多算法超螺旋滑模观测器控制的电梯曳引系统

摘要

为了降低电梯制造成本的同时又不影响运行舒适度,提出了一种基于支持向量机联合粒子群算法优化的变系数超螺旋滑模观测器,用于电梯转子的转速观测.针对经典永磁同步电机模型强耦合、高阶次及非线性等特征,导致传统滑模观测器存在求解难、抖振严重等问题,建立定子电流状态观测器模型,简化求解过程,提高消抖效果.在用超螺旋滑模观测器观测转速时,引进支持向量机对观测器参数进行拟合,得到支持向量机辨识标准模型,在此基础上,用离子群算法再对支持向量机辨识标准模型寻优,有效地提高了运算速度,并降低了观测器对噪声的敏感性.最后,在实验部分验证了所提策略的有效性.%To reduce the manufacturing cost while the comfort of the elevator running was hot affected,a super-twisting observer with variable gains based on support vector machine combined with particle swarm optimization was proposed.For classical synchronous motor model existing strong coupling,higher-order and non-linear characteristics,resulting in its solution was difficult to obtain and chattering was very serious.In order to overcome these mismatches,a stator current state observer was proposed.While super-twisting observer was used to determinate speed,the support vector machine effectively reduce observer sensitivity of the measurement noise.The results of the experimentation indicate that the proposed section strategy is very effective and reliable.