基于单神经元控制器的永磁同步电机矢量控制仿真

The Simulation of Vector Control of the Permanent-magnet Synchronous-motor Based on Single Neuron Controller

在永磁同步电机状态方程和单神经元控制器数学模型的基础上,提出基于Matlab/Simu-link的永磁同步电机和神经元控制器S函数仿真模型,并将仿真模型应用到矢量控制系统中,以代替传统的基于PID控制器的矢量控制系统。在Hebb和Delta学习算法的基础上,应用增量式的神经元控制算法实现神经元控制器的参数优化和自动调整功能并提高学习能力和自适应性。仿真结果表明了基于S函数的永磁同步电机仿真模型的可行性和采用神经元控制器的永磁同步电机矢量控制系统具有较强的自适应性和鲁棒性。

Base on state equation of the permanent-magnet synchronous-motor （PMSM） and the mathematic model of the single neuron controller, the S-function simulation model of the PMSM and single neuron controller in Maflab/simulink environment were put forward and applied into the vector control system instead for the PID controller vector control system. The increment neuron control algorithms were adopted to achieve the parameter optimum and automatic adjustment and improve the self-study capability and self-adaptability of the single neuron controller on the basis of the Hebb and Delta study rule. The simulation results show that the simulation model of the PMSM based on S-function is feasible, and the single neuron controller can enhance the self-adaptability and robustness of the PMSM vector control system.