Research On Temperature Distribution Modeling,State Of Charge Estimation And Thermal Regulation Strategy Of Power Lithium-ion Battery

Lithium-ion batteries(LIBs)have been widely used in the field of electric vehicles(EVs)due to their advantages such as high energy density,long cycle life,and low self discharge rate.However,LIBs are a complex system that couples electrochemical reactions with heat generation characteristics,and the temperature distribution is uneven,making it difficult to obtain an accurate thermal model;The high degree of parameter coupling makes online identification difficult,resulting in poor adaptability of the electrochemical mechanism model;The state of charge(SOC)of LIBs is easily affected by uncertainty and time-varying operating conditions,leading to many challenges in SOC estimation and temperature control.In order to improve the accuracy and reliability of battery management system(BMS)and ensure the safe and reliable operation of EVs,this thesis conducts in-depth research on modeling method for surface temperature distribution of LIBs,modeling of electrochemical mechanisms and online identification of time-varying parameters,robust estimation method for SOC,and temperature regulation strategy.The main research work and innovative points of the thesis are as follows:(1)Aiming at the problem of difficulty in obtaining an accurate temperature distribution model of LIBs due to the strong nonlinearity and complex boundaries,a spatiotemporal neural network temperature distribution modeling method for LIBs was proposed.Firstly,a time dynamic prediction model based on neural network was constructed to address the strong nonlinearity of temperature distribution in time dynamics.Secondly,in response to the nonlinear problem of spatial position distribution,a spatial distribution model based on Gaussian kernel function was proposed to characterize the nonlinear relationship between the spatial positions.Then,in response to the spatiotemporal coupling characteristics of temperature distribution,a spatiotemporal neural network temperature distribution modeling method for LIBs was proposed by combining the temporal dynamic prediction model with the spatial distribution function,and a two-step solution method was developed to identify the unknown model parameters.Finally,the convergence of the model was analyzed theoretically,and the effectiveness of the method was verified through various operating conditions experiments.(2)In response to the problem of poor accuracy of electrochemical mechanism models in variable temperature environments and low SOC stages,a reduced order electrochemical thermal coupling model for LIBs and an online identification method for time-varying parameters were proposed.Firstly,based on the spatiotemporal coupling characteristics of solid-phase diffusion process,a spectral method based reduced-order model of solid-phase diffusion process was established,and a lithium concentration distribution model was obtained.Secondly,a temperature distribution model for LIBs was constructed using spatiotemporal neural network.Based on this,a liquid phase potential model based on the double layer effect was established,which was integrated with ohmic potential and polarization potential to establish a terminal voltage model.Then,in response to the difficulty of online identification of electrochemical mechanism model parameters,a time-varying electrochemical mechanism parameter model based on temperature and SOC was established,and an adaptive multi-innovation UKF parameter online identification method was proposed to improve the adaptability of the mechanism model.Finally,the effectiveness and correctness of the model were verified through various experiments.(3)Aiming at the problem that temperature,uncertain interference and time-varying operating conditions will affect the accuracy of SOC estimation,a robust fuzzy model and multi-innovation UKF algorithm were proposed to estimate the SOC of LIBs.Firstly,to address the issue of traditional fuzzy models being unable to handle uncertain disturbances such as non-Gaussian noise,a robust fuzzy terminal voltage modeling method that minimizing the mean and variance of model errors was proposed to ensure the accuracy of terminal voltage observation.Secondly,in order to further improve the accuracy of state estimation,a multiinnovation UKF algorithm was proposed and combined with a robust fuzzy model to update the estimated SOC of LIBs by using residual online feedback.Finally,the effectiveness of the proposed method was verified through multiple operating conditions experiments.(4)In response to the complex heat generation process and uneven surface temperature distribution of LIBs,as well as the slow cooling speed and limited cooling capacity of air-cooled/liquid-cooled thermal management systems,a temperature control system and thermal regulation strategy for LIBs using semiconductor refrigerators were designed.Firstly,a temperature control system of lithium battery based on semiconductor refrigerators was designed.Secondly,a thermal regulation mechanism model including spatiotemporal neural network temperature distribution,electrochemical heat generation,and SOC estimation was established.Then,aiming at the unsteady heat transfer characteristics of the cooler,a modeling method of the dynamic temperature distribution of the cooler based on the spectrum method was proposed to improve the modeling accuracy of the cooler.Finally,aiming at the nonlinearity and parameter uncertainty of the thermal management system,an adaptive fuzzy sliding mode thermal control strategy was proposed to realize the adaptive robust control of the system under disturbed conditions.The effectiveness and correctness of the proposed strategy were verified by experiments under various working conditions.