Prediction Of Driving Condition For Plug-in Hybrid Electric Vehicles

Predictive control strategy for hybrid vehicles uses the information of vehiclefuture driving conditions to optimize the engine and motor power allocation, in order toachieve low fuel consumption, low emission, adequate and reliable recovery ofregenerative braking energy and extend battery life in the actual driving operation. Theprediction of vehicle operating condition is difficult; the plug-in Hybrid Electric Vehicleis regarded as the research object to study the forecast of vehicle future conditions. Themain contents are as follows:①The mathematical model of the engine and motor in a plug-in parallel hybridelectric vehicle was established. The acceleration characteristics of the vehicle in pureelectric driving mode, the pure engine driving mode and mixed driving mode wereanalyzed to provide a reference for establishing a speed prediction model based onfuzzy theory.②The vehicle speed prediction problem in future driving condition was studied.For the problem of vehicle speed prediction in neighboring moment, based on theanalysis of vehicle speed data and driver operating behavior, the speed predictionmethod based on fuzzy theory was studied, and speed prediction method based onhistorical data and speed prediction method based on historical data-fuzzy theory wereproposed respectively. In the vehicle speed prediction mode based on fuzzy theory, thevehicle speed prediction mathematical mode was established, which took the vehiclecurrent speed, relative acceleration/braking pedal position and the changing rate ofacceleration/braking pedal as inputs, and the acceleration/deceleration of future momentas output.The prediction of transmission gear was researched. Gear prediction model basedon fuzzy control in driving condition was set up, whose inputs are vehicle current speedand demanding power, while output is the gear of future moment. And gear predictionmodel based on rules was set up in brake condition.③Simulation analysis was carried out using the fuzzy speed prediction model inthe typical driving cycles, the actual road conditions and the slope conditionsrespectively. Compare of simulation analysis in typical driving cycles and actual roadcycles were performed with speed prediction model based on historical data and speedprediction model based on the historical-fuzzy theory. From the simulation results, it is proved that speed prediction method with historical data-fuzzy theory is better than theother two methods especially in suburban driving cycle and when using more historicaldata. Both historical data-fuzzy theory based speed prediction method and historicaldata based speed prediction method can predict more precisely in the suburban cyclethan the urban cycle. Simulations were carried out based on gear fuzzy prediction modelin ECE driving cycle and EUDC driving cycle respectively, which were proved correctin transmission gear prediction.