Disturbance Estimation Based Model Predictive Control Method for High-speed Railway Low-frequency Oscillation Suppression

Abstract

The invention discloses a high-speed rail low-frequency oscillation suppression method based on model predictive control of disturbance estimation, which includes the following steps: A. d-q Under the rotating coordinate system, the dynamic characteristic equation of the EMU network side rectifier taking into account the parameter error; B. Define the disturbance amount, and get it according to the dynamic characteristic equation d-q The current prediction model of the EMU network side rectifier in the rotating coordinate system considering the disturbance; C. Taking the current and the disturbance as state variables to obtain the state space model of the disturbance estimator and construct the disturbance estimator; D. The estimated value of the disturbance Substituting the current prediction model in step B to obtain the current prediction value, and substituting the obtained current prediction value into the cost function to obtain the control voltage of the next sampling period; E, the control voltage is obtained through coordinate transformation α‑β Coordinate system components output control pulses through sinusoidal pulse width modulation to complete high-speed rail low-frequency oscillation suppression; the invention improves the anti-interference ability and robust performance of the control system, and the voltage variation has the smallest fluctuation in the cycle, and is insensitive to parameter changes.

Description

technical field [0001] The invention relates to a control method for a rectifier on the grid side of a motor train, in particular to a method for suppressing high-speed rail low-frequency oscillation based on model predictive control based on disturbance estimation. Background technique [0002] With the rapid development of high-speed railways, the new "AC-DC-AC" electric locomotive has been widely used in electrified railway systems due to its advantages of high power factor, high power, and large traction; the traditional "AC-DC-AC" The control methods of locomotives are mainly divided into two categories, indirect current control and DC current control; indirect current control is represented by "phase-amplitude control", and direct current control includes hysteresis current control, predictive current control and transient current control; State direct current control is currently the most used control strategy in electric locomotives and high-speed EMUs; the control e...

AI Technical Summary

Problems solved by technology

[0002] With the rapid development of high-speed railways, the new "AC-DC-AC" electric locomotive has been widely used in electrified railway systems due to its advantages of high power factor, high power, and large traction; the traditional "AC-DC-AC" The control methods of locomotives are mainly divided into two categories, indirect current control and DC current control; indirect current control is represented by "phase-amplitude control", and direct current control includes hysteresis current control, predictive current control and transient current control; State direct current control is currently the most used control strategy in electric locomotives and high-speed EMUs; the control effect of traditional linear control methods has been difficult to improve, so nonlinear control methods, such as predictive control and passive control, are introduced The control of the converter is necessary; the predictive control is accompanied by the proposed control algorithm in the complex actual production process, and is widely used because of its good robustness and good complex system control performance; the commonly used model prediction Control includes predictive current control, predictive direct power control, etc. It only considers the predicted current obtained by the predictive model to track the set value to improve the control effect, but does not consider the problem that the parameters of the model parameters and the parameters of the actual controlled object cannot completely match

Images