A data-driven energy consumption prediction method for building electrical equipment based on data-driven

A data-driven energy consumption prediction method for building electrical equipment based on data-driven is proposed to address the issues of unstable prediction results and low accuracy in existing methods. Multiple sensors are selected to collect voltage, power, temperature and humidity data of electrical equipment. The mean filling method is used to fill in the missing values of the collected data. The K-means algorithm is used to detect anomalies in the filled data, identify and remove abnormal clusters or samples. Based on the data processing results, particle swarm optimisation algorithm is used to train energy consumption data, construct an energy consumption prediction model and achieve energy consumption detection through this model. The experimental results show that the highest prediction accuracy of this method is 98.5%, and the difference between the predicted results and actual energy consumption is small, indicating that the stability and robustness of this method are strong.