基于神经网络的机械臂电机速度控制器的片上可编程系统

针对机器人伺服控制系统高速度、高精度的要求,介绍了一种全数字化、基于神经网络控制的直流电机速度伺服控制系统设计方案。速度控制器采用BP网络参数辨识自适应控制,并将其在一片现场可编程门阵列(FPGA)中硬件实现,同时以NiosII软核处理器作为上位机,构成一个完整的速度伺服控制器的片上可编程(SOPC)系统。试验结果表明,该系统具有较高的控制精度、较好的稳定性和灵活性。

SOPC of Robot Machinery Arm Motor Speed Controller Based on Neural Network

Aiming at the control problem of the high-speed high-precision servo control system of robot,a fully digitized design scheme of a speed servo controller of direct-current motor based on BP neural network was presented.Speed controller uses the BP neural network parameter identification,which is hardware implemented on one-chip field programmable gate arrays(FPGA),and develops a whole speed servo controller based on SOPC by using a Nios II containing internally as superordination controlling machines.Experiment results show that this system has high precision and good robustness.