Research On Control Strategies For Engaging Process Of Automatic Clutch For Vehicle

Automatic transmission is the development tendency of transmission for vehicle. Automatic mechanical transmission (AMT) has many merits such as low cost, simple structure, high transmission efficiency and good fuel mileage; moreover it needs few changes with the structure of present manual transmission, so AMT is a promising transmission which is suitable for the situation of our country. Automatic clutch is the most important module of AMT and the core and difficulty of ATM control. At the same time, automatic clutch can be developed as an independent production, which can decrease driving intension and improve comfort and safety of driving compared with manual transmission. The control difficulty of automatic clutch focuses on two aspects: how to figure out the optimum engagement rule and how to track the target engagement position precisely. The present researches on them are not mature enough. Aiming the two problems mentioned before, adaptive control strategies were deeply studied. Finally, aiming at a sample car automatic clutch control system was developed and long terms experiments were carried out.1) The merits and shortcomings of clutch driving mechanism existing were analyzed and a new driving mechanism based on DC motor was designed. In the scheme a whirl spring was adopted to compensate load toque, by which the load of motor can be reduced and system's dynamic performance will be improved. Additionally, a mechanical limited mechanism was devised. The design method of whirl spring mechanism was represented in detail. Furthermore, the integrated dynamic model of mechanism including clutch, worm reduction box, compensating spring and motor was established, which is the theoretical foundation of position tracking control.2) Utilizing eddy current dynamometer system, the steady load characteristic of DA465Q engine was measured, and the mathematical model of engine torque was established by way of data fitting method. Moreover the dynamic simulation model of transmission system of vehicle was established. The factors influencing clutch engagement were analyzed. According to the control aim of starting, an improved strategy of constant engine speed in part process is proposed as a total control principle. For the realization of the strategy, a compound control strategy combining fuzzy control and rule coordinating control is proposed. Throttle opening, engine speed relative deviation and engine speed rate were chosen as control parameters to construct fuzzy algorithm, while the minimum engine speed for engaging, the difference of input and output axis speed of clutch and out axis speed of clutch were chosen auxiliary control parameters to design rule controller. The starting simulation results of diverse conditions indicate that the strategy had good control effect and good adaptability, and it provided the theoretical gist for the control software design.3) A self-learning compensating strategy was put forward to solve the problem of clutch abrasion. The position of unloaded displacement termination point was measured dynamically on line by using the character that the torsion transmission characteristic of clutch kept relatively constant. The method is feasible and practical.4) A compound filtering method based on self-adaptive median decision filter and IIR filter was proposed to solve the problem of rotation speed signal filtering. Self-adaptive median decision filter was used to filter the big impulse disturbing signal and low order butterworth filter was used to smooth the signal. The experiment results show that the compound filtering method can accurately filter the singular value with the smallest filtering window under big density disturbance and has good smoothing ability compared with common medium filter and pure FIR filter. Real-time and accurate filtering can be implemented.5) Accurate and quick position tracking of automatic clutch is difficult on account of its nonlinear characteristic. In this subject a neuron adaptive PID controller with feedforward and adjustable gain was designed. The feedforward control was able to accelerate the response of controller, the weight of neuron was adaptively adjusted by self-study function, and PSD algorithm was proposed to adjust its gain. The clutch position tracking experiments showed that the controller could adapt the nonlinear change of clutch load and target engaging speed with small position error. This indicated that the controller had good robustness and can meet position tracking control command.6) The electronic control unit (ECU) of automatic clutch was developed for the sample car based on infineon XC164CS microcontroller. According to vehicle's running conditions and control strategy, the control software was developed. Moreover, an auxiliary debugging software was developed based on VB for convenience to debug. The control system starting and shift experiments under diverse condition were carried out on the sample car. The experiments showed that the vehicle can start successfully on no matter flat road or slope road. While starting on flat road, the vehicle started smoothly according to driver's intention with low engine speed fluctuating; while starting on slope it can prevent vehicle from countermarching and engine halting. This indicates that the automatic clutch control system has good performance. Until now the sample car has traveled over 15,000 km, and the comfort and safety is satisfying, which indicates that the control strategy is effective.