Model Predictive Control For A Plug-in Hybrid Electric Vehicle

Compared with the conventional hybrid vehicles, the plug-in hybrid electric vehicles (PHEVs) are equipped with larger capacity of energy storage device, and can charge the energy storage devices by using the ordinary grid, which increases All-electric range(AER), greatly reduce the fuel consumption and exhaust emissions. PHEV has become one of the most important technical means in energy conservation in developed countries'development plan of a new generation automobile.As the batteries of PHEV with much higher capacities, PHEV should make full use of the battery energy when arriving at the destination,and the control strategy is different from the traditional ones. PHEV with global economic optimal control strategy based on different driving cycles shows that"Blended Mode"is better than the traditional"Depleting- Sustaining"mode.However,the global optimal strategy is unable to use in Real-time control.With the application and popularization of GPS\GIS in vehicles,predictive control for PHEV can reduce more fuel consumption and emission by future driving status messages supported by on-board navigation system.The mathematics model of PHEV's power train were established and the main parameters of components were calculated.The dynamic programming was applied in model predictive control configuration, and a model predictive control mathematics model was established for the fuel economy of a PHEV based on spatial-domain, then the global optimal dynamic programming was transformed to local optimization in prediction horizon.A global optimal method for PHEVs was studied with dynamic programming, and the results indicate that the state of charge (SOC) is reduced slowly from the maximum at initiation to the low threshold on the destination of the trip, therefore, theoretical slope of SOC was suggestion, which was regarded as reference SOC slope of model predictive control, practical SOC at current location was considered as initial value, and the reference SOC of prediction horizon end-point was regarded as target value,The target value was modified according to the especial driving modes in the future. The results show that fuel consumption is reduced obviously by using this method.In order to improve the system operation efficiency, the fuel economy predictive control program of PHEV was written with higher C language instead of MATLAB language. A co-simulation was made between the C language and the MATLAB\Simulink with the help of MATLAB external interface files.The problem dealing with engine frequently open and close was presented,New Europe Driving Cycles as calculation conditions was Adopted,model predictive control strategy of PHEV was simulated and analysed under the different mileages, The results show that model predictive control based on the method that theory SOC slope was modified are veryfied effective. Finally, reducing the "dimension disaster" of dynamic programming and choosing reasonable prediction horizon were studied.