Hierarchical clustering-based short-term prediction of tower inclination angle by Bi-LSTM

Abstract A hybrid prediction algorithm based on hierarchical clustering and Beluga Whale Optimization (BWO) algorithm, and Bi-LSTM network is proposed for the high voltage transmission lines so as to face the inclination angle of the pole towers caused by different weather reasons in the future and take safety protection measures. Using the dataset of historical tower tilt angle with a timestamp, it is verified by example that the short-term prediction of tower tilt angle based on hierarchical clustering BWO-Bi-LSTM tower tilt angle is able to predict the future tower tilt angle under any weather, which overcomes the limitations of the traditional Bi-LSTM model prediction method in accuracy and randomness of the influence of weather factors on tower tilt angle. and it has great practical value for the high-voltage lines in the tower. t has important practical value and significance for maintaining the safety of tower inclination of high voltage lines.