工業技術研究院

技術簡介

開發空調箱自動化檢測方法與系統，即時監測空調箱操作錯誤以及導入數據強化前處理的機器學演算法提升故障分類準確度。同時，系統包含調控決策功能，當故障成立後進行數據整合並且提供其他子系統因應的操作條件來維持整體建物的總空調能耗。

Abstract

An automated detection method and system for Air Handling Units, real-time monitoring of operating errors and data-enhanced machine learning algorithms are used to improve fault classification accuracy. This system also includes a decision analysis function that integrates data when a fault is established and provides corresponding operating conditions for other subsystems to maintain the total air conditioning energy consumption of the entire building.

技術規格

•控制參數監測：多段式函數轉換，以二維四象限圖像監測控制參數與目標參數的相對偏移量，使設備可維持良好的運轉效率 •故障分類技術：具數據強化前處理的分類演算法，提高分類準確度 •濾網/皮帶更換建議：監控且記錄處理效率及運行電流的變化趨勢，主動計算累計衰減率，並提供需於何時更換之建議 •調控決策：故障成立後進行數據整合，提供其他子系統因應的操作條件維持整體建物的總空調能耗

Technical Specification

•Control parameter monitoring: multi-stage function conversion, using two-dimensional images to monitor the residual difference between the control parameters and the target parameters to maintain good operating efficiency of the equipment. •Fault classification technology: Classification algorithm with data-enhanced pre-processing to improve classification accuracy. •Filter/belt replacement recommendations: monitor and record the changing trend of current, proactively calculate the cumulative attenuation rate, and provide replacement recommendations. •Control decisions: After the fault is established, data integration is performed to provide corresponding operating conditions for other subsystems to maintain the total air conditioning energy consumption of the entire building.

技術特色

一種用於空氣調節檢測方法與控制系統，可連結空調系統的子系統設備，主動診斷空調箱設備以達到互連決策控制，系統包含對場域目標與控制項檢測、多系統協調控制以及異常資料庫的故障分類。

應用範圍

系統整合商、維運保養業者

接受技術者具備基礎建議（設備）

伺服器、網路資訊設備

接受技術者具備基礎建議（專業）

伺服器管理、SQL資料庫管理

聯絡資訊