Abstract: The intelligent optimization algorithms have been widely used in solving dynamic optimization problems for their ability to converge to global optimum at a specific probability without being trapped in local optimums. However,the practical industrial application fields of such random mechanism-based algorithms are restricted due to their demerits,such as slow convergence and low searching efficiency.Therefore, an improved knowledge-based evolutionary algorithm structure was presented for improving the efficiency of intelligent optimization algorithms which were used to solve dynamic optimization problems. The structure included a discretization method of candidate solutions in time and control domains,the evolution of the repository,population evolution steered by knowledge. When applied to dynamic optimization problems of four typical chemical processes,such as batch reactor,which owned distinguishing control feature,this algorithm demonstrated its competitive optimal searching ability,meanwhile verifying its satisfying convergence efficiency by using a smaller scale of population guided by knowledge and consuming less computational cost.

Key words: dynamic optimization, knowledge evolution, evolutionary algorithm, chemical processes, biochemical processes

摘要： 智能优化算法在动态优化问题的求解中，一方面可以一定的概率收敛到全局最优，避免局部极值而得到了广泛应用；但另一方面，基于随机机制的仿生智能算法也面临收敛速度慢、寻优效率较低的瓶颈，限制了其工业实时应用的场合。为此，从提高智能优化算法在动态优化问题的求解效率出发，提出了一种改进的基于知识引导的进化算法结构，主要包括候选控制策略-时域与控制域的离散策略、知识库空间的进化、知识引导的种群进化。该算法分别在批式反应器等4个典型化工动态优化问题上进行了仿真验证，计算结果表明，该方法能够以较小的种群规模通过知识的引导，以较少的计算代价找到较好的全局解，有效提高了算法的收敛效率。

关键词: 动态优化, 知识进化;进化计算;化工过程;生化过程