基于收敛性提升的粒子群算法及其在火电厂配煤优化研究  

Particle Swarm Optimization Algorithm Based on Convergence Improvement and Its Application in Coal Blending Optimization of Thermal Power Plant

作　　者：李前胜 解继刚[1] 关怀 陈筑 李扬 李俊 王永富[2] LI Qiansheng;XIE Jigang;GUAN Huai;CHEN Zhu;LI Yang;LI Jun;WANG Yongfu(Dalian Power Plant,Huaneng Power International Inc.,Dalian 116100,China;School of Mechanical Engineering and Automation,Northeastern University,Shenyang 110819,China)

机构地区：[1]华能国际电力股份有限公司大连电厂,辽宁大连116100 [2]东北大学机械工程与自动化学院,辽宁沈阳110819

出　　处：《控制工程》2024年第10期1849-1855,共7页Control Engineering of China

基　　金：流程工业综合自动化全国重点实验室开放课题(2021-KF-11-02)。

摘　　要：针对大型火电机组的燃煤煤种复杂、混煤掺烧决策困难的现状,以及配煤优化过程存在多种设计约束和物理约束而导致传统优化算法的寻优过程难以收敛的问题,提出了一种改进粒子群优化算法。该算法将自适应约束处理机制引入传统粒子群优化算法中,基于距离测度和自适应惩罚项对违反约束的粒子进行自适应处理,引导寻优过程实现收敛;同时,采用平滑非线性权重递减策略代替传统粒子群优化算法的定值惯性权重设置方法,防止算法的寻优过程陷入局部最优。基于现场数据的仿真结果表明,所提算法在存在多约束条件的非线性函数寻优过程中具有明显优势,能够实现不同评价指标的均衡优化。For the current situation of complex coal types and difficult decision-making of mixed coal combustion in large thermal power units,and the problem that the optimization process of the traditional optimization algorithm is difficult to converge due to a variety of design and physical constraints in the coal blending optimization process,an improved particle swarm optimization algorithm is proposed.In this algorithm,the adaptive constraint processing mechanism is introduced into the traditional particle swarm optimization algorithm,and the particles violating the constraints are processed adaptively based on the distance measure and adaptive penalty term,so as to guide the optimization process to achieve convergence.At the same time,the smooth nonlinear weight decreasing strategy is used to replace the fixed inertia weight setting method in the traditional particle swarm optimization algorithm to prevent the optimization process from falling into local optimization.The simulation results based on field data shows that the proposed algorithm has obvious advantages in the optimization process of nonlinear function with multiple constraints,and can realize the balanced optimization of different evaluation indexes.

关 键 词：数学模型 自适应约束处理 粒子群优化算法 混煤掺烧 

分 类 号：TP1273[自动化与计算机技术—控制理论与控制工程]