Lithium-ion batteries have been widely used in national strategic emerging industries such as electric vehicles and energy storage projects, and have drawn wide attention from academia and industry. During the 13th Five-Year Plan period, China continued to list "new energy vehicles" as the key special topics of the state's key R & D plans and put forward higher requirements on the special-purpose power batteries for new energy vehicles. The design and manufacture of low-cost, high-capacity, long-life and high-safety lithium-ion batteries are the basic goals proposed by new-energy vehicles for the next generation of power batteries. The battery management system (BMS) is an important part of the development of powertrain assembly and large-scale energy storage system for new energy vehicles. The battery state of charge (SOC) and state of health (SOH) prediction model are designed and constructed. For the entire battery state Control, improve battery life and system energy density, give full play to the battery capacity is of great significance. Currently, the SOC prediction model used in BMS design mostly adopts the open-circuit voltage method or the battery internal resistance method, which is practical but low in accuracy. However, there is a lack of research on the SOH prediction model of lithium-ion batteries and no breakthrough research results . Recently, Professor Ma Zifeng of Shanghai Jiaotong University, based on the multi-objective and uncertain design theory of fuel cell and carbon dioxide capture and storage system, cooperated with Zhongji Battery Research Institute and Shanghai Electrochemical Energy Device Engineering and Technology Research Institute Center for the solution to lithium battery high security, long-life battery management system difficult and hot issues, the chemical complex system of multi-objective, uncertainty and operational optimization theory applied to the BMS design theory, the establishment of Lithium - ion battery equivalent circuit model. In the MATLAB software, Simscape simulation platform of single Li-ion battery was developed to realize model parameter identification and dynamic simulation. For the first time, a polynomial open circuit voltage model based on nonlinear semi-infinite programming (NSIP) was proposed. The open circuit voltage (OCV) And prior knowledge of monotonicity of SOC are explicitly integrated into the modeling process and a global optimization method is proposed to ensure that the built OCV model can satisfy the monotonicity relationship in principle and improve the accuracy and stability of SOC prediction . In addition, the research team also creatively proposed a multi-scale Gaussian process model framework that can decouple the global trend of capacity decay and local capacity regeneration and fluctuation, and can simultaneously achieve the short-term and long-term accuracy of lithium-ion battery SOH The predictions provide a solid theoretical and methodological basis for battery state of charge (SOC) estimation and residual life expectancy (RUL) prediction. Using the proposed method to fit the data of three standard batteries of NASA's Ames Prediction Center, the results show that the long-term capacity decay, short-term capacity regeneration and fluctuation of lithium ion batteries can be deconstructed adaptively The type of trend greatly improves the predictive performance of the model. The modeling idea of ​​this research and development is a generalized method. As a pure data-driven modeling method, it is helpful to overcome the shortcomings of the existing mechanism models such as difficult parameter calibration, large amount of calculation, and difficulty in real-time on-line application . Airless Bottles,Eye Cream Plastic Bottle,Container Cosmetic Packaging Bottle,Pp Material 15 Ml Bottle Zhejiang Xinlei Packaging.Co.Ltd.(China) , https://www.xinleipack.com