Research On Starting Control Strategy Based On Magnetic Powder Clutch For Vehicles

The magnetic powder clutch is one kind of the magnetic clutch. Compared with routine dry diaphragm spring clutch, it has more compact structure and better linear relationship between transmitted torque and excitation current. It is easy to control. Therefore the magnetic powder clutch is increasingly used in mini vehicles. In order to raise the research and development level on key parts and components for vehicles in our country and improve the development of the domestic automatic transmission, it is of great significance to further study engaging control of the magnetic powder clutch. Basing on the development situation of the automatic transmission control technique for vehicles, considering the need of the project and theoretical research and improving the engaging quality and service life, this dissertation studies systematically to build the model of the engaging, control optimally the engaging process, control precisely the current of the magnetic powder clutch and design the electronic control unit. At last, the developed control system of magnetic powder clutch is installed on the Ha Fei-MinYi prototype vehicle and related road tests are done.The working principle and the characteristic of the magnetic powder clutch are analyzed and the static model of the magnetic powder clutch is established by using the measured data. The method of building the speed characteristic model of the engine by using BP neural network is given. This model with different throttle openings is established by some data measured on DA465Q engine. On the basis of analyzing the dynamic mathematical model of the vehicle transmission system, the simulation model of the starting process is established. This is good for the research on the control algorithms.Control strategies of the clutch engaging process when starting are studied. A control strategy which combines the partial constant speed of the engine with the variable engaging speed of the clutch is proposed by analyzing the engaging quality of the starting clutch and factors which may influence the clutch engaging rule.For the difficulty of on-line setting of the fuzzy control quantization factor, a control algorithm based on the particle swarm optimized fuzzy control quantization factor is introduced in controlling the variable engaging speed of the clutch. Compared with the traditional fuzzy control algorithm, this algorithm has the following merits: the vershoot of the engine speed is greatly deduced on the process of starting; the maximum collision of the vehicle and the slipping work of the clutch are improved.Aimed at the non-linear phenomenon in the clutch drive circuit and the difficulty of setting control parameters to the optimum value by the traditional PID when the vehicle is operated in the complex condition, the algorithm of controlling the clutch current by PID setted with Radial Basis Function neural network is put forward in terms of the rapid convergence rate and strong overall approaching capability of the neural network. Using this method, the anti-disturbance ability is stronger and the adaptive capability of the system is better than those of controlling by PID. It sets a foundation for tracking accurately to the clutch current.The control system of the magnetic powder clutch with Ha Fei-MinYi is designed by using Infineon XC164CM as the core. XC164CM is a 16-bit MCU. For the bad working environment during the operation of the vehicle, the reliability and anti-disturbance ability of the electronicl control unit are taken into consideration.The test bench of the vehicle transimmion system is built. A road test assistant testing system is designed. Starting experiments are made on Ha Fei-MinYi prototype vehicle under many kinds of operation conditions. The experimental results show that control system can adapt to different working conditions well. When starting, the speed fluctuation of the engine is small, the vehicle is able to start smoothly, and the purpose of the driver can be reflected well. At present, the prototype vehicle has run over 20 thousand kilometers, the system runs well, and the vehicle comfort and safety can have achieved better effects.