基于BP神经网络的太阳能路灯SOC预测

SOC Prediction of Solar Street Lamps Based on BP Neural Network

针对光伏发电应用领域太阳能路灯系统的过充电或过放电现象对蓄电池本身特性产生影响、降低使用寿命的问题,采用单片机和LabVIEW进行太阳能路灯蓄电池电压检测,采用BP神经网络进行太阳能路灯蓄电池荷电率(SOC)预测。BP神经网络将测得数据建立SOC(State of Charge)预测模型,LabVIEW可视化面板实时显示测量数据、波形及预测结果,实现太阳能路灯智能化控制。测试结果表明,系统能够实时检测蓄电池充电电压,并预测电池工作状态,BP神经网络蓄电池SOC预测值与蓄电池电量实测误差为0.1%~0.4%,满足网络误差要求。

In view of the problem that the overcharging or overdischarge phenomenon of solar street lamp system in the application field of photovoltaic power generation affects the characteristics of the battery itself,single chip computer and LabVIEW are used to conduct solar street lamp battery voltage detection,and BP neural network is used to predict the state of charge(SOC)of solar street lamp battery.A BP neural network SOC prediction model is established for the measured data,and the LabVIEW visualization panel displays the measurement data,waveform and prediction results in real time,so as to realize the intelligent control of solar street lamps.The test results show that the system can detect the battery charging voltage in real time and predict the operating state of the battery.The error of the predicted value of the battery SOC and the measured battery power is about 0.1%~0.4%,which meets the network error requirements.