Mass And Grade Identification And Gear Decision Of Commercial Vehicle Equipped With AMT

In recent years,with the development of technology,commercial vehicles continue to develop in the direction of intelligence.Commercial vehicles equipped with mechanical automatic transmission have good economy,power and comfort,which plays an important role in reducing environmental pollution.However,due to the use characteristics of commercial vehicles,their total mass and driving road grade vary widely.Therefore,the control performance of the control system which depends on the two parameters of vehicle mass and road grade will be greatly affected.Due to the complex driving conditions of the vehicle,in the research of mass and grade identification,considering the complex working conditions,and designing the identification algorithm with high accuracy and good stability,which is of great significance to formulate a reasonable and accurate shift strategy for the automatic transmission and improve the performance of the vehicle.This paper takes a light commercial vehicle equipped with mechanical automatic transmission as the research object,carries out the recognition research of vehicle quality and road grade,and explores its application in gear decision-making.The main research contents of this paper are as follows:(1)In order to establish the simulation environment for mass and grade identification research.Firstly,the commercial vehicle model is established according to the parameter information of the whole vehicle,engine and gearbox and with the help of AVL CRUISE dynamics simulation software;In order to build the gear decision-making function for the automatic transmission control module in the model,the two parameter dynamic shift law is analyzed and designed.On this basis,the gear decision-making model is designed in Simulink / Stateflow,and the signal is connected with the whole vehicle model established by AVL CRUISE to complete the establishment of the simulation environment.(2)The recursive least squares identification algorithm is analyzed and studied,and the joint estimation model of quality and grade based on double forgetting factor recursive least squares recursive algorithm(FFRLS)is established by using Matlab / Simulink;The road grade is designed by using the cycle conditions generated in cruise.Based on this,the estimation performance of the basic algorithm under the conditions of constant mass,variable mass and ramp starting is simulated and analyzed;On this basis,the problems of the original algorithm are deeply analyzed,and an improved algorithm for initializing covariance is proposed.The simulation results show that the improved algorithm solves the problem that the original algorithm can’t accurately estimate the variable mass,and can improve the convergence speed and accuracy of grade estimation after starting on the ramp.(3)Road grade estimation based on long-term and short-term memory cyclic neural network(LSTM)is studied.In order to select the real data of the road speed model based on the training data of the network,and obtain the road speed data based on the training data of the neural network of the commercial vehicle based on the depth of the tree.The experimental results show that the designed LSTM model can track the road grade effectively,and the grade estimation at the beginning of the ramp has short convergence time and high estimation accuracy.(4)The application of mass and grade identification in gear decision-making is explored.Firstly,based on the formulation method of basic dynamic shift law,the shift law and gear decision-making characteristics under the separate and joint action of quality and grade are analyzed.On this basis,the gear correction strategy based on fuzzy control is designed.Finally,FFRLS estimation algorithm and gear decision-making model are integrated.The experimental results show that the designed gear decision system based on mass and grade recognition can make reasonable gear judgment and effectively eliminate the shift cycle.