基于ANFIS模型的蓄电池放电剩余电量估计

Estimation of Remaining Electricity in Battery Discharge Based on ANFIS Model

针对变电站直流电源的蓄电池放电后剩余电量(SOC)难以评估的问题,在简述传统蓄电池SOC估计模型的基础上,分析了人工神经网络和模糊逻辑相结合的自适应神经模糊推理系统(ANFIS),并对蓄电池的SOC进行预测。该蓄电池剩余容量模型具有更强的泛化能力,适应性和高精度性。通过对电池充放电过程的分析,确定了SOC的关键参数,并在Mat Lab平台上对实验模型进行了修正。试验结果表明SOC预测与实际SOC的差异小于3%。该模型可以有效反映出电池的特性。SOC估计算法能够满足精度要求,并且测试结果具有较高的实用价值。

Aiming at the problem that it is difficult to evaluate the SOC of DC power supply battery after discharging in substation,this paper analyzes the adaptive neural fuzzy inference system(ANFIS)based on artificial neural network and fuzzy logic,and predicts the SOC of storage battery.The residual capacity model of the battery has stronger generalization ability,adaptability and high accuracy.Through the analysis of battery charging and discharging process,the key parameters of SOC were determined,and the experimental model was modified on the MatLab platform.The experimental results show that the difference between SOC prediction and actual SOC is less than 3%.The model can effectively reflect the characteristics of battery.SOC estimation algorithm can meet the accuracy requirements,and the test results have high practical value.