结构不对称条件下测速发电机的卡尔曼测速算法

针对基于测速发电机、光电编码器等测速传感器的传统测速系统存在低速测量性能，实时性，对结构不对称、电磁扰动等的抗干扰性等都较差的问题，提出了一种高精度宽范围实时滤波测速方法。以爪极永磁式交流测速电机为例，详细分析了其机械结构特性和工作原理，并在结构不对称条件下建立了爪极永磁式交流测速电机测速模型，提出了一种面向工程应用、计算量小的交互双模自适应降阶无迹卡尔曼滤波算法来实时估计永磁转子转速。该算法同时运行降阶和全阶交互双模自适应算法，当采用降阶估计值保障系统实时性时，在计算耗时较长的全阶算法运行完成一次后修正一次降阶估计值，提高测速精度。仿真结果表明，提出的测速算法对于结构不对称扰动具有良好的鲁棒性，能够适用于宽范围条件下的转速高精度实时跟踪，具有一定的工程指导意义。

Kalman speed measurement algorithm for tachogenerator under structural asymmetry conditions

Aiming at the poor performance of traditional speed measurement system based on speed detectors such as tachogenerators and optical encoder that not good at low speed operation, real-time capability and anti-interference done by the storm structural asymmetry, electromagnetism and so on, a high-precision and wide-range real-time filter speed measurement method is presented. The mechanical structure characteristics and working principle of the claw pole permanent magnet AC tachogenerators were analyzed in detail. The speed measurement model of the claw pole permanent magnet AC tachogenerators was built under the condition of structural asymmetry. The proposed reduced-order adaptive unscented Kalman filter algorithm was engineering application oriented and based on the interactive dual-mode, which could achieve the real-time measurement for permanent magnet rotors and had the advantage of low computational complexity. The reduced order and full order interactive dual-mode adaptive algorithms were running parallel. While using the reduced-order estimation to ensure the real-time performance of the system, a long time full order algorithm was used to correct the reduced-order estimation and improve the speed measurement accuracy. The simulation results indicate that the proposed speed measurement method not only has the good robustness to structural asymmetry, but also has the excellent performance to track the rotation speed with high precision in real-time under wide-range condition, which has certain engineering guiding significance.