CIESC Journal ›› 2016, Vol. 67 ›› Issue (9): 3812-3816.DOI: 10.11949/j.issn.0438-1157.20151407

Previous Articles     Next Articles

Determination of optimal initial steam pressure of turbine based on relevance vector machine

LIU Chao, NIU Peifeng, DUAN Xiaolong, LI Guoqiang, ZHANG Xianchen   

  1. Key Laboratory of Industrial Computer Control Engineering of Hebei Province, Yanshan University, Qinhuangdao 066004, Hebei, China
  • Received:2015-09-23 Revised:2016-06-13 Online:2016-09-05 Published:2016-09-05
  • Supported by:

    supported by the National Natural Science Foundation of China (61573306, 61403331).

基于相关向量机的汽轮机最优运行初压的确定

刘超, 牛培峰, 段晓龙, 李国强, 张先臣   

  1. 燕山大学工业计算机控制工程河北省重点实验室, 河北 秦皇岛 066004
  • 通讯作者: 牛培峰
  • 基金资助:

    国家自然科学基金项目(61573306,61403331)。

Abstract:

In order to calculate the optimal initial pressure effectively, a heat rate forecasting model is presented based on the optimized relevance vector machine (RVM), in which the blended biogeography-based optimization based on the simulated annealing (B-BBO-SA) is adopted to optimize the parameter of RVM. Then, B-BBO-SA is employed to seek the optimal initial pressure under all loads based on the forecasting model. The comparison with B-BBO-SA-SVM show that the B-BBO-SA-RVM has better generalization abilities. In addition, there are some differences between the found optimal initial pressure and the designed one, and the found one can better ensure the turbine run safely and economically.

Key words: turbine, BBO, relevance vector machine, model, optimization

摘要:

为了有效计算最优初压,首先提出一种基于模拟退火和混合迁移的生物地理学优化算法(B-BBO-SA)优化相关向量机(RVM)的热耗率预测模型;然后在该模型的基础上利用B-BBO-SA算法寻找各个负荷下热耗率最小时所对应的主蒸汽压力。通过与B-BBO-SA-SVM进行比较,B-BBO-SA-RVM具有更好的泛化能力;另外,得到的最优初压与设计初压存在着一定的差别,它更能准确地指导汽轮机的安全、经济运行。

关键词: 汽轮机, 生物地理学优化算法, 相关向量机, 模型, 优化

CLC Number: