基于遗传编程的火电机组煤耗特性建模研究

Coal Consumption Characteristics Modeling Research on Thermal Power Units Based on Genetic Programming

遗传编程算法能够对非线性系统的结构和参数进行同步辨识,是进行全局最优搜索的智能算法。对火电机组煤耗量影响因素进行数据预处理并进行相关性分析,找出对煤耗量影响大的因素。采用多目标遗传编程算法对火电机组煤耗特性曲线模型进行辨识,进化目标为原始适应度、表达式复杂度以及最大偏差的和最小。仿真结果说明,遗传编程对煤耗特性曲线模型的辨识是有效的,非常适合解决非线性系统建模问题,并在算法上实现了结构辨识与参数辨识的统一,具有实用性。

In the paper, genetic programming algorithm was applied to identify both structures and parameters of nonlinear system, which is an intelligent algorithm for optimal searching. Thermal power units coal consumption was affected by many factors, data were processed in order to make a correlation analysis. Multi - objective genetic programming algorithm was used to identify the model of coal consumption curve, which searching aim is to minimize the sum of the original fitness, the complexity and the maximal deviation. The results indicate that the genetic programming algorithm can search out the structures and parameters at the same time which is both valid and practical for nonlinear system.