Research On The State Observation And Management System Of Power Battery

In recent years,new energy vehicles has been got the rapid development.The lithium iron phosphate battery with its excellent characteristics has become the new choice of power source for new energy vehicles.In order to improve mileage and safety performance of new energy vehicles under complicated condition,the state observation and reliable management system of lithium iron phosphate have become the key to promote the development of new energy vehicle,which is of great significance.In this dissertation,the lithium iron phosphate battery management system is taken as a research center,and as the two key technologies of this center,the state observation and management control have been carried on the thorough research.Firstly,the background and significance of the subject were introduced in this dissertation.The status quo at home and abroad of relevant research had also been reviewed.And,the main content of the research and arrangements were given.Secondly,the operating mechanism of lithium iron phosphate were analyzed in this dissertation.On the basic of this analysis,the factors that affect the accuracy of the battery model were pointed out.A general nonlinear equivalent electrical circuit model was deduced based on the electrochemical thought,which reveals the polarization and diffusion the nature of the performance.At the same time,the existing mathematical model was improved and an optimal mathematical model was proposed with the consideration of coulomb efficiency,temperature,deterioration,model optimization factors relaxation effect,hysteresis,etc,which lays the foundation for the battery SOC estimation.In order to reduce the computational complexity of traditional EKF algorithm and improve the stability of the algorithm,an adaptive sigma kalman filter algorithm was put forward based on optimal Gaussian approximation kalman filter for the SOC observation.The square root of state estimation error covariance was introduced to improve the positive semi-definition of the state covariance.Sigma sampling sequence was also constructed,and the estimated state variable and the observed variable were updated based on the iterative minimum mean square error estimation to achieve a precise estimate of the battery's SOC.The lithium iron phosphate battery degradation and its remaining useful life prediction were also researched.The battery capacity attenuation experiment was designed to derive the battery capacity attenuation model for observation.Then,an adaptive sigma kalman particle filter algorithm was proposed based on the LiFePO4 battery capacity attenuation model for RUL prediction.The phenomenon of lack of particles was overcome adopting the proposed method,and the method was more simple and effective.In order to improve the accuracy of RUL prediction,a data driven auto regression model based on the capacity changes was built in this dissertation.Then,an ASKPF algorithm was proposed based on the data driven fusion model to achieve the goal of battery RUL prediction.The experiments verify the feasibility and effectiveness of the described method.Again,in order to achieve the consistency of each monomer battery equalization control for battery pack,equalization circuit and control method were researched through the energy point of view.A non dissipative secondary equalization circuit based on the energy storage inductances was proposed in this dissertation,which had the characters of simple structure and easy to cascade control.The mathematical model was built and the MOSFET sequence control was given at the same time.A control scheme of adaptive equalization current was proposed to control the equalization circuit.Experiments under no-load condition,charging conditions and efficiency test were given to verify the effectiveness and feasibility of the proposed equalization circuit and control scheme.In order to solve the problem of power battery pack charging,a bipolar DC-BUS charging system was proposed based on the V2 G integrated energy conversion system.For the former PWM main converter,a general programmed PWM modulation algorithm was proposed based on the Volt-second balance principle.A general switch sequence was designed in a single period of carrier.Meanwhile,the modulation algorithm is greatly simplified,and has more clear physical meaning.On the purpose of achieving the full range neutral voltage balance and low frequency oscillations suppression,this dissertation puts forward the nine-segment special solution planning programmed PWM algorithm and the neutral voltage closed-loop control strategy without sector judgement and vector composition analysis.At the same time,the topology and control method of integrated DC/DC system was research.And the multistage constant current and constant voltage trickle charging strategy was also designed.charging conditions and efficiency test were given to verify the effectiveness and feasibility of the proposed equalization circuit and control scheme.Balanced and unbalanced charging experiments verified the effectiveness and feasibility of the charging system and strategy.Finally,a multi-communication network battery pack management systems was designed based on the MSCAN bus and Modbus_RTU protocol in this dissertation,taking the Freescal carle automotive embedded processor as core.And the high precision battery information acquisition chip was used as the acquisition unit in the proposed battery pack management systems.The hardware and software design methods were also introduced to improve the reliability of the system and reduce the development cycle.