Study On Energy Management Strategy For Plug-In Hybrid Electric Vehicle

As the global oil crisis and environmental pollution become increasingly serious, improving energy use efficiency and reducing environmental pollution have become the primary task of the automotive industry development. Plug-in hybrid electric vehicle (PHEV) can not only significantly reduce human dependence on oil resources, but also be effective in reducing urban air pollution. So it has become one of most important technical means of vehicle energy saving and emission reduction. As the core technology of PHEV design, the quality of energy management strategy (EMS) directly impacts on vehicle power, economic and emission performance. In this dissertation, for obtaining a reasonable EMS which can be applied on the vehicle, and a better way to design and optimize PHEV EMS, we conduct a in-depth study of PHEV energy management strategies, by following the design ideas of "system modeling - matching design - offline optimization - online application - test to verify" and using the methods such as simulation modeling and mathematical optimization.In order to provide necessary simulation environment for PHEV powertrain matching and real-time EMS verification, a PHEV simulation system was built. A forward-looking simulation structure was selected and the whole simulation system was divided into four parts: driver model, vehicle controller model, powertrain model and vehicle model. To improve the modeling efficiency, standardize the data transmission ports between the powertrain components. By using theoretical modeling and experimental modeling method, we built engine model, motor model, power battery model, gearbox model, clutch model and vehicle model in Matlab/Simulink language. For verifying the system simulation model, we applied it to simulate a sample hybrid electric vehicle with parallel powertrain configuration. The results shown that: the simulation model can meet the precision and real-time requirements.To obtain the optimal powetrain matching design, a PHEV powertrain matching method was developed. According to the basic parameters of the prototype vehicle and American technical performance indicators, PHEV performance constraints were put forward. Then the PHEV powertrain layout structure was proposed in accordance with the characteristics of PHEV. Through the established system simulation model, a series of matching designs of powertrain parameters were obtained. After doing cost analysis on powertrain components, we found the optimal design by the self-developed optimizing program, which was not only able to meet the performance constraints but also consider the actual situations. For verifying the matching design, a simulation was started. The results demonstrated that: the vehicle with optimized matching design fully met all the performance constraints. Base on the optimal design, a sample plug-in hybrid electric vehicle was developed, which could provide a good vehicle platform for the PHEV energy management strategy application.For obtaining optimal controls, which can be used to establish online EMS at the macro sense, under different mileage, an offline global optimization method was brought forward. Bellman principle was used to convert the optimal control problem of vehicle performance into step-by-step decision-making problem. By the rational discretization of reachable state set and admissible control set, we managed to get numerical solution of the global optimization algorithm. Through Matlab and C++ mixed programming code calculation, the optimal controls under different PHEV mileage were obtained. The optimization results showed that:①. When the mileage is less than 55km (NEDC×5), PHEV should be run under the motor-dominant driving mode; When the mileage is greater than 121km (NEDC×11), the vehicle should be run under the engine-dominant driving mode; When the mileage is greater than 55km and less than 121km, the vehicles should be run under the power-balance driving mode of dynamic.②. When the vehicle mileage is 55km, the energy obtained from the power grid could be used most efficiently. So, we got the best vehicle economic performance at this mileage. When the vehicle mileage is less than 165 kilometers, the vehicle average equivalent fuel consumption is 2.9 L/100km, which is improved by 63% compared with the prototype vehicle.A method of designing PHEV online energy management strategies was developed based on the offline global optimization calculation results. That is, through the method of statistical analysis and multivariate nonlinear regression, we summed up the macro-distribution rules of PHEV powertrain energy flow under different optimal control and developed powertrain operating mode switching rules and power distribution rules under power-balance and engine-dominant driving mode. From the correlation analysis of regression formula, we knew that: the formula regressed by the Levenberg-Marquardt algorithm, whose calculated values was highly relevant to the optimized values, could be used to make an engine operating map for optimizing the distribution of powertrain energy flow. Finally, for the removal of human subjective factors during powertrain operating mode switching rule-making process, genetic algorithm (GA) has been chosen to optimize the control parameters of powertrain operating switching rules. With the optimized control parameters, PHEV was very close to its best economic performance. It fully demonstrated that: the established online control simulation system was with sufficient accuracy, and the optimized online EMS which was developed through the above design method was well close to the optimal control.For the application and verification of PHEV energy management strategies designed above, an engine power control system was developed and several road tests were carried out. The engine power control system was developed by using of Infineon's new generation automotive embedded control chips XC164. Through functional design, hardware development and test verification, the engine power control system developed in this dissertation could achieve all the functions well, which were engine direct start-stop control, engine output torque control and CAN communication. In this way, it was certainly able to meet the control requirements from the hybrid powertrain. Then, road tests were conducted after EMS being applied to the PHEV. The results shown that: the online energy management strategy obtained by above methods, could meet the design requirements of PHEV. The strategy was able to make PHEV utilize energy from external power grid effectively and distribute powertrain energy flow reasonably according to different vehicle mileages. Thereby, the advantages of conventional fuel vehicles and electric vehicles were integrated, and a substantial increase in vehicle economic performance was obtained.