An Intelligent Control Method For The Shifting Of The Dry-type DSG

Abstract:Currently, the car as a convenient means of transport, in people’s daily life occupies an important position, vehicle ride quality requirements continue to increase. Shift as an important part of vehicle control, to some extent determines the vehicle’s comfort, economy. The dual-clutch transmission with its superiority, a growing concern. Therefore, the intelligent control algorithm applied to the dual-clutch transmission control system has great significance on the development of automobile industry.In this paper, DCT mechanical structure works as an entry point, combined with the actual road conditions, driver’s operating habits, car quality indicators reality factors, analyzed the problems of dual-clutch transmission shift process, and solve these problems. species shift intelligent control system structure and algorithm ideas. The algorithm mainly shifts strategies and shifting process control in two parts.First of all, the premise of a comprehensive vehicle economy and power control objectives for the affected by many factors, dual-clutch transmission stall and control objectives presented in the complexity of the law, the use of neural network approach to traffic data, drivers intent, learning, establish a model of the relationship between the stall with integrated indicators, and use the genetic algorithm to optimize the neural network connection weights heavy. Finally AdaBoost ensemble learning methods, to further improve the accuracy of the genetic neural network model.Then, according to the goal of shifting process quality indicators, the characteristics of the actual work of the DCT two clutches, shift fast response and the need for multi-objective optimal. A shifting process control based on simulated annealing algorithm, the switching speed of the combination and separation of the two clutch double clutch transmission optimization and the use of the the Logistic mapping algorithm, improved simulated annealing neighborhood perturbation method to enhance the algorithm search capabilities.Finally, the analysis of a concrete realization of the DCT gear control system hardware architecture, software architecture, and the core software control algorithm. Structures based on the simulation platform, and contrast simulation results demonstrate the effectiveness of the proposed method.