condition, and the ultrasonic enhanced functionalized nanocomposites Fe₃O₄-mPD/SP(50∶50) modified with amino, imino and sulfonic acid are synthesized in one step in presence of ultrasonic irradiation, which were characterized by TEM, XRD, IR, TGA, VSM and BET. The adsorption properties of Cr(Ⅵ) were investigated, which show that the pH value of the solution has a significant effect on the adsorption of Cr(Ⅵ) by Fe₃O₄-mPD/SP(50∶50), and the adsorption is best when pH=2. The amount of adsorbent, competitive anion(Cl^-,NO₃^-,SO₄^2-) and temperature all had effect on the adsorption of Cr(Ⅵ). The adsorption mechanism of Cr(Ⅵ) by Fe₃O₄-mPD/SP(50∶50) was discussed, which show that the quasi-second order kinetic equation can describe the adsorption behavior of Cr(Ⅵ) on Fe₃O₄-mPD/SP(50∶50) well, k₂=1.324×10^-3 g·mg^-1·min^-1, q_e=77.157 mg·g^-1; The isothermal adsorption data were in accordance with the Freundlich model, K_F=34.464 mg^1-(1/n) ·L^1/n·g^-1, n=3.861. The adsorption of Cr(Ⅵ) is a spontaneous process, ?G⁰<0，?S⁰=73.368 J·mol^-1·K^-1, ?H⁰=19.375 kJ·mol^-1. It is inferred that the adsorption mechanism includes electrostatic adsorption, redox and ion exchange.

In the modeling of chemical process, due to the high dimensionality and non-linearity of the process data, the calculation amount is greatly increased and the modeling difficulty is increased. In order to solve this problem, a regularization based functional link neural network (RFLNN) is proposed. In the proposed RFLNN method, there are two salient features through using the regularization

Aiming at the characteristics of large data volume and high dimensions in the current industrial control system, a grey wolf optimization integrated random black hole (RBHGWO) algorithm incorporating a random black hole (RBH) strategy is proposed. When the wolf group updates the position of the next generation grey wolf, the proposed algorithm simulates the attraction of black holes, so that the individual in the wolf group can move faster towards the current global optimal solution, and enhances the convergence speed of the proposed algorithm. Meanwhile, individuals are randomly attracted by black holes, which maintain the local search ability of the proposed algorithm. Compared with particle swarm optimization (PSO), random black hole particle swarm optimization (RBHPSO), GWO algorithm and survival of fitness grey wolf optimization (SFGWO) algorithm using test functions, the experimental results show that the RBHGWO algorithm has fast convergence speed and excellent convergence accuracy. Moreover, based on the data set of Tennessee-Eastman (TE) simulation platform, the situation of industrial control systems is simulated by attacking from the covert intrusion. The experimental results show that the RBHGWO algorithm has obvious advantages in convergence accuracy, iteration speed and stability in the feature selection of intrusion detection of industrial control systems.

The existing ethylene cracking furnace optimization usually only considers the two objective functions, namely the yield of ethylene and propylene; the convergence effect of genetic algorithm is not good enough. Therefore, an improved NSGA-Ⅱ algorithm is proposed to study a multi-objective operating solution for the fixed cycle operation optimization problem of the ethylene cracking furnace, which is to increase the ethylene and propylene yield while reducing the raw material and steam flow to improve the overall operation. We quantified the specific problem into a mathematical model and analyzed the effects of feed gas-to-hydrocarbon ratio, feedstock flow rate, and outlet temperature on the yield of ethylene and propylene. The experimental results show that compared with the original operating conditions, the proposed optimization scheme has good feasibility.

Fuzzy cognitive maps (FCM), as a modeling tool for complex systems, can handle the nonlinearity and uncertainty of the system. However, time-delay among industrial process variables is always ignored in traditional FCM models. The causal relationship between variables can t be calculated accurately. Therefore, the prediction result is inconvincible and unpredictable. The time-delay-mining fuzzy time cognitive maps (TM-FTCM) method is proposed to enhance the accuracy of the time-delay prediction model. The cross-correlation functions (CCF) helps to find the time-delay factors hiding in the big data, thus revealing the actual structure of the model. Furthermore, the optimization of self-impact factors, bias and transfer functions enhances the efficiency of the prediction process. The TM-FTCM method has been verified by numerical simulations and actual chemical plant process data to be efficient and practical.

The injection molding is a typical batch process and consists of three stages: filling, packing-holding and cooling. Packing-holding is an important stage to determine product quality and has obvious nonlinear and time-varying characteristics. Piece-wise affine (PWA) model can effectively describe the dynamics of packing-holding stage, but the hard switching of PWA model will lead to the output jump problem. Therefore, we propose a novel time-partitioned PWA fusion model, in which the switching regionis introduced between adjacent sub-models and model fusion is realized by linear weighting method. Then, an identification algorithm based on separable nonlinear least-squares is developed to estimate model parameters. Based on this, a multi-model PID control method based on internal model tuning is designed.Experimental results of injection molding machine show the effectiveness of the algorithm.

In the context of intelligent manufacturing, cost accounting and value stream analysis for digital resources have important practical significance for enterprises intelligent manufacturing investment and transformation. A kind of multi factor production cost model is established based on the composition domain reference model architecture for intelligent manufacturing. According to the cost accounting requirements of different management levels, the coupling relationship model of cost nodes between different management levels is established, and the unified modeling and data consistency of multi-level cost models are realized. It can be used to value flow analysis for enterprise intelligent manufacturing, and provide theoretical basis and cost calculation model for improving the cost efficiency of intelligent manufacturing and investment decision.

Gold cyanidation leaching processes suffer from slow-dynamics and strong uncertainties. To this end, an economic model predictive control (EMPC) based dynamic real-time optimization scheme is proposed for the gold cyanidation leaching process. Unlike traditional model predictive control, EMCC takes economic indicators directly as the objective function of rolling optimization, and solves the optimal operation sequence within the rolling window at each sampling time. Compared to the static optimization, the EMPC-based scheme ensures dynamic optimality, which enables to obtain a better economic performance. To deal with measurement noise and unknown parameters, the extended Kalman filter (EKF) is employed to identify the states and parameters simultaneously by constructing an extended system, such that the precision and reliability of EMPC are enhanced. Simulation results show that the proposed EMPC+EKF strategy can effectively improve the economic performance of gold cyanidation leaching process.

To coordinate and match fluctuations of hydrogen demand in chemical production systems with the intermittent and variability of renewable energy, an optimization model for the electricity-hydrogen energy storage system coupled with a renewable energy system and a hydrogenation system in the chemical production systems was established, aiming to minimize the total cost of the energy storage system and to determine the optimal capacity configuration and power dispatching scheme of the system. The design and operation characteristics of the energy storage system were analyzed and discussed under the different configuration of the system and the penetration rate of renewable energy. The results show that when the hydrogen demand of the chemical production system is completely satisfied by the renewable energy, the energy storage combined by battery system and hydrogen tanks can effectively reduce the total cost of the system. In the electricity-hydrogen storage system, the configuration of the battery can effectively reduce the maximum load of the electrolyzer. The battery is used to stabilize fluctuations in the supply and demand in a short term, whereas the hydrogen tank is used to cope with mismatches between the supply and demand in a long-term. The total cost of the system increases dramatically with the increase of the penetration rate of renewable energy. To maintain the stability of the purchased hydrogen flow rate, the capacity of the electrolytic cell and the energy storage system needs to be increased in the system to solve the fluctuation and mismatch between the power generation side and the load side.

Based on the integration of the hydrogen network and the characteristics of the AB₅ hydrogen storage materials LaNi_4.75Fe_0.25 and LaNi_4.85Al_0.15, the application of hydrogen storage purification in the hydrogen network was studied. According to the relationship between the hydrogen-saving capacity of unit hydrogen storage material and the purification parameters, and that between utility saving and purification parameters, the optimal purification feed purity, the maximum utility saving, the amount of hydrogen storage material and the hydrogen absorption time can be determined. The proposed method is used to optimize the hydrogen network and hydrogen storage purification unit of a refinery. The results show that, in this refinery, the optimal purification hydrogen feed purity is 70% and the utility consumption can be reduced by 23.72%; more hydrogen can be saved when LaNi_4.85Al_0.15 is used in the purification unit instead of LaNi_4.75Fe_0.25, and the amount of LaNi_4.85Al_0.15 demanded in the purification unit is 991.26 kg.

The heat exchanger network plays important roles in energy recovery, energy saving and consumption reduction of chemical processes. The traditional heat exchanger network synthesis method optimizes parameters and structures with economic goal, which leads to strong coupling for process units and increases control difficulties. Either a single heat exchanger device or the entire network is inevitably interfered by uncertain factors in the practical chemical processes, thus, it is important to synthesize a controllable heat exchanger network that can withstand the disturbance caused by uncertain factors. In this paper, a synthesis method of controllable heat exchanger network is proposed with the consideration of the interaction between the network structure and the controllability of the heat exchanger network. The influences of decision variables on network structure are explored based on multi-scenario optimization. Then the flexibility index is used to measure the feasibility of the system, and the heat exchanger network is further optimized to resist uncertainties. Finally, a case study is carried out to verify the feasibility and effectiveness of the proposed method.

Hazardous gas release will threaten social security and cause damage to property and environment. Establish an effective emergency response system with monitoring devices can help reduce the potential risk and provide evaluation for release accident emergencies. The gas concentration is detected by the gas concentration sensor arranged in the risk area, and the leakage assessment is carried out according to the meteorological conditions and the forward diffusion model. Because the data of leakage scenarios is not easy to obtain, the large eddy simulation method based on Fires Dynamics Simulator (FDS) is used in this paper to generate date through simulating release scenarios with turbulent condition. On this basis, the impacts of the cost function to the modeling process and source term estimation are studied. The results of source term estimation shows the function of the square sum of concentration deviation is better than other functions. However, the function cannot avoid the condition that the estimation index of some scenes is far from the average value because of turbulent flow. In addition, a combination of the estimation results of different cost functions can improve the situation to a certain extent.

A three-dimensional unsteady coal gasification model of BGL gasifier is established in this paper. The model considers the shrinkage process of coal particles, the coal core pyrolysis model, gas phase turbulence model, gas-solid flow model, gas-solid heterogeneous reaction model, gas phase homogeneous reaction model, energy conservation equation and phase-to-phase heat transfer model. This model fully considers the temperature and composition distribution of the three-dimensional space inside the gasifier. Through the optimization of the stoichiometric parameters of the coal pyrolysis section model, the CO/H₂ molar ratio is about 1.59, which is consistent with the BGL furnace pyrolysis section operating mechanism; The three-dimensional unsteady simulation of the gasification section of the gasifier is used to simulate the composition of the outlet gas (CO, H₂, CO₂, CH₄, H₂O, O₂) compared with the literature results, and the error is less than 4%, which proves the accuracy of the BGL model. Based on this model, we analyze the main parameters of the coal gasification process in this paper. The results show that the coal gasification efficiency increases with the increase of steam-oxygen ratio. When the steam-oxygen ratio is determined to be between 1—1.3, it can meet the process and production requirements, which is suitable for the characteristics of lignite used in this paper; the coal gasification efficiency will decrease with the increase of oxygen-coal ratio, but the content of effective gas in syngas shows a trend of increasing first and then decreasing. When the oxygen-coal ratio is around 0.17, the effective gas content will reach a peak value. With the increase of coal particle diameter, the temperature in BGL furnace decreases, the maximum temperature is dropping from 2536.77 K to 2047.81 K; as for the composition, the increase of coal particle diameter will reduce the production of CO, H₂ and CH₄, and increase the amount of CO₂.

For the optimization of the heat exchange network, not only the“quantity”of energy recovery, but also the“quality”dissipation problem in the energy recovery process must be considered. On the premise of the maximum energy recovery(MER) of the heat exchanger network, based on the entransy dissipation theory in the process of irreversible heat transfer, aiming at the highest entransy efficiency which represents the energy recovery quality, we establish a multi-objective mixed integer nonlinear programming (MOMINLP) model, which comprehensively consider the quantity and quality of energy recovery and the economics of the heat exchanger network. According to the characteristics of the model targets, the heat exchanger network is optimized step-by-step based on the ε-constraint algorithm, the maximum energy recovery and minimum total annual cost (TAC) of the heat exchanger network are accurately solved in turn with the BARON software. Then, with the solved results multiplied by the relaxation coefficient ε_i as the constraint condition for narrowing the search area, the Pareto frontier of the entransy efficiency under the constraints of energy and economy is obtained. Through calculating the classic 10SP1 case, the optimal solution as the minimum ratio of economy to efficiency is obtained when the relaxation coefficient is 1.05, and compared with the results of other literature optimization methods, the multi-objective constrained optimization method in this paper can find the comprehensive optimal solution faster. Finally, the entransy efficiency of different optimization schemes is compared through the heat exchanger networks composite curve in T-Q diagram, which indicates that divide the heat exchanger network into internal part and surplus part firstly, and then optimize the scheme according to the economy and entransy efficiency, this method can reduce the irreversible loss of internal part and increase the temperature of the surplus part.

A reactive distillation (RD) process for the production of ethanolamine from ethylene oxide (EO) and ammonia catalyzed by water was presented. The feasibility and maximum selectivity of different products selectivity in a reactive distillation column were discussed. Aspen Plus software package was used to simulate the process. The effects of operating pressure, water feed flow, molar feed ratio of NH₃ to EO, boilup ratio and EO feed location on the RD column performance were investigated and the optimal operating parameters considering different product selectivities were obtained. The results showed that MEA selectivity could be increased with a larger NH₃/EO ratio, water feed flow and boil up ratio, while DEA selectivity could be increased with a smaller NH₃/EO ratio, water feed flow and boil up ratio. Under optimum conditions, the selectivity of MEA and DEA could reach 70.30% and 41.89%, respectively. The studies indicated that the production of ethanolamine by RD had more operational flexibility and obvious advantages compared with the literature methods.

This paper proposes a lexicographic economic model predictive control (MPC) without terminal constraints for multi-objective optimization of circulating fluidized bed boiler (CFBB) combustion systems subject to nonlinearity, constraints, and multivariable coupling. By the idea of lexicographic multi-objective optimization, the hierarchical receding horizon optimization control problem is formulated for the CFBB combustion system, where stabilization of the combustion process is viewed as the most important control objective and combustion economic performance is viewed as the second important control objective. By designing the terminal region condition of the stability objective function, the stabilizing lexicographic economic MPC scheme with no explicit terminal constraints is established for the CFBB combustion system. This not only reduces the online calculation amount of the multi-objective combustion controller, but also realizes the stable control and economic performance optimization of the CFBB combustion system in parallel. Finally, the validity of the proposed method is verified by simulation comparison.

To make the effluent total phosphorus reach the real-time standard in wastewater treatment process (WWTP), an effluent total phosphorus control strategy, based on fuzzy neural network (FNN), is proposed to control the biochemical phosphorus in this paper. First, the manipulated variables, based on the mechanism analysis of biochemical phosphorus removal process, were considered as the external carbon (EC) and dissolved oxygen (DO) transfer coefficient. Second, an FNN-based process controller was designed to control the effluent total phosphorus. And a gradient descent algorithm was applied to adjust the parameters of controller. Finally, the proposed FNN-based process controller was tested on the benchmark simulation model No. 1 (BSM1) to evaluate its effectiveness. The results demonstrated that the proposed FNN-based process controller can guarantee the standard discharge of effluent total phosphorus. The results show that the FNN-based effluent total phosphorus controller can ensure that the total effluent total phosphorus is discharged and has a good control effect.

Copper removal process(CRP) is an important step in hydrometallurgical zinc process. Due to the changeable production conditions, diverse mineral resources and complex mechanisms, there exist uncertainties in the process of copper removal, which affect the stability and reliability of production. In this paper, the chance constrained optimization for the copper removal process under uncertainty is studied, considering the uncertainty of the inlet flow rate, the returned underflow rate and the inlet copper ions concentration. Firstly, the uncertainties of the copper removal process are analyzed, and the distribution characteristics of the uncertain parameters are analyzed by using statistical methods. The idea of chance constraint is introduced, and the problem of the copper removal process under uncertain conditions is formulated as a chance constrained optimization problem. Then the chance constrained optimization problem is transformed into a nonlinear programming problem by the back-mapping approach(BMA). Finally, sequential quadratic programming is used to solve the resulting problem. Monte Carlo simulation verify the effectiveness of the proposed method and demonstrate that the robustness of the system is significantly improved.

17 generalized alpha functions for Peng-Robinson equation of state (PR EoS) were evaluated using the vapor pressures of 11 kinds of non-polar, weakly polar and polar compounds. The estimative abilities of the generalized alpha functions were compared according to the average absolute deviation of each kind of compounds. The results indicated that the vapor pressures of alkanes, aromatic hydrocarbons, gases and halogenated hydrocarbons can be accurately predicted with Robinson-Peng (1978), Wang (2004) and Li-Yang (2011) alpha function. However, the predictions of vapor pressures for weakly polar and polar compounds, such as alcohols, ethers, esters, acids, and water, with Robinson-Peng (1978), Wang (2004) and Li-Yang (2011) alpha function were less accurate than that predicted with Forero (2016) alpha function. In conclusion, the generalized alpha functions with acentric factor and polar factor as variables are more accurate than that generalized only with acentric factor for the prediction of vapor pressures of polar compounds, such as alcohols, acids and water.

The optimization of batch chemical production scheduling under uncertainty is a challenging topic in the study of production scheduling problems. A robust optimization model based on mixed integer linear programming (MILP) is proposed to optimize production scheduling decisions under uncertain conditions. Considering the operating costs and raw material costs in the production process, a deterministic mathematical model with the target of net profit maximization was constructed. Then three uncertainties, demand, processing time, and market price were considered. The robust optimization model which could adjust the degree of conservatism was established and transformed into a robust counterpart model. The example results showed that the robust optimized batch production scheduling model had lower profit than the deterministic one, but the reliability of the scheduling scheme was enhanced with more production tasks and less equipment idle time, which achieved production operational and economic optimization under uncertainty.

The statistical model based on the kernel entropy principal component analysis method only uses the data under normal operating conditions for modeling, and ignores some known categories of previous failure data in the monitoring system database. This paper proposed a fault discriminant enhanced the kernel entropy component analysis algorithm by using fault information contained in the previous fault data to enhance the fault detection performance. The model adopted unsupervised learning and supervised learning methods to monitor two types of data features: nonlinear kernel entropy component and fault discriminant component. In addition, the proposed algorithm employed Bayesian inference to convert the corresponding monitoring statistics into failure probability. The weighted probability of the two sub-models generated the monitoring statistics of the overall probability. Through numerical simulation and Tennessee Eastman process simulation experiments, the FDKECA algorithm is more effective than the traditional KECA since the improvement of the fault detection rate.

Aiming at the problem that the load of the ball mill is difficult to accurately judge during the grinding process, a ball mill load identification method based on improved empirical wavelet transform (EWT)-multi-scale entropy and kernel extreme learning machine (KELM) is proposed. Firstly, according to the diversity and complexity of cylinder vibration signal, the EWT spectrum segmentation method is improved. By constructing the signal simulation model, the decomposition effect of EWT, EMD is compared, and the effectiveness of the method is proved. Secondly, the intrinsic modal function (IMF) is obtained by using the improved EWT algorithm to decompose the vibration signal of the cylinder under different load states, and then the effective IMF component is selected for reconstruction by correlation analysis. Finally, the multi-scale entropy of the reconstructed signal is extracted as the eigenvector to characterize the different load states of the mill, and the mean value of multi-scale entropy deviation is calculated. The results show that there are obvious differences in the mean value of multi-scale entropy and multi-scale entropy of the three load signals, and the relationship between the three load signals is underload > normal load > overload. The extracted multidimensional feature vector is normalized and used as the input of KELM and the load state of the mill is used as the output. The kernel target alignment (KTA) algorithm is used to optimize the kernel parameters and the optimal model of mill load state identification is established. The feasibility of the method is verified by grinding experiments. Compared with SVM, the overall recognition rate of EWT- is 3.4% higher, and for EMD-multi-scale entropy, EWT-multi-scale entropy is increased by 12.3% and 8.9%, respectively.

For the complex industrial process, the lack of labeled samples under multiple working conditions cannot be used for soft sensor modeling, and the original model is out of alignment. In this paper, an unsupervised soft sensor modeling method combining locally linear embedding (LLE) and geodesic flow kernel (GFK) is studied. The method firstly extracts the common feature between each working condition by locally linear embedding, and then projects the common feature of the known working condition data and the unknown working condition data into the manifold space, and uses the geodesic flow kernel frame on the manifold space for domain transfer. Finally, the soft-measurement model is established by partial least squares regression method, and the soft-measurement value of the dominant variable is obtained. The practicality and effectiveness of the proposed algorithm are verified by soft sensor of component variables under different working conditions in TE process and soft sensor results of ball mill load parameters under different working conditions.

At present, most graphic tools are only suitable for the design or retrofit of heat exchange networks (HEN) with the goal of saving energy. The increase of energy recovery is often accompanied by the increase of the number of heat exchange units, and the number of heat exchange units obviously affects the total equipment costs. A new method for the retrofit of heat exchanger networks by using heat exchanger load diagram (HELD) is proposed. Based on pinch analysis, selecting targeted matches and reconstructing the HELD, the problem is simplified. By shifting the heat flow curve horizontally, those cross-pinch heat loads are re-matched to reduce utility consumption, and the number of heat exchangers in the retrofitted heat exchanger network can be reduced as much as possible with the help of some heuristics. An industrial case of paper mill is retrofitted to demonstrate the effectiveness of the proposed approach. Compared with reported solutions, two new retrofit networks with similar energy saving values and fewer heat exchangers are obtained.

The aging of power batteries is likely to cause problems such as inaccurate state of charge, difficult screening of retired batteries, and thermal safety, which limit the development of electric vehicles and the use of old batteries. Furthermore, it may offer theoretical foundation to solve the above problems by studying on the aging characteristics and analyze the variation of electro-thermal performance between fresh and aged batteries in a variety of environments or conditions by comparison. Therefore, it is researched and discussed based on the power battery test system through experiments. It is found that effective capacity, energy and energy efficiency reduces gradually as well as the increase of cycle numbers. Meanwhile, there is a slight swelling phenomenon accompanied by aging at the bottom section. The rate characteristics and the adaptability of ambient temperature become worse with the increase of aging degree. Moreover, the resistances and temperature sensitivity of resistances increase obviously. The variation of electrical performance affects its thermal performance directly so that working temperature and temperature gradient of battery increase obviously with comparison of fresh battery. The difference of working temperature between fresh and aged batteries is more serious especially for discharging under low ambient temperature or by high rate.

In view of the environmental pollution caused by SCN^- accumulation and efflux in the process of cyanide waste water recycling, CuSO₄ was used as a precipitant to treat cyanide waste water bearing SCN^-. The effects of precipitant dosage, precipitation time and precipitation temperature on SCN^- precipitation effect and coexisting ion concentration were investigated. The results show that the SCN^-precipitation rate is up to 87.5% when 1.6 times the chemical reaction amount of CuSO₄ is added into it at the room temperature for stirring 60 min, the concentration of T_Cu, S₂O₃ ^2-and SO₃ ^2-is sharply reduced. At the same time, the purity of CuSCN products with more than 97% is obtained. Part of SO₄ ^2- is precipitated after the waste water only need neutralized by lime, its activity is further restored and the treated water could be directly returned to the concentrate leaching process and the remaining CN^- in it is fully used, the toxic cyanide water recycling is realized.

By coupling the chemical agglomeration and turbulent agglomeration technology, the agglomeration and removal characteristics of coal-fired fine particles under the coupling effect of chemical and turbulent agglomeration, as well as the influence of particle concentration, flue gas temperature, flow rate of chemical agglomeration solution and flue gas velocity on them were studied experimentally. The results show that the chemical turbulence coupling agglomeration can further promote the agglomeration and growth of fine particles and the removal of fine particles by electrostatic procipitator under typical conditions, and its effect is better than that of the chemical turbulence agglomeration alone. With the increase of fine particle concentration, the agglomeration and removal efficiency decreased gradually, from 49.2% and 96.7% to 35.3% and 88.2% respectively. With the increase of flue gas temperature and flow rate of chemical agglomeration solution, the agglomeration and removal efficiency of fine particles first increased and then decreased, and they reached the maximum value at 180℃ and 12 L/h, which were 44.7% and 94.8%. With the increase of flue gas velocity, the agglomeration and removal efficiency of fine particles increased gradually, from 30.5% and 86.3% to 50.2% and 97.5% respectively.

In the reaction of catalytically activated peroxymonosulfate (PMS) degradation of pollutants in water, the reaction efficiency was improved by adding cobalt-based perovskite. The degradation effect of LaCoO₃/PMS on naproxen（NAP） was evaluated by different system experiments. The effects of LaCoO₃ dosing, PMS dosing, initial pH, Cl^- concentration and humic acid (HA) on NAP removal rate and mineralization ability of the system were investigated. The degradation rate of NAP increased with the rising PMS concentration and increasing LaCoO₃ dosage. The best removal efficiency was obtained at pH 5.0. Moreover, the presence of Cl^- in solution slightly promoted the degradation of NAP, however, humic acid (HA) inhibited the reaction to some extent. LaCoO₃ still has good stability after five reuses. In addition, the radical quenching test showed the \mathrm{S}{\mathrm{O}}_{\mathrm{4}}^{·-} was the main radical specie of LaCoO₃/PMS reaction.

The response surface methodology based on center composite design is used to predict the optimum experimental conditions for electrochemical degradation of dye wastewater. The impacts of pH, voltage, electrode spacing and their interaction on the electrochemical decoloration of methylene blue solution were designed and analyzed by using Ti/SnO₂-Sb as the anode. The accuracy analysis gets R²=0.9942, indicating that the selected mathematical model is well fitted. The response surface method can be used to predict and optimize the electrochemical degradation of dye wastewater process conditions. The best condition for the optimized decolourization ratio of 98.68% is as follows: pH is 6.98, the voltage is 6.0 V, and the electrode spacing is 1.01 cm. The following experiment shows that the predicted results are reliable and the practical decolourization ratio reaches 98.47%. This method can be used to optimize the process parameters of electrocatalytic degradation of dye wastewater and provide an optimization scheme for the actual wastewater treatment.

A high boron removal reverse osmosis membrane was prepared by modifying the polyamide reverse osmosis membrane by the“swelling-embedding-shrinking”method. Swelling by methanol increases the distance between the polymer chains, providing a place for the embedding of fatty acid (decanoic acid) molecules. Under the action of pressure and concentration polarization, the modified molecules are selectively embedded in the pores of the polyamide membrane. When the methanol molecules leave, the polyamide membrane shrinks to immobilize the decanoic acid molecules in the polymer network. This experiment increases the boron removal rate of the reverse osmosis membrane by increasing steric hindrance and reducing polarity (reducing hydrogen bonding sites). The experimental results show that the boron removal rate and the salt rejection rate of the modified membrane are both increased. The salt rejection rate is increased from 90.36% to 96.46%, and the boron removal rate is increased from 47.85% to 77.32%. In addition, the boron concentration in the permeate is less than 2.4 mg/L to meet emission standards (WTO). Although the permeability of water and boron decreased, the permeability selectivity of water and boron increased, which proves that the method is advantageous for increasing the selectivity of water and boron.

The technology of digital inkjet printing requires inkjet printing dye with high purity and low salt. Based on preparation of GO / nano TiO₂ self-assembled nanofiltration membrane, the size of nano TiO₂ particles and the blending ratio with GO were explored. The optimal GO/TiO₂ composite nanofiltration membrane was obtained in which the TiO₂ particle size was 60 nm and the blending ratio with GO was 1∶1. The pure water permeance was 10.69 L/(m²·h·bar), the rejection of NaCl and Na₂SO₄ was 12.6% and 15.7% respectively, and the rejection of Eriochrome black T, Congo red and Coomassie brilliant blue R were all higher than 99%. The home-made continuous constant volume diafitration device was used to conduct the dye desalination and concentration experiment. The concentration of obtained ink was increased from the initial 2.0 g/L to 9.74 g/L, and the concentration of NaCl and Na₂SO₄ was reduced from the initial 10 g/L to 5.3 mg/L and 11 mg/L respectively, meeting the requirements of digital inkjet printing dye.

Mg-Al layer double hydroxides (Mg-Al LDHs) modified asphalt was prepared by melt blending with Karamay asphalt and waste crumb rubber (CR), which has UV aging resistance. The physical properties including softening point, penetration index (PI), ductility were tested, and the three indicators were “normalized”into overall desirability by Hassan s mathematical method. The Box-Behnken mathematical relation model between overall desirability and affecting factors was established, and the technology of preparing composing LDHs/CRMA was optimized. The optimum preparation conditions were: shear temperature 173℃, shear time 89 min, and shear rate 3500 r/min. The aging resistance of composite modified asphalt was evaluated by UV aging simulation test. The FT-IR and SEM revealed that the introduction of LDHs can reduce the production of oxygen-containing functional groups during asphalt aging, and the anti-UV aging performance was improved.

Using BaCl₂, TiCl₄ and NaOH as raw materials, BaTiO₃ particles with nanometer spherical morphology were prepared by the method of microchannel continuous precipitation. The effects of barium concentration, volume flow rate, reaction temperature and flow ratio of two branches on the morphology and particle size of the products were investigated. And the products were compared with those obtained by batch precipitation. The results show that spherical BaTiO₃ particles with uniform particle size distribution can be prepared under the condition of 90℃, 0.15 mol/L Ba concentration, 3 mol/L alkali concentration, and 2 ml/min of each branch volume flow. The particle size of prepared powders was 80－150 nm. The obtained BaTiO₃ diffusion layer was applied to the creatine kinase dry sheet, and the repeatability was good. Besides, the apparent concentration gradient was showed, and the relationship between the light reflectance value and the concentration was corresponding to the reaction phenomenon. The results have high reliability, indicating that the BaTiO₃ particles obtained by the microchannel continuous precipitation method are feasible to be applied in medical test dry film.

Poly(vinylidene fluoride)(PVDF) has been considered as one of promising membrane materials due to its excellent properties such as high mechanical strength, chemical stabilityetc. and it is widely applied in the preparation of microfiltration and ultrafiltration membranes. Unfortunately, its hydrophobic nature is susceptible to fouling and remove the foulants with frequently cleaning process may pose irreversible damage to the membranes resulting in reduced membrane lifetime. Therefore, to improve the hydrophilicity and anti-fouling property, a thermo-responsive inorganic-organic nanofiber additive (palygorskite-g-poly(N-isopropylacrylamide), PGS-g-PNIPAM) was first fabricatedvia “graft from” technic and then was incorporated into PVDF matrix. 3-Methacryloxypropyl trimethoxy silane (MPS) was used for introducing ethylene groups onto the PGS surface. A series of thermo-responsive PGS-g-PNIPAM, with a wide range of grafting yields, were prepared by grafting poly(N-isopropylacrylamide) (PNIPAM) onto PGS-MPS surfaces with the free radical graft polymerization. A series of characterizations including XPS, SEM and TG were conducted and the results indicate that PNIPAM were successfully grafted onto PGS surface and the composite nanofiber owns thermo-responsiveness; the graft ratio gradually reached saturated state as the NIPAM amount increased. The PVDF/PGS-g-PNIPAM hybrid membranes were preparedvia a combination of immersion precipitation (IP) and thermally induced phase separation (TIPS) process. PGS-g-PNIPAM was incorporated into PVDF matrix and the influence of graft ratio, on membrane structures and properties was investigated in detail. The results show that the thermo-responsiveness of membranes ascended firstly and then descended lastly as the graft ratio increased and reached the maxium value at the graft ratio of 21.33% for the PGS-g-PNIPAM. The PGS-g-PNIPAM nanofibers, which uniformly dispersed in the PVDF matrix, acted as nucleating seeds and expedited the crystallization process of PVDF. As the increasing graft ratio of the PGS-g-PNIPAM, the maximum pore size, mean pore size, pure water flux (PWF) and hydrophilicity of PVDF/PGS-g-PNIPAM hybrid membranes increased. The hybrid membranes exhibited superior bovine serum albumin (BSA)-fouling resistance.

Cobalt carbonate is a typical type of conversion anode lithium battery material, with rich resources, high capacity, safe and reliable and other advantages, but there are some unresolved problems, such as poor conductivity, at the same time in the lithium ion embedding and excitability process of serious volume changes. In this paper, the different components of Zn doped Zn_xCo_1-xCO₃ by hydrothermal method, by adjusting the quality of Zn and Co raw materials to control the molar ratio of Zn/Co, the study shows that when molar ratio of Zn and Co is 0.3∶0.7, doping products have good cycle and multiplier performance, zinc ion doping lithium ion conductivity. The study of Zn_xCo_1-xCO₃ (x=0.12,0.3, 0.5) showed that it had both an alloy and a conversion reaction during charging and discharging process, which improves the conductivity of the whole electrode and then exhibits excellent electrochemical properties.

In order to study the characteristics of heat flow and temperature in the combustion process of aviation kerosene tank fire, an oil combustion simulation bench was set up. By analyzing the experimental images and data, it is found that the radiant heat flux of kerosene pool fire decreases gradually with the increase of radial distance and height, in which the radiant heat flux is more sensitive to the increase of height and decreases more rapidly with the increase of height. The heat radiation reached the maximum intensity before the heat convection, and then there was a jump in the heat convection intensity in the later stage of the stable combustion, and the intensity exceeded the heat radiation intensity and became the main heat transfer mode in this stage. On the central line of the kerosene pool fire, the flame in the lower area burns continuously with low oxygen concentration, while the oil fire in the upper area absorbs air and has high oxygen concentration, which is the main reason for the difference in the time when the highest temperature at different heights arrives. For different sizes of oil pool fires, the peak heat flux increases with the size of the oil pool, and the peak heat flux of the square oil pool is significantly higher than that of the round oil pool.