基于LS-SVM和滑模变结构的无刷直流电动机混合控制

Hybrid control of brushless DC motor based on LS-SVM and sliding mode variable structure

针对一类不确定离散时间系统,通过分析趋近律中参数对系统到达稳定状态的影响,结合LS-SVM的建模优势,提出了一种基于LS-SVM趋近律的全局滑模变结构控制方法。分别用两个LS-SVM模型自适应调整趋近律中的参数δ和ε,克服了常规变结构控制方法中需要预先设定趋近律中参数的限制。通过在无刷直流电机(BLDC)位置伺服系统变结构控制中的应用,结果表明该方案使系统一开始就处于滑平面上,消除了系统抖振,而且具有良好的跟踪能力和较强的鲁棒性。

A global sliding mode variable structure control method based on LS-SVM reaching law for a class of uncertain discrete-time systems is proposed by analyzing the effect of the reaching law parameter on the system stability and using the advantage of LS-SVM modeling.Parameters and,which should be pre-determined in the conventional reaching law,are regulated adaptively by two LS-SVM models respectively.Through using this method to control the tracking servo sliding mode variable structure control system of BLDC （brushless DC motor）,the results demonstrate that the designed controller can make system state on the sliding surface at the beginning of the controlling,can eliminate system chattering and can provide the servo system with perfect tracking and stronger robustness.