基于粒子群算法的液压APC系统分数阶PID控制器设计
|
摘要
为提高冷轧液压伺服位置控制系统在复杂工况下的暂态性能和稳定性,提出一种径向基函数神经网络在线自适应调节分数阶PID控制算法。为提高网络精度,减少冗余隐层节点,采用带有2次变异机制的粒子群算法离线同时优化网络结构和初始参数,同时选择BP算法在线调整网络参数,使FOPID控制系统具备良好的自适应能力。仿真结果表明,该控制系统能够快速准确跟随输入信号,且能明显抑制外在干扰和系统参数扰动,控制效果优于其他对比控制算法。
For controlling the complicated nonlinear system effectively,the fractional order PID neural network controller is proposed based on particle swarm(PSO) algorithm in this paper.The controller,combined with fractional order PID control and BP neural network and PSO algorithm,uses the BP neural network to set the fractional order PID controller's parameters online and utilize PSO algorithm with two variation mechanism instead of back propagation algorithm in network training,to avoid the slow convergence speed,falling into local optimum easily and computing complex faults.The result of simulation shows that the fractional order PID neural network controller based on PSO algorithm could control the nonlinear systems effectively,moreover,control effect is better than that of neural network adaptive PID controller and neural network adaptive PID controller based on PSO optimization BP PID.
For controlling the complicated nonlinear system effectively,the fractional order PID neural network controller is proposed based on particle swarm(PSO) algorithm in this paper.The controller,combined with fractional order PID control and BP neural network and PSO algorithm,uses the BP neural network to set the fractional order PID controller's parameters online and utilize PSO algorithm with two variation mechanism instead of back propagation algorithm in network training,to avoid the slow convergence speed,falling into local optimum easily and computing complex faults.The result of simulation shows that the fractional order PID neural network controller based on PSO algorithm could control the nonlinear systems effectively,moreover,control effect is better than that of neural network adaptive PID controller and neural network adaptive PID controller based on PSO optimization BP PID.
引文
[1]刘宏民,贾春玉,单修迎.智能方法在板形控制中的应用[J].燕山大学学报,2010,34(1):1-5.
[2]PODLUBNY I.Fractional-order systems and PIγDμcontrollers[J].IEEE Transactions on Automatic Control,1999,64(8):1287-1300.
[3]LIU C,SHU T,CHEN S,et al.An improved grey neural network model for predicting transportation disruptions[J].Expert Systems with Applications,2016,45:331-340.
[4]亢克松,方一鸣,夏天,等.伺服电机驱动的连铸结晶器振动位移系统模糊自整定PID控制[J].燕山大学学报,2015,39(4):334-340.
[5]杨景明,马明明,车海军,等.多目标自适应混沌粒子群优化算法[J].控制与决策,2015,30(12):2168-2174.2.
[6]乔俊飞,韩红桂.RBF神经网络的结构动态优化设计[J].自动化学报,2010,36(6):865-872.
[7]PANCHAPAKESAN C,PALANISWAMI M,RALPH D,et al.Effects of moving the center's in an RBF network[J].IEEE Transactions on Neural Networks,2002,13(6):1299-1307.
[8]李巍华,翁胜龙,张绍辉.一种萤火虫神经网络及在轴承故障诊断中的应用[J].机械工程学报,2015,51(7):99-106.
[9]刘顺安,谢丹彤,尚涛,等.基于分数阶PID的液压变压器配流盘控制性能[J].北京工业大学学报,2013,39(10):1452-1458.
[10]DEB K,PRATAP A,AGARWAL S,et al.A fast and elitist multiobjective genetic algorithm:NSGA-II[J].IEEE Transactions on Evolutionary Computation,2002,6(2):182-197.
[11]魏立新,郑翠红,李莹,等.记忆径向基神经网络在冷轧液压自动位置系统的优化控制[J].计量学报,2015,37(1):47-52.
[12]LIU L,PAN F,XUE D.Variable-order fuzzy fractional PID controller[J].Isa Transaction,2015,55(1):227-233.
[2]PODLUBNY I.Fractional-order systems and PIγDμcontrollers[J].IEEE Transactions on Automatic Control,1999,64(8):1287-1300.
[3]LIU C,SHU T,CHEN S,et al.An improved grey neural network model for predicting transportation disruptions[J].Expert Systems with Applications,2016,45:331-340.
[4]亢克松,方一鸣,夏天,等.伺服电机驱动的连铸结晶器振动位移系统模糊自整定PID控制[J].燕山大学学报,2015,39(4):334-340.
[5]杨景明,马明明,车海军,等.多目标自适应混沌粒子群优化算法[J].控制与决策,2015,30(12):2168-2174.2.
[6]乔俊飞,韩红桂.RBF神经网络的结构动态优化设计[J].自动化学报,2010,36(6):865-872.
[7]PANCHAPAKESAN C,PALANISWAMI M,RALPH D,et al.Effects of moving the center's in an RBF network[J].IEEE Transactions on Neural Networks,2002,13(6):1299-1307.
[8]李巍华,翁胜龙,张绍辉.一种萤火虫神经网络及在轴承故障诊断中的应用[J].机械工程学报,2015,51(7):99-106.
[9]刘顺安,谢丹彤,尚涛,等.基于分数阶PID的液压变压器配流盘控制性能[J].北京工业大学学报,2013,39(10):1452-1458.
[10]DEB K,PRATAP A,AGARWAL S,et al.A fast and elitist multiobjective genetic algorithm:NSGA-II[J].IEEE Transactions on Evolutionary Computation,2002,6(2):182-197.
[11]魏立新,郑翠红,李莹,等.记忆径向基神经网络在冷轧液压自动位置系统的优化控制[J].计量学报,2015,37(1):47-52.
[12]LIU L,PAN F,XUE D.Variable-order fuzzy fractional PID controller[J].Isa Transaction,2015,55(1):227-233.