复数型扩展卡尔曼滤波算法在感应电机 直接转矩控制系统中的应用

针对传统算法中存在的数字信号处理器(DSP)运算速度要求高因而容易产生较大的延迟的问题.提出一种复数型扩展卡尔曼滤波观测器(ECKF)对感应电机进行状态估计,将得到的定子磁链和电机转速应用于直接转矩控制系统中,实现感应电机的无速度传感器控制.采用感应电机复数模型进行滤波器设计可以简化感应电机状态方程的维数并有效减少滤波算法计算量.由于复数型扩展卡尔曼滤波器在实现过程中没有矩阵求逆的运算,并且与常规扩展卡尔曼滤波器相比具有更低的维数,因此DSP的运算时间得到了有效的降低,提高了滤波器状态估计的快速性.仿真和实验结果验证了所提出的复数型扩展卡尔曼滤波器有效性和可行性.

Application of extended complex Kalman filtering algorithm in direct torque control system of induction motor

In order to solve the problem of high computation speed of digital signal processing (DSP) can generate large delay in traditional algorithms. This paper proposed an extended complex Kalman filter (ECKF) to estimate the state of induction motor, the stator flux and motor speed obtained are applied in the direct torque control system in order to achieve speed sensorless control of induction motor. A complex-valued model for induction motor is adopted that can simplify the observability analysis and reduce the computation complexity of the filtering algorithm simultaneously. The DSP operation time is reduced effectively due to that the ECKF is realized without matrix inversion and has the lower dimensions of the matrices compared with the traditional extended Kalman filter, therefore, the processing time of DSP is reduced effectively and the rapidity of the filter is improved. Simulation and experimental results validate the effectiveness and feasibility of the developed method.