基于遗传Hopfield混合神经网络的多轴控制系统设计与实现

Design and Realization of Multi-axis Control System Based on Modified GA-Hopfield Neural Network

采用ASIC核心的PCI多轴控制器,通过对步进电机的运动控制及状态反馈,设计完成了八轴控制系统.在汉字雕刻路径问题上,结合遗传算法和Hopfield神经网络算法的优点,提出遗传Hopfield混合神经网络算法对汉字雕刻路径进行优化,较好地解决了雕刻过程混乱、雕刻总路径过长的问题.给出了该系统的硬件结构以及遗传Hopfield混合神经网络优化算法和仿真结果.实际应用表明,该多轴运动控制系统具有很好的稳定性和控制精度.

This paper designed an eight-axis control system by adopting ASIC as the central core, and using the PCI multi-axis controller to control the stepper motor movement through state feedback. Combi- ning the advantages of Hopfield neural network and the genetic algorithm, the control system adopted GA- Hopfield neural network to optimize the path of Chinese sculpture. This method solved the perplexity sculpture process problem and shortened the total sculpture path. This paper presented the hardware structure of the system, the GA-Hopfield neural network optimization algorithm and its simulation. The application has shown that the multi-axis motion rnnt^r,1 ＊o~ k -1 _._t.,,.