Green building energy consumption composite prediction method

Abstract

The invention discloses a green building energy consumption composite prediction method, relates to the technical field of electric power demand side energy consumption analysis and prediction, and particularly relates to a building electric energy consumption prediction model. The method comprises: monitoring personnel information, building body information, building external environment conditions and other information on the basis that historical electric power consumption data are collected on the basis of a big data center. The big data center can observe the energy consumption condition of the building demand side in real time; meanwhile, the energy consumption is evaluated and diagnosed through big data mining. On the premise of recording climate and power consumption, an FCM fuzzy C-means method is adopted to carry out data cleaning on power data of a demand side. An analytic hierarchy process is adopted to carry out data reduction on collected building basic information and main factors influencing energy consumption, and then a swarm intelligence algorithm (PSO-BP prediction model) is adopted to propose an optimal energy-saving strategy.

Description

technical field [0001] The invention belongs to the technical field of building energy conservation and electric energy analysis and prediction, and in particular relates to a compound prediction method for green building energy consumption. Background technique [0002] With China's urbanization process, the proportion of building energy consumption in the country's total energy consumption continues to rise, and is expected to eventually reach about 35%. According to the "China Building Energy Consumption Research Report 2018" statistics, China's total building energy consumption in 2016 was 899 million tons of standard coal, accounting for 20.6% of the country's total energy consumption. This result is 0.6% higher than that in 2015, and electricity as a major building energy consumption accounted for 46%, and public buildings as a major energy consumption terminal accounted for 38%. Therefore, accurately analyzing and predicting the collected building power consumption d...

AI Technical Summary

Problems solved by technology

In the face of massive data, the traditional building energy consumption prediction method is difficult to realize the energy consumption information management of actual public buildings, and the energy consumption prediction does not consider the physical meaning, and there is an error of 30%-50%, which poses a great obstacle to the application of actual projects

[0011] In addition, although the algorithm of optimizing the BP neural network by using particle swarm can greatly avoid the defect of the model falling into local optimum, the defects of poor correlation between model input variables and output variables and high redundancy leading to poor prediction accuracy of the model are still there. well resolved

Images