基于BP-NSGA-Ⅱ优化的高速电梯轿厢水平振动变论域模糊PID控制OA北大核心CSTPCD
Variable Universe Fuzzy PID Control of Horizontal Vibration of High-speed Elevator Cars Based on BP-NSGA-Ⅱ Optimization
针对影响高速电梯乘坐舒适性和安全性的轿厢水平振动问题,提出一种基于反向传播(Backpropagation,BP)神经网络和非支配排序遗传算法-Ⅱ(Non-dominant Sorting Genetic Algorithm-Ⅱ,NSGA-Ⅱ)的变论域模糊PID控制方法.首先建立基于达朗贝尔原理的轿厢动力学模型,其次在传统变论域模糊PID控制的基础上建立以量化因子作为输入,轿厢水平振动加速度均方根和位移均方根作为输出的BP神经网络模型,最后将该模型作为NSGA-Ⅱ算法的适应度函数,通过NSGA-Ⅱ算法优化量化因子来提高系统控制精度.仿真分析结果表明:基于BP神经网络和NSGA-Ⅱ算法的变论域模糊PID控制方法对轿厢水平振动的抑制效果优于变论域模糊PID控制方法.
Aiming at the problem of horizontal vibration of the cars that affects the comfort and safety of high-speed elevators,a variable universe fuzzy PID control method based on backpropagation(BP)neural network and non-dominant sorting genetic algorithm-Ⅱ(NSGA-Ⅱ)is proposed.Firstly,the dynamics model of the car is established based on D'Alembert's principle.Then,based on the traditional variable universe fuzzy PID control,a BP neural network model with quantization factor as input and root mean square of horizontal vibration acceleration and displacement of the car as output is established.Finally,the model is used as the fitness function of NSGA-Ⅱalgorithm,and the quantization factor is optimized by NSGA-Ⅱ algorithm to improve the system control accuracy.The simulation analysis results show that the variable universe fuzzy PID control method based on BP neural network and NSGA-Ⅱ algorithm has better suppression effect on the horizontal vibration of the car than the original variable universe fuzzy PID control method.
陈岁繁;杨松;李其朋
浙江科技学院 机械与能源工程学院,杭州 310023
土木建筑
振动与波变论域模糊PID控制量化因子BP神经网络NSGA-Ⅱ算法
vibration and wavevariable universe fuzzy PID controlquantification of the factorBP neural networksNSGA-Ⅱ algorithm
《噪声与振动控制》 2024 (002)
63-69,81 / 8
国家重点研发计划资助项目(科技助力经济2020)(SQ2020YFF0423771);浙江省科技计划资助项目(2020C01053)
评论