Pneumatic Servo Control Of Gear Shifting For An Automated Manual Transmission(AMT)

As a kind of engineering technology, pneumatic technology can transfer energy and signal with the working medium of compressed air. It has the advantages of low-cost, high power-to-weight ratio, clean operation, simple design, easy maintenance and use. In addition, it is also flame-proof, anti-explosion and anti-electromagnetic interference. These properties can provide strong support for the development of electro-pneumatic automated manual transmission (AMT) systems. However, pneumatic systems have many weaknesses that can make accurate control difficult, such as the compressibility, low stiffness and damping of air, the flow characteristic of valve port and the friction characteristics of cylinder. So, the development of pneumatic-driven AMT systems is trapped. Due to recent improvements in microelectronic technology, many high cost-effective pneumatic control valves, executed components and sensors are constantly emerging. And, with the development of modern control schemes, more and more promising control schemes have gradually been presented. These make a great progress in pneumatic servo control, and also broaden the application range of pneumatic technology. Hence, it is possible to develop a high-accuracy pneumatic-driven AMT system successfully.The main content of each chapter is summarized as following:In chapter 1, the literatures that related to AMT and pneumatic servo control technology are reviewed, and the key point of the research is pointed out. Relative research significance, difficulties and contents of this doctoral dissertation are illustrated.In chapter 2, the kinetics models of AMT systems at each stage are analyzed. According to the influence factors of shifting quality, we determine the key stage of pneumatic servo control in AMT systems. To prepare for the design of the controller later, the nonlinear models of pneumatic system in AMT systems are established, such as flow model of valve port, thermodynamic model of cylinder chamber and friction model of pneumatic cylinder. The model parameters and the relation between the PWM signal’s duty cycle and the control law are identified experimentally. A new method to obtain the clutch load characteristics is presented to lay the foundation of achieving high-accuracy pneumatic servo control of AMT systems.In chapter 3, in order to avoid stick-slip phenomenon encountered in pneumatic cylinder, stick-slip influence factors in practical application, such as supply pressure, sonic conductance of orifice, external force and initial volume of inlet chamber, are studied experimentally. Based on the stick-slip criterion concluded from existing studies and analysis of the stick-slip phenomenon, the critical stick-slip experiments of different working conditions are conducted. Through the observation of the experimental curves, it’s found that the final piston velocity is constant. An empirical discriminant function with supply pressure, sonic conductance of orifice and external force as parameters is established by processing the critical stick-slip experimental data for polynomial surface fitting, and its validity and feasibility are verified by performing experiments with another pneumatic cylinder of the same type. The results show that the empirical discriminant function can be used for instructing practical engineering applications by efficiently predicting whether the stick-slip will occur or not for the same type pneumatic cylinder with different supply pressures, sonic conductance of orifices and external forces.In chapter 4, sliding mode control algorithm which is strongly robust to some nonlinear uncertainties and bounded disturbances is adopted to study the motion trajectory tracking control for on/off valve controlled pneumatic cylinder system subjected to variable-stiffness and hysteretic load force. At some stage, there even exists negative stiffness. Based on sliding mode control algorithm, non-model based sliding mode control strategy is proposed, and a sliding mode control law is implemented directly without using pulse-width modulation to track the reference trajectory. Meanwhile, the load characteristics of clutch could be estimated indirectly by pneumatic driving force during trajectory tracking. In theory, the estimation error of load characteristics is small. So, the clutch load characteristic is closer to the actual value in the trajectory-tracking control. Since the disturbance rejection of the actuator with low stiffness is weak from force to motion, it will be worse for the actuator with negative stiffness. For improving motion trajectory tracking accuracy, simultaneous motion and stiffness maximizing control is presented. In addition, we study adaptive robust control algorithm in depth, and present a modified direct adaptive robust control algorithm. Experiments show that the above presented control algorithms are all effective.In chapter 5, to achieve rapid disengagement and smooth engagement of clutch in AMT systems, servo control of an electro-pneumatic clutch actuator controlled by solenoid on/off valves is considered. Nevertheless, the realization of a high-accuracy servo control usually depends on model-based nonlinear control technologies requiring the full-state knowledge of the system. Pressure sensors are not used in AMT systems due to cost considerations. Thus, the pressure states in the chambers are unknown. To address this issue, a nonlinear pressure observer taking the place of pressure sensors was proposed in this paper. It is stable in the sense of Lyapunov and independent of load. Furthermore, the pressure observer’s polytropic exponent was analyzed and then determined experimentally. Finally, a compound sliding mode controller with the pressure observer is presented. Extensive experiments demonstrate the feasibility and effectiveness of the proposed pressure observers-based controller.In chapter 6, to achieve pneumatic servo control of the shifting process for AMT systems without pressure sensors, non-model based sliding mode control algorithm is adopted to control the gear’s switching, and pressure observer based compound sliding mode control strategy is developed to control the disengagement and engagement of clutch. During the disengaging process of clutch, the same non-model based sliding mode control algorithm is employed. While during the process of clutch engagement, simultaneous motion and stiffness maximizing control is implemented to ensure smooth engagement. Based on the theory of hierarchical control, a pneumatic servo controller of gear shifting for AMT systems is synthesized to achieve the task of gear shifting under the cooperation of clutch and gear. The core idea is to complete the respective control activities independently on the basis of the objective gear and the statuses of clutch and gear. Experiments show that the presented control strategy of gear shifting for AMT systems is feasible and effective.In chapter 7, the achievements of the thesis are summarized and the further research work is put forward.