Intelligent regulation method for the thermal inlet of a dynamic distributed central heating system

Abstract

Disclosed is an intelligent regulation method for the thermal inlet of a dynamic distributed central heating system, including the following steps: first, performing multi-source heterogeneous data collection, covering building indoor temperature, outdoor weather and user behavior pattern data; then constructing a temperature prediction model based on LSTM neural network with the introduction of an attention mechanism to realize dynamic weight fusion for generating a fused feature vector; then generating a thermal inlet water pump flow regulation sequence; subsequently deploying edge computing nodes at the thermal inlet to perform local preprocessing and preliminary data prediction before uploading; finally establishing a climate zone characteristic database to dynamically adjust model parameters according to the climate zone characteristics where the building is located. The method comprehensively improves the system's adaptability to complex conditions and optimizes the regulation accuracy and energy efficiency of the thermal inlet.