Research On Real-time Estimation And Optimal Control Of Automatic Diaphragm Clutch Engagement Torque

In this paper, with automatic diaphragm spring clutch system applicated in hybrid electric bus as the research object, and aiming at the special requirements of coaxial parallel hybrid electric vehicle clutch control, the real-time estimation algorithm of automatic diaphragm clutch engagement torque and based on the linear quadratic theory for the clutch torque optimal tracking control method are proposed and designed, providing a theoretical basis for HEV automatic clutch torque control strategy research.Based on the dynamic analysis of HEV power transmission system and automatic clutch, the dynamic model of HEV automatic clutch and power transmission system is established. The mathematical model of the clutch actor is established based on mechanism analysis of the actuator of the mechanical and electrical clutch.Considering axial elastic characteristics of the driven plate and clutch cover load characteristic and separation characteristic as well as empirical formula of the friction coefficient, the clutch torque characteristics calculation model is established using release bearing travel as the input, the clutch torque as output, and bench test was carried out.Based on dynamics equation of clutch driven disc and clutch disc, with adjacent engine (or motor) speed and clutch transfer torque as the state vector, state space model of clutch torque estimation is designed and clutch transmission torque estimation algorithm is proposed based on Kalman filter, performed by software programming in the MATLAB/Simulink software. The clutch transmission torque is estimated under the HEV starting and mode switching, compared with simulation values. The error analysis shows that the method is feasible and effective.Based on analysis of clutch actor and torque control characteristic, piecewise linearization method of clutch travel and torque characteristics is proposed and the establishment of and the system state space model is established with clutch transmission torque and the actuator parameters as state vector. Combined with the Kalman filter and the linear quadratic optimal control method, clutch torque controller is design. Studying the effect of the weight matrix of quadratic performance index function on the clutch torque control,The weighted matrix is optimized by genetic algorithm.Based on the mathematical modeling of the automatic clutch, clutch torque estimation and linear quadratic optimal control algorithm, the simulation platform of HEV automatic clutch based on Matlab/Simulink is developed, carrying out simulation study on vehicle starting and pure electric drive/engine driving mode switching process. The optimal feedback gain matrix was simulated by trial and error and genetic algorithm, comparing simulation value with the expected value of the clutch transmission torque were compared and analysising the error characteristics of the two methods. The results show that the designed method of linear quadratic optimal control of the clutch transmission torque is feasible based on Kalman filter real-time torque estimation, and the method can more accurately track clutch setting the target torque and further improve the control performance through the optimal feedback gain through genetic algorithm.