The power load prediction of green building based on multidimensional data mining

In order to solve the problems of low recall and precision and high-prediction error in traditional prediction methods, a power load prediction of green building based on multidimensional data mining is proposed. The initial clustering centre and feature weight of fuzzy k-means algorithm (FKM) clustering algorithm are optimised, and the improved FKM clustering algorithm is used to mine multi-dimensional green building power load data. The multi-dimensional data mining results were taken as sample data, and the Least Squares Support Vector Machine (LSSVM) model parameters were optimised by Particle Swarm Optimisation with Extended Memory (PSOEM) algorithm. The sample data were input into the optimised model to obtain the power load prediction results of green buildings. The experimental results show that the average recall rate and precision rate of the proposed method are 96.31% and 96.13%, respectively, and the prediction error rate fluctuates between -2% and 2%, indicating high-prediction accuracy.