Simulation and experimental results verify the effectiveness of the algorithm

Posted by bearingkit on May 6th, 2021

Literature designed a robust Kalman filter and adaptive speed estimation to reduce the impact of parameter changes on system stability. The experimental results show that when the rotor inductance Lr and the rotor resistance Rr change, the new series Kalman filter is compared with the traditional EKF. The installation of sensors on the rotating shaft will seriously affect the performance of the motor. The stable suspension of the rotor is realized under the combined action of the winding magnetic field.

However, this method not only enhances the robustness of the system, but also affects the estimation accuracy when the parameters are accurate. In order to achieve decoupling control between torque and radial suspension force, it is necessary to accurately detect the rotor speed. Among China mechanical bearing manufacturers them, the EKF algorithm is not affected by the voltage DC offset, can effectively suppress noise, has high estimation accuracy and wide estimation range, so it has been widely used in motor speed sensorless control. However, the EKF algorithm has poor anti-interference ability and is sensitive to motor parameters.

The influence of the change on the speed estimation can improve the accuracy and performance of the system, reduce the calculation load of the chip, and enhance the practicality of the algorithm. The series structure of three reduced-order state equations is adopted. The proposed method was verified by simulation and experiment, and the results were compared with the traditional EKF. During motor operation, motor parameters will change with changes in temperature, frequency and load, which will affect the performance of speed sensorless, so how to improve Kalman filter The robust performance of the processor has become a current research focus.

Bearingless motors have excellent characteristics such as low cost, low cogging torque, simple structure, and wide field weakening speed range. Literature proposes an improved SRUCF filter, by introducing time-varying fading factors and weakening factors, real-time correction of the filter gain matrix and error covariance square root matrix, so that the system has stronger robustness in the state of sudden changes or load disturbances , But the algorithm is too complicated, the calculation amount is large, and it is only applicable when the motor parameters change small. Effectively reduce steady-state estimation errors and offsets to ensure stable operation of the system.

Extend the motor parameters as the state vector to be identified to the system model to realize the online detection of motor parameters. Simulation and experimental results verify the effectiveness of the algorithm. , But this method is mainly aimed at the interference in the stator voltage and state vector, and does not consider the change of the motor's own parameters.