摘要
针对传统经验方法难以得到合理的直线电机进给伺服系统控制参数,提出一种改进的自适应遗传算法对直线电机伺服系统五个控制参数进行同步优化。在分析直线电机进给系统控制原理的基础上,采用实数编码对控制参数进行编码生成种群,根据控制需求确定了由误差积分项、超调量惩罚项、稳定时间项组成的适应度函数,为加快遗传算法的收敛速度、防止算法进入局部收敛,采用非线性变化的自适应交叉率和变异率对算法进行改进。仿真结果显示,改进的自适应遗传算法在保证系统无超调的情况下使稳定时间达到4.1 ms,优于"三环参数整定法"的39 ms和传统遗传算法的5.9ms,相比于三环参数整定法系统达到稳态所需要的时间减少了89.4%;采用改进的自适应遗传算法后正弦信号的平均相对跟随误差绝对值为24.70%,优于"三环参数整定法"和传统遗传算法的51.02%和24.77%。因此,改进的自适应遗传算法在解决直线电机伺服系统多控制参数同步优化问题时,性能优于"三环参数整定法"和传统遗传算法。
As it is difficult to obtain the reasonable control parameters of linear motor feed servo system by using the traditional methods, the improved adaptive genetic algorithm was proposed to optimize these five control parameters of the linear motor servo system simultaneously. The control principle of the linear motor feed system was analyzed, and the real number coding was adopted to encode the control parameters and generate the population. According to the control requirements, the fitness function composed of the error integral term, the overshoot punish term and the stability time term was determined, and the algorithm was improved by using the nonlinear adaptive crossover rate and mutation rate to accelerate the convergence speed of genetic algorithm and avoid plunging into local convergence. The simulation results show that the stability time of the improved adaptive genetic algorithm is 4.1 ms better than 39 ms of the "three-loop parameter setting method" and 5.9 ms of the traditional genetic algorithm under the condition of ensuring the system without overshoot. The time required to reach the steady state of the system is reduced by 89.4% comparing with the "three-loop parameter setting method". The average relative following error of the sine signal after the improved adaptive genetic algorithm is 24.70%, which is better than 51.02% of the "three-loop parameter setting method" and 24.77% of the traditional genetic algorithm. Therefore, the improved adaptive genetic algorithm is superior to the "three-loop parameter setting method" and traditional genetic algorithm in the synchronous optimization of multi-control parameters of linear motor servo system.
作者
杨泽青
张炳寅
刘丽冰
彭凯
黄凤荣
Yang Zeqing;Zhang Bingyin;Liu Libing;Peng Kai;Huang Fengrong(School of Mechanical Engineering,Hebei University of Technology,Tianjin 300130,China)
出处
《机械科学与技术》
CSCD
北大核心
2019年第12期1921-1927,共7页
Mechanical Science and Technology for Aerospace Engineering
基金
国家自然科学基金项目(51305124)
天津市自然科学基金项目(16JCYBJC19100,17JCTPJC53700)
河北省自然科学基金项目(E2017202294)
河北省重点研发计划项目(16211803D)资助
关键词
直线电机
参数优化
进给伺服系统
遗传算法
linear motor
parameter optimization
feed servo systems
genetic algorithm
