CAJ | 학술논문

针对开关磁阻电机转矩脉动大、非线性严重和数学模型难建立等问题，将单神经元自适应PID控制和传统PI控制相结合，构成了开关磁阻电机直接转矩控制系统。根据预设定的误差限，利用神经元自学习能力消除系统较大的误差信号，小误差信号通过传统PI进行消除，使得系统能够更好的适应各种扰动。最后，在Matlab／Simulink环境下，对四相开关磁阻电机进行了仿真，仿真结果表明，复合控制方法能够有效的减小转矩脉动，提高系统的动、静态性能。
침대개관자조전궤전구맥동대、비선성엄중화수학모형난건립등문제，장단신경원자괄응PID공제화전통PI공제상결합，구성료개관자조전궤직접전구공제계통。근거예설정적오차한，이용신경원자학습능력소제계통교대적오차신호，소오차신호통과전통PI진행소제，사득계통능구경호적괄응각충우동。최후，재Matlab／Simulink배경하，대사상개관자조전궤진행료방진，방진결과표명，복합공제방법능구유효적감소전구맥동，제고계통적동、정태성능。
For large torque ripple , serious nonlinear problem and the mathematical model is difficult to establish , the direct torque control system was design for switched reluctance motor , which combined single neuron adaptive PID control with the traditional PI control .According to the error limit , the self-learning ability of neurons was used to eliminate a large error signal , and the small error signal was eliminated by the traditional PI control .The simulation for the four-phase switched reluctance motor was done in the Matlab/Simulink environment .The simulation results show that control strategy can effectively suppress torque fluc -tuation and obtain good system characteristics .