Building energy consumption prediction method based on BP neural network of MapReduce

Abstract

The invention provides a BP neural network building energy consumption prediction method based on MapReduce. By using MapReduce and BP neural network, the connection weights in neural network are adjusted and corrected through parallel computation, and the problem of low accuracy of building energy consumption prediction is solved, and the BP neural network is easy to fall into local extremum andlow precision of solution.

Description

technical field [0001] The invention belongs to the field of building energy consumption prediction, in particular to a MapReduce-based BP neural network building energy consumption prediction method. Background technique [0002] With the continuous acceleration of the urbanization process, the energy problem has become increasingly prominent. Building energy conservation has become a research hotspot in today's social development. A comprehensive evaluation and comprehensive analysis of building system energy consumption is the premise and basis for energy-saving renovation or energy-saving design. The establishment of a prediction model that reflects changes in energy consumption is an effective way and an important means to analyze and understand the changes and development characteristics of building energy consumption on a macro scale to provide decision-making basis for energy conservation in public buildings. [0003] BP neural network prediction has a strong nonline...

AI Technical Summary

Problems solved by technology

[0004] The technical problem solved by the present invention is to provide a BP neural network building energy consumption prediction method based on MapReduce, which uses MapReduce in combination with BP neural network to adjust and correct the connection weights in the neural network through parallel computing, and realizes the building energy consumption prediction method. The energy consumption prediction of the building solves the problem of low accuracy of building energy consumption prediction results, and the problem of BP neural network easily falling into local extreme values ​​and low solution accuracy

Images