Joint Estimation of Vehicle Mass and Road Grade Based on Recursive Kernel Density Optimization

The vehicle mass and road grade will affect the working state of the vehicle and power system and is also the basic data of intelligent vehicle control. However, the estimation effect of mass and grade is limited under the small acceleration of commercial vehicle. Aiming at the reliability and prediction accuracy of the estimation algorithm under small acceleration, a joint estimation method of vehicle mass and road grade based on recursive kernel density optimization is proposed. The estimation process is improved by combining a priori analysis and a posteriori estimation, and then the estimation results are evaluated and optimized. The priori analysis is to mine the conditions suitable for estimation from the data, locate the estimated time, and calculate the average vehicle mass to provide a reference for estimating the vehicle mass. The posterior probability analysis uses recursive kernel density estimation method to evaluate and optimize the estimated vehicle mass generated in real time, and uses the optimized result for real-time grade estimation. The results show that the joint estimation algorithm of vehicle mass and grade based on kernel density optimization has good accuracy.