Surrogate Model Of Temperature Prediction For Wet Dual Clutch Based On GA-BP Algorithm And Simulink

When a Wet Dual Clutch is operating,a large amount of heat is generated due to the relative sliding and friction between the dual steel sheet and the friction plate.One part of the generated heat is taken away by the cooling oil,and the other is conducted into the clutch body so that the temperature of the clutch components is increased.If the clutch cooling system has no enough dissipating capacity and cannot taken away the generated heat in time,the temperature of the friction pair will increase,resulting in the degradation even failure of the clutch working performance.It is necessary to control the cooling system parameters by predicting the working temperature of the dual steel plate and the friction plate of the clutch,so that the temperature of the friction pairs is ensured to be below required value,in this way,the clutch can operate continuously and reliably.In this paper,the heat transfer of a Wet Dual Clutch is modeled to solve the problem of predicting the temperature of friction pairs and cooling oil based on thermal network theory,and the temperature sample library for the main working components and cooling oil of the clutch under different working conditions is obtained with the established themal model.Then,a proxy model for quickly predicting the temperature of the main components and cooling oil of the clutch is established on the basis of the genetic neural network algorithm(GA-BP).The work and main conclusions are as follows:(1)A heat transfer path diagram of the clutch is drawn,meanwhile,the main components of the clutch and the flow area of the cooling oil are discretized based on the structural characteristics and heat transfer mode of the clutch.Then,the simulation model of clutch heat transfer analysis is established by AMESim software on the basis of the thermal network theory.(2)A three-dimensional simulation model of the clutch cooling oil filed is established by Fluent software,which is to analysize the influence of the clutch inlet flow,working oil temperature and the input speed on the cooling oil field,and then to obtain the characteristic velocity of the cooling oil for calculating the convective heat transfer coefficient in the heat transfer model.The results show that the input speed has a great influence on the cooling oil field in the clutch,while the inlet flow and the working oil temperature have little influence on the flow field which can be ignored in calculating the convective heat transfer coefficient.Finally,the cooling oil flow field of clutch at different speeds is simulated,and the characteristic speed of cooling oil on the surface of each component is extracted.(3)According to the convective heat transfer mode between the components of a Wet Dual Clutch and the cooling oil,a calculation method of convective heat transfer coefficient between the main components of the clutch and the cooling oil is determined,and the convective heat transfer coefficient of each mass block of clutch at different speeds and oil temperatures are calculated according to the characteristic velocity of each components surface.According to the structural characteristics of clutch friction pair,the three-dimensional models of heat flow distribution coefficient of odd and even gear friction pairs are established.Through simulation calculation,the heat flow distribution coefficient of odd and even gear friction pairs are obtained.Compared with the experimental results,the heat transfer simulation model is verified.The results show that the equilibrium temperature simulation results of dual steel sheet and oil chamber are close to the experimental result,and the relative error is small.(4)The sample library of temperature prediction proxy model of clutch main components and cooling oil is obtained by using the established heat transfer model of Wet Dual Clutch.Then,the proxy model for predicting the temperature of clutch main components and cooling oil under different working conditions is established on the basis of GA-BP algorithm and Simulink.The results show that the temperature of clutch dual steel sheet and oil chamber predicted by the established proxy model is close to the experimental results,and the error is small.