Research And Implementation Of Parameter Identification And Short-term Fault Early Warning Of High Specific Power Battery

In recent years,with the rapid development of economy,people’s material living standards have been continuously improved.The demand for urban transportation has gradually increased,and more and more families have begun to buy cars.While cars are convenient for people to travel,they also cause problems such as environmental pollution and energy shortages.But the emergence of new energy vehicles is expected to solve these problems.China attaches great importance to the development of new energy vehicles,and has also released many key research and development plans and special projects on new energy vehicles.The project High specific power long-life power battery and new super capacitor technology development is one of them.The research content of this paper is related to it.This paper mainly studies on the parameter identification and short-term fault early warning of high specific power battery,and applies the research achievements to the short-term fault early warning system of high specific power battery.There are three main contributions of this paper.First,an improved particle swarm algorithm is proposed for parameter identification.After understanding and comparing common high specific power battery models,a suitable battery model is select;then,based on the equation of state of the model,choose to use the particle swarm optimization algorithm in the intelligent optimization algorithm for parameter identification;then improve the particle swarm optimization Algorithm.Finally,experiments were performed to prove the effectiveness of the improved particle swarm algorithm in battery parameter identification.Second,effective early warning models of short-term faults were trained.Fully analyzed the data reported by electric vehicles,a method for preprocessing the data is proposed.Afterwards,an improved random forest algorithm is proposed to train each short-term failure early warning model.Finally,the appropriate model evaluation indicators,Accuracy and AUC(Area Under Curve),are selecte to evaluate each early warning model objectively and determined that all early warning models are effective models.Third,a short-term fault early warning system is designed and developed.Starting from the current status of existing systems and user expectations,the problems and pain points that the early warning system needs to solve is understood.And the needs of the early warning system based on this is analyzed.Then the system design is proposed,including the system architecture design,the design of database and the detailed design of the core module(the early warning module).Finally,the implementation of the early warning system is described.The final effect and test result of the early warning system is shown.