基于群智优化小波神经网络的机械臂路径控制研究

Research on path control of mechanical arm based on swarm intelligence optimization wavelet neural network

为了提高机械臂路径控制的准确性,采用群智能优化的小波神经网络算法对机械臂路径进行跟踪,以便实现精准有效的控制。首先分析了二连杆机械臂动力结构,然后建立基于小波神经网络的机械臂路径控制模型,根据机械臂状态变量构建粒子群,通过粒子位置更新获得稳定的小波神经网络模型主要参数。在仿真过程中通过差异化设置隐藏层节点数M和粒子群速度权重ω主要参数,实验证明,当M=12,ω=1.2时,可以获得最优的机械臂目标路径跟踪性能,角度平均误差和位移平均误差均最小,相比于小波神经网络的机械臂路径跟踪,经过了粒子群优化后的跟踪性能提升明显。

To improve the accuracy of manipulator path control,the wavelet neural network algorithm optimized by swarm intelligence was used to track the path of,so as to achieve accurate and effective control.Firstly,the dynamic structure of the two link manipulator was analyzed.And then the path control model of the manipulator based on wavelet neural network was established.The particle swarm optimization was constructed according to the state variables of the manipulator,and the main parameters of the stable wavelet neural network model were obtained by updating the particle position.In the simulation process,the number of hidden layer nodes M and the main parameters of particle swarm optimization speed weightωwere set by differentiation.Experiments showed that when M=12,ω=1.2,the optimal path tracking performance of the manipulator can be obtained,and the average angle error and displacement average error were the minimum.Compared with the path tracking of manipulator based on wavelet neural network,the tracking performance after particle swarm optimization was significantly improved.