Wheel hub motor driving vehicle speed estimation method based on multi-model fusion

Abstract

The invention discloses a wheel hub motor driving vehicle speed estimation method based on multi-model fusion. The method comprises the following steps that a vehicle-mounted sensor signal is collected, filtering processing is carried out on an original collected signal, a vehicle speed Kalman estimation filter is established, self-adaptive adjustment is carried out on noise variances by combiningvehicle driving state information, and an exponential damping memory factor is utilized to correct a state vector estimation error to obtain a vehicle speed estimation value based on self-adaptive exponential weighted damping memory Kalman filtering; a vehicle body acceleration integral vehicle speed estimator is designed; and the two kinds of vehicle speed estimation models, namely the vehicle speed estimation value based on self-adaptive exponential weighted damping memory Kalman filtering and the vehicle body acceleration integral vehicle speed estimator are weighted and fused based on theprinciple of minimum total mean square error in combination with vehicle driving conditions. The wheel hub motor driving vehicle speed estimation method based on the multi-model fusion aims to achieve the effects of being good in real-time performance, high in precision and strong in applicability.

Description

technical field [0001] The invention relates to a method for estimating vehicle speed, in particular to a method for estimating the speed of a vehicle driven by a hub motor based on multi-model fusion. Background technique [0002] Vehicle speed is a state parameter that must be referred to in the study of vehicle handling stability control, and the accuracy of vehicle speed information can affect the effect of dynamics control. For traction control system (Traction Control System, TCS), electronic stability system (Electronic Stability Control, DSC), direct yaw moment control (Direct Yaw-Moment Control, DYC), brake anti-lock braking system (Anti-lock Braking System , ABS) and other commonly used active safety systems need to obtain accurate and reliable vehicle speed information as control input, and adjust the longitudinal / lateral force and yaw moment in time according to driving, braking, steering and other working conditions to improve vehicle handling stability and dri...

AI Technical Summary

Problems solved by technology

Some scholars use a nonlinear state observer to estimate the longitudinal vehicle speed. This type of algorithm generally needs to construct a nonlinear vehicle model and a dynamic tire model to complete nonlinear iterative calculations. The amount of calculation is large and the real-time performance is difficult to guarantee.

Some scholars have proposed a vehicle speed estimation method based on the integral of the longitudinal acceleration of the vehicle body, but this method is likely to cause problems such as continuous accumulation of signal noise errors due to long-term integration, which seriously affects the estimation accuracy

In addition, some researchers use algorithms such as Kalman filter and particle filter to estimate the longitudinal vehicle speed. The filter algorithm is more dependent on signals such as wheel speed, so this method has a better estimation effect when the vehicle is in a steady state. Once the road conditions are poor, or the wheels are in a large slip or locked situation, the filtering algorithm may have a large error in the vehicle speed estimation

Furthermore, the traditional Kalman filter algorithm belongs to the infinite growth memory filter. When making the optimal estimation at a certain moment, all the data before that moment must be used. What is the proportion of the old data in the filter value, while the proportion of the new time data is relatively small , when there are model errors and ignorant time-varying noise in the system, the correction effect of the new observation data on the state estimation is too small, and the influence of the error on the state estimation value cannot be effectively suppressed, resulting in error accumulation, which in turn leads to excessive filtering errors or even divergence , so the traditional Kalman filter algorithm for vehicle speed estimation is not reliable enough

Images