Non-linear control of a gear shift process in a dual-clutch transmission based on a neural engine model. (October 2021)

• Record Type: Journal Article
• Authors: Bera, Piotr; Sikora, Wojciech; Wędrychowicz, Dariusz
• Is Part Of: Control engineering practice. Volume 115(2021)
• Publication Date: 2021-10
• Subjects: Dual-clutch transmission -- Combustion engine dynamic model -- Neural network engine model -- Inertia phase; Automatic control -- Periodicals; 629.89
• DOI: 10.1016/j.conengprac.2021.104886
• Languages: English
• ISSNs: 0967-0661

Abstract

Abstract: The paper presents an engine control algorithm, based on an artificial neural network (ANN), which ensures a quick and smooth inertia phase of an upshift in a dual-clutch transmission (DCT). The main emphasis is placed on 3D characteristics, which can be directly implemented in the ECU, to control the inertia phase, when the engine speed decreases to reach the target clutch speed. They are developed based on an ANN model which approximates data obtained during engine test bed measurements in dynamic states. Moreover, to provide an overall gear change algorithm, the dual-clutch assembly activation mechanism was also analysed. Graphical abstract: Highlights: Inertia of the engine influences its performance in dynamic states significantly. Dual clutch transmission control with the use of static characteristic is imprecise. Neural network allows the engine dynamic model to be obtained. Neural engine model allows 3D characteristics to be determined. Developed 3D characteristics control throttle angle during upshifts.