Research On Intelligent Self-learning Method For AMT Vehicle Starting Control And Its Applications

With the rapid development of economy in our country, China has become the world’s largest automaker and automobile market. However, most of the vehicle and key parts of the market are of foreign enterprises or joint ventures. The independent brand car companies develop slowly. As the "2025" made in China, intelligent manufacturing, energy conservation and emissions reduction, innovation-driven development strategy is put forward, would greatly prompted the transformation and upgrading of industries in China. all walks of life began to pay attention to technology innovation. It will also bring car technology innovation opportunities.Because of the AMT has simple structure, high transmission efficiency, better driving experience, the companies to develop vigorously. This article is based on enterprise cooperation study the AMT start control.First of all, clarified the research background and significance of this topic. Through comprehensive literature analysis found that the starting clutch control problems and deficiencies. Clutch control much focus on the joint speed precision and neglected the joint displacement and the coordinated control of engine. Pay attention to the intelligence and efficiency of the algorithm,but ignore the difficulty of the automobile application. On the basis of analysis, to determine the content in this paper, the research.The second study the structure of the starting clutch control system. The second study the characteristics of the clutch. Through the analysis of the composition of the transmission system and the characteristics of components, bulid engine dynamics model, the clutch torque model, vehicle model, vehicle start joint control model and clutch of mathematical model of hydraulic system.Emphatically analyze the key components of the diaphragm spring clutch deformation and stress characteristics, and lay a foundation for the research of control strategy.Study on the control strategy structure features and related algorithm for clutch starting up. Controller based on the practical application, the layered hierarchical control algorithm is used. the structure of Vehicles start controller is divided into perception layer, policy makers and executive management. According to high perception layer and the requirements of the intelligent decision makers, the two layer fuzzy control algorithm. According to the executive level requires strong adaptability, it adopts the intelligent fuzzy control. And study the specific algorithm.Study algorithm implementation and modeling simulation. According to the function of each layer control, the specific control algorithm is desig ned. According to driver’s initial intention, engine speed and the clutch disc speed relative factors, such as difference of policy makers to determine the speed of clutch joint and the joint displacement. Executive control actuators complete laws set by the decision makers. Through the Simulink modeling and simulation verify the performance of the controller.Finally, based on d SPCAE, complete rapid prototyping real vehicle experiment. Through the DSPACE platform, controller model in simulink can be converted to the C code, and it is used to control the clutch joint directly. In the process of testing it can online debugging relevant parameters, in order to obtain good control effect. It to verify whether the performance of the controller can meet the drive r start the initial intention of small, small impact degree, friction work, engine running smooth start requirements.