Font Size: a A A

Optimization Of Energy Management Strategy For Parallel Hybrid Electric Vehicle

Posted on:2009-04-02Degree:DoctorType:Dissertation
Country:ChinaCandidate:J WuFull Text:PDF
GTID:1102360245496155Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
As the energy security and environmental issues are getting more and more remarkable in the world,developing clean vehicles with high fuel economy and low exhaust emissions has become a primary task in the automotive industry.Under the background mentioned above,hybrid electric vehicle(HEV),which incorporates the advantages of both the electric vehicle and the conventional ICE(internal combustion engine)vehicle,is regarded as the most practical solution for clean vehicles of high fuel efficiency,low emissions at present.As a new type vehicle with different energy,HEV's performance largely depends upon its energy management strategy(EMS).With the prerequisite of satisfying the driving requirements,EMS should control the torque distribution between engine and electric motor to obtain high fuel economy,low emission and good driving performanceaccording to the characteristic of hybrid powertrain and real-time driving cycle,.The energy management problem of HEV is a complicated problem involving non-linear dynamic optimum design.As a main bottleneck that impacts the vehicle performance and the process of HEV industrialization,the energy management problem of HEV has not been solved and needs to breakthrough.Because hybrid powertrain is a nonlinear dynamic system integrated with electrical,mechanical,chemical and thermodynamic devices,the synergetic operation of the hybrid powertrain itself and different components is complex,so it is difficult to construct an accurate mathematical model of the hybrid powertrain.Meanwhile,the unpredictability of driving conditions and driver's habbits also increase the difficulty of the design of EMS.Thus,aimed at the problems of existing parallel HEV's EMS,this dissertation explores how to optimize and improve the EMS using intelligent theories,such as fuzzy control theory,neural network theory,cluster algorithm and multi-objective optimization.And the main framework is as follows:In the first chapter,the background of task,present situation and key technology of HEV are introduced,so is the present situation of HEV's EMS.After that,considered the defects of existing EMS,it is pointed out that the optimization and improvement of EMS is the key to improve the HEV performance.The operation modes of parallel HEV and control thought of EMS are analyzed in this paper.Based on the engine efficiency map and battery internal resistance curve,an EMS is proposed for a parallel HEV with continuously variable transmission(CVT)using dynamic logic threshold approach.The dynamical transitions between the various operation modes are realized by setting a set of alterable logic thresholds and the torque distribution between engine and motor,and CVT ratios are determined through optimal operation lines of hybrid powertrain system in different operations to improve the vehicle performance.Based on the analysis of traditional instantaneous optimization EMS,four factors that influence the energy distribution,are confirmed.The instantaneous optimization EMS is summarized as a set of energy management rules.The energy management rules that gotten by offline simulation using instantaneous optimization EMS,are classified by fuzzy C-mean cluster and selected as the training sample of neural network.The BP neural network controller is used to control the energy distribution of hybrid powertain, then a real-time EMS is proposed for PHEV based on BP neural network.The simulation results demonstrate that,compared with instantaneous optimization EMS,the proposed EMS not only satisfies the fuel economy,but also increases real-time performance of energy management effectively.Aimed at the problems of fuzzy EMS that is widely adopted in parallel HEV,this paper proposes a novel fuzzy EMS based on particle swarm optimization(PSO).Firstly,a fuzzy controller of energy management,which uses the torque request to the hybrid system and the battery SOC as the inputs,and the control coefficient of engine torque as the output,is constructed.The scheme of fuzzification interface,database,rule library and defuzzification interface has been made.Then,membership functions and rules of fuzzy controller,which mainly depend on expert's experience,are optimized simultaneously by using PSO based on the optimization object of fuel economy.Finally,the simulation results show that,compared with unoptimized fuzzy EMS,the fuzzy EMS based on PSO reduces fuel consumption and maintains the battery SOC within its operation range more effectively.The driving cycles of HEV are various and the EMS should adjust its control parameters automatically to adapt to the different driving cycles.So,a fuzzy EMS based on driving cycle recognition is proposed for parallel HEV in this paper.The EMS is composed of driving cycle recognition and fuzzy torque distribution controller.The present driving cycle is recognized by learning vector quantization combined with vehicle travelling parameters in driving cycle recognition.Based on the identification results of driving cycle recognition,fuzzy torque distribution controller selects corresponding membership function and rule to control hybrid system.The simulation research demonstrates that the fuzzy EMS based on driving cycle recognition improves EMS performance more effectively.The match of powertrain components coupling with EMS effect is another key that has great relationship with vehicle performance.So,the optimization of EMS parameters and the match of powertrain components together are considered as a multi-objective optimization problem,whose optimization variables are the principal component parameters of powertrain and the parameters of logic threshold EMS,with a view to reducing the fuel consumption,exhaust emissions as well as the manufacture cost of HEV. Then the multi-objectives optimization problem is converted into a single-objective optimization problem by goal-attainment method.The multi-objective problem is optimized using PSO,whose constraint condition is dynamical performance of HEV. Computer simulations are hence carried out,which show the this scheme gives much preferable results to that optimized by the ADVISOR method,and thereby the fuel consumption and exhaust emissions of HEV can be effectively reduced without sacrificing HEV's dynamic performance.As one of the major key techniques,EMS needs to develop and consummate continuously.This paper optimized and improved the EMS of parallel HEV using relative knowledge and validated the porposed strategies with the simulation.The significance of the work in this dissertation lies in improving the domestic research and development of the HEV and promoting the industrialization of the HEV.
Keywords/Search Tags:Hybrid electric vehicle, Energy management strategy, Fuzzy control, Neural network, Particle swarm optimization, Multi-objective optimisation
PDF Full Text Request
Related items