Research on Switchgear Fault Warning Based on Improved HHO-LSTM

Switchgear is a key component of the safe and stable operation of the power supply system. If a switchgear malfunction occurs, it not only interferes with the normal power supply process, but also may trigger a serious safety accident. To ensure the safe and reliable operation of switchgear and to minimize the risk of potential accidents, it is essential to establish a mechanism capable of accurately forecasting changes in the switchgear's operational status and to implement an effective fault early warning system. As a result, an efficient intelligent fault warning model IHHO-LSTM is proposed by combining the improved Harris hawk algorithm with the long and short-term memory neural network, which introduces adaptive inertia weights, nonlinear escape energy factor, and Tent chaotic mapping to improve the original Harris hawk algorithm, which are designed to improve the algorithm's global searching ability and accelerate the convergence speed so as to enhance the overall efficiency of the solving process. LSTM neural network models are excellent in the field of prediction, but their performance is largely dependent on the hyperparameter settings. Traditional hyperparameter optimization methods are time-consuming and inefficient. To enhance predictive accuracy, this research employs an improved Harris Hawk Optimization algorithm to swiftly identify the optimal hyperparameters for the LSTM model. The method is used to analyze the historical data of switchgear operation in a city in North China, and by analyzing the experimental results, it can be seen that the fault early warning model is able to accurately predict the switchgear state volume data and provide fault early warning for it.