摘要
针对伺服控制系统负载转矩扰动的问题,基于永磁同步电机矢量控制基本原理,提出了一种负载转矩扰动辨识算法。并结合神经网络控制理论,采用两层神经网络设计了一种负载转矩扰动辨识器。采用DSP控制器搭建了伺服控制系统,同时给出了软硬件设计方法,最后针对电机转速响应特性进行了仿真实验并与传统控制方法比较。仿真结果表明:基于神经网络的伺服系统转矩扰动辨识方法能够有效地抑制负载扰动对系统的影响,提高了控制系统动态响应速度和跟踪性能。
Aiming at the problem of servo control system load torque disturbance, a load torque disturbance identifica- tion algorithm was put forward based on the principle of permanent magnet synchronous motor vector control. Combined with neural network control theory, two layer neural networks were adopted to design the load torque disturbance identifier. The servo control system was designed by DSP controller and the hardware and software design method were given at the same time. Finally the simulation experiment was conducted in allusion to motor speed response characteristics and it was compared with the results of traditional control methods. The simulation results show that the load torque disturbance identification method based on neural network can effectively restrain the load torque disturbance influence on servo system and the control system dynamic response speed and tracking performance is also improved.
出处
《微特电机》
北大核心
2015年第9期74-76,80,共4页
Small & Special Electrical Machines
关键词
负载转矩扰动
辨识器
神经网络
伺服系统
load torque disturbance
identifier
neural network
servo system
