Research The Technology Of Feature Extraction For Distributed Energy Storage Battery Network

With the development of renewable energy,the increase of random energy demand leads to the energy fluctuation and uncertainty of the Micro-grid,and the distributed storage battery charge-discharge is an important technical,which can effectively eliminate the energy fluctuations and uncertainties.Therefore,fully understood with the performance of the energy storage battery,build a battery model,extract characteristics of battery,and accurately predict the battery state have important significance which enhances the utilization of energy storage battery and extends the service life of battery.Initially,the paper had analyzed the influence of self-discharge characteristics,internal resistance,voltage,current and temperature of a lead-acid battery with its performance,and the interdependence and mutual restraint of parameters between batteries,an accurate high-order battery model was established.The parameters of the high-order model were identified through pulse discharge identification method and stage discharge method,and the high-order battery model was continually estimated,modified and optimized by the extended Kalman Filtering(EKF),which improved the accuracy of SOC estimation.Simulation analysis and experimental tests verified that the accuracy and validity of the higher-order model.Additionally,the adaptive reduced order optimization method for energy storage battery was researched based on the model.Many parameters were filtered by the fuzzy optimization strategy,and the complexity of solving the high-order model was reduced when the accuracy was ensured.Simulation analysis and experimental tests verified the validity of the reduced order model,which applied to the series batteries and the parallel batteries.Finally,the high-order battery model and optimization algorithm of adaptive reduced order could be applied to a management system for distributed storage battery,which achieved the sharing of information in real-time,reduced the load for system maintenance and improved the management for system efficiency.It can provide technical supports for distributed energy storage battery of real-time monitoring and management.