Research On Control Strategies Of Plug-In HEV Based On Multi-Objective Optimization

Vehicle industry is facing great challenge from energy shortage and also contributes to air pollution and comes with large quantities of global warming greenhouse gases. To reduce the reliance on the non-regeneration energy and improve running efficiency, Plug-In Hybrid Electric Vehicles (PHEV) are widely considered to be the most promising long-distance transportation vehicles instead of traditional engine vehicles.The energy management strategy has important theoretical significance on the performance of hybrid electric vehicle. However, the information of driving cycle is needed in advance in the global optimization, which is rarely practical. The energy management strategy based on the Particle Swarm Optimization (PSO) algorithm is designed to take both the vehicle’s fuel economy and minimum emission as objective function. Compare with deterministic Rule-Based energy management strategy, PSO algorithm can improve fuel economy and emission.CVT is an ideal transmission system for HEV, which has predominance at property and structure. The impulsion affects the endurance life of detail of vehicle and passenger ride comfort characteristic. The purpose of this paper is to solve above problems by is researching on the relevant problems concerning CVT vehicle clutch control and the improvements of drivability and NVH. Based on the design of the particle swarm algorithm and the proportional controller, key influence factors of the joint flexibility of the automatic clutch control are summarized. The engine speed and the load signal are used as the inputs, fuzzy control table is set up, relationships among the systematic parameters are derived, the binding extent of the clutch is determined, torque coordination control is realized and the ride performance is raised. The results shows degree of jerk can be reduced effectively and can be satisfactory to the requirement of German related regulations. Simulation results demonstrate the effectiveness of the proposed methods.The vehicle passive safety technology can only solve the problems caused by traffic accidents. The active safety technology, which can prevent and reduce accidents, would suffice for more far-reaching applications. In this paper Elman neural network is adopted to predict driver’s behavior ahead of time. The "people oriented" driver-vehicle-road closed loop model is set up. The system would record the habits of the driver and warn in time when the behaviors of the driver deviate from the forecasted trajectory to a certain extent. Real time simulation is carried out, which is based on 3D urban road that acquired by GPS equipment. The results indicate that Elman algorithm can be used to establish the warning system of driver’s improper operation and provide the reliable and valuable information for safe driving.