Acceleration estimation method and apparatus thereof, and locomotive motion control method and locomotive

Abstract

The invention provides an acceleration estimation method. The method comprises the following steps of converting motion of an object into a SISO linear discrete dynamic system so as to acquire a system mathematics model of the object; carrying out recursive least squares processing on the system mathematics model of the object so as to acquire a recursive least squares model; according to a sampling period, collecting a speed of the object so as to acquire a speed variation of the object in the sampling period; according to the sampling period and the speed variation, processing the recursive least squares model so as to acquire a real-time estimation value of a system parameter; and according to the real-time estimation value of the system parameter, outputting a real-time estimation value of an accelerated speed of the object. The invention also provides an acceleration estimation apparatus, a locomotive motion control method and a locomotive. By using the acceleration estimation method and the apparatus thereof, and the locomotive motion control method and the locomotive provided in embodiments of the invention, through estimation of the system parameter, the real-time estimation value of the accelerated speed is acquired and can be used to guarantee continuous approximation to a real acceleration signal; and real-time performance is high and the methods, the apparatus and the locomotive can directly take part in control.

Description

technical field [0001] The invention relates to the technical field of locomotive motion control, in particular to an acceleration estimation method and device, and a locomotive adopting the acceleration estimation method. Background technique [0002] In the motion control of objects, acceleration is an important control quantity, such as the body acceleration and wheel set acceleration of rail vehicles, etc. Acceleration signals in the control system are mainly obtained through software calculations. Acceleration calculation methods in the prior art mainly include direct differential calculation method and high-precision first-order numerical differential algorithm, but these methods are very sensitive to noise, and the calculated acceleration has a lot of noise and cannot directly participate in the control. The usual practice is to filter the original signal before calculating the acceleration. However, the filtered signal has a time lag, which affects the control perfor...

AI Technical Summary

Problems solved by technology

[0013] It can be seen from formulas (2) and (3) that the acceleration is obtained by directly dividing the velocity difference at different sampling moments by the sampling period T by using the conventional numerical differentiation method. If the velocity signal has a large noise, due to the sampling If the period T is very small, the noise will be further amplified, so the calculated acceleration must have a lot of noise

Images