Neural Network Based Torque Observer for Vehicle Transmission Shifting Processing Modelling
To reduce shock during transmission gear shift, a Neural Network based torque observer for vehicle transmission shifting processing modelling is proposed in this paper. In contrast to traditional method for design of the closed-loop control algorithms to improve shift quality, the modelling of transmission input torque , which is needed by the observer for accurate clutch pressure estimation, is addressed and implemented using a Neural Network based observer. The resulting observers are validated via off-line simulation tests, as well as experiment tests at different sampling frequencies on a test vehicle bench, for demonstration the observer performance and establishment the feasibility of the approach. It concludes that the Neural Network based torque observer can obviously help to reduce the transmission output shaft torque during gear shift.
Neural NetworkTorque ObserverVehicle TransmissionShiftingModeling
Pingkang Li Xiuxia Du Yunpeng Liu
Beijing Jiaotong University, Beijing, 100044, P. R. China Northern Vehicle Research Institute, Beijing, 100072, P. R. China
国际会议
北京
英文
2007-08-05（万方平台首次上网日期，不代表论文的发表时间）