High-speed Train Traction Motor State Estimation And Its Implementation

High-speed trains are powered by traction motors directly,which traction and braking of high-speed trains can be achieved by adjusting output torque of traction motors.Therefore,regular and efficient operation of traction motors is the key to ensuring the running safety of high-speed trains.In practical applications,the control of the traction motors requires realtime acquisition of mechanical information.Traditional method of acquisition is using sensors which installed on non-transmission shaft of the traction motor.If the sensorless method is adopted,the mechanical quantities such as motor speed and load torque can be indirectly obtained by electrical quantities through state estimation.It can reduce occupation of internal space of traction motor and the corresponding cost,and increase stability of system.Therefore,the research on state estimation of high-speed train traction motors has significant value on engineering.This paper takes high-speed train traction motor MT205 as the object,and builds its vector control system.On this basis,state estimation of the traction motor is realized by coding extended Kalman filter algorithms and cubature Kalman filter algorithms.Hardware is the carrier of real-time operation of algorithms.In practical applications,state estimation algorithms of traction motor need to be implemented on hardware.As the core part of high-speed train control,the traction control unit needs to collect the data about traction motors in real-time,which can realize real-time control of traction motors,real-time detection of system faults and supervision and protection of system.If the real-time estimation of traction motor states can be implemented on the traction control unit,the traction control unit can estimate mechanical quantities of traction motors in real-time using electrical quantities obtained in real-time,so as to apply them to other functions of the traction control unit.In order to realize real-time state estimation of traction motor in the traction control unit of highspeed train,this paper adopts the digital signal processor TMS320F28335,which has the same architecture as the traction control unit of a type of high-speed train,as the hardware carrier of the state estimation algorithm execution.And studies hardware implementation of traction motor state estimation algorithm based on extended Kalman filter.In order to verify the realized hardware function in real-time,this paper designs a hardware-in-the-loop simulation platform based on real-time simulator and TMS320F28335.The structure,connection configuration and working principle of the hardware-in-the-loop simulation platform are analyzed,and the hardware-in-the-loop simulation platform is built on the basis of the above works.Real-time state estimation experiments of traction motor MT205 are carried out with different orders of state space model of traction motor.