Bus voltage prediction method based on deep learning algorithm

The invention discloses a bus voltage prediction method based on a deep learning algorithm. The method comprises the following steps: S1, obtaining bus voltage, active power, reactive power and frequency historical data; s2, preprocessing the acquired data, wherein the preprocessing comprises missing value cleaning and normalization; s3, training an LSTM model by using historical data; and S4, predicting the bus voltage situation by using the trained model. According to the bus voltage prediction method based on the deep learning algorithm provided by the invention, the bus voltage fluctuation can be predicted. 本发明公开了一种基于深度学习算法的母线电压预测方法，包括如下步骤：S1、获取母线电压、有功功率、无功功率、频率历史数据；S2、将获取的数据进行预处理，包括缺失值清理和归一化；S3、使用历史数据训练LSTM模型；S4、使用训练好的模型对母线电压态势进行预测。本发明提供的基于深度学习算法的母线电压预测方法，能够实现对母线电压波动进行预测。