Load Forecasting for Steel Enterprises Based on Prophet and Bi-LSTM Algorithms

Power load forecasting is of great significance in the production and operation activities of steel enterprises. Based on the existing Bi LSTM power load forecasting problems of gradient vanishing and explosion, this paper proposes a power load forecasting method that integrates Prophet algorithm and Bi LSTM algorithm. On the basis of the Bi-LSTM model, the Prophet algorithm is used to solve the problem of gradient vanishing and exploding in the Bi LSTM model after using the backpropagation algorithm, thereby improving the accuracy of power load forecasting. The experimental results show that using the model proposed in this article for load forecasting has higher accuracy than traditional RNN, LSTM, and Bi LSTM algorithms, verifying the effectiveness of this algorithm in improving the accuracy of load forecasting.