船舶锅炉水位PID控制器参数多目标优化整定

PID controller parameters tuning for marine boiler by multiobjective optimization

为有效提高船舶锅炉水位控制性能,提出一种基于多目标遗传算法和多属性决策(MADM)技术的PID控制器参数整定方法.将系统的跟随性能、扰动抑制性能及鲁棒性能作为目标函数,采用改进非劣分层多目标遗传算法(NSGA-II)求出Pareto最优解.使用客观赋权的信息熵法对Pareto最优解属性进行权值计算.采用逼近理想解的排序方法(TOP-SIS)进行MADM研究,给出了Pareto最优解排序.仿真结果表明,该方法通用性好,PID性能良好,适合工程实际应用.

To increase the performance of water level control in marine boiler effectively, a proportional integral differential （PID） controller parameters tuning method combining multiob- jective optimization technique with multi-attribute decision making （MADM）method was developed. Taking the accurate setpoint tracking, disturbance attenuation and robust stability performances as the objective function, the Pareto optimal solution was obtained by using non-dominated sorting genetic algorithm II （NSGA- II）. Weight calculation of Pareto optimal solution attributes was proposed by using the objective weight information entropy method. And a MADM approach is adopted by technique or order preference for similarity to ideal solution （TOPSIS） to give the ranking of Pareto solutions. Simulation results indicate that the proposed method has better generality and PID performance, which is suitable for the application of actual project.