Development of an Integrated Expert Diagnostic System for The Steam Boiler

[[abstract]]本研究之目的在針對工廠中小型蒸汽鍋爐建立一套綜合診斷專家系統程式，此類鍋爐的圖形監測多只有監視而無控制功能。製程設備參數可區分為製程控制參數與結構完整性參數，製程控制參數用於操作與控制，結構完整性參數則多用於設備診斷，此二種參數間存在著相關性，但是工廠往往是分開蒐集與處理，並未將二者結合使用。本研究以蒸汽鍋爐為研究對象，由鍋爐的設計建造規範、運轉操作流程、維修保養程序以及現場操作人員的經驗等相關資料蒐集與整理，建立故障樹(FTA)、失效模式與效應分析(FMEA)與決策樹，探討各種參數間的相關性與影響關係，進而建立綜合診斷規則與診斷系統。研究後發現，專家系統與不同檢測週期參數的結合應用，在經過診斷後所提供的資訊，除了可能的故障部位、故障的原因、故障的可能後果，亦提供異常處置對策，這些資訊確實能在異常發生初期時便針對根源做處置。本系統除了主要診斷功能外，也將提供其他額外計算功能，如：剩餘壽命計算，新檢測週期計算以及鍋爐效率計算…等。[[abstract]]The research focuses on developing an integrated expert diagnostic system for steam boilers. Control systems found on steam boilers commonly seen in small to medium sized process plants function mainly as providing parameter readings and less in actually controlling the process. The user must rely on strong empirical knowledge in addition to inspection techniques in order to diagnose boiler problems.Two types of equipment parameters are utilized in developing the expert system, one is operation related and the other is structural integrity related. The domain knowledge and rules of diagnosis were collected and transformed from a wide variety of materials including design/fabrication standards, operation conventions, maintenance and inspection procedures, and, last but no the least, operators’ experience. Then bits and pieces of raw information were then represented in the form for failure mode and effect analysis (FMEA) and fault tree analysis (FTA). The expert system is designed not to replace the current boiler control system in the field but to serve as a decision support tool for the operator when the anomaly emerges. The most difficult part of the research is to overcome the widely scattered data collection intervals ranging from a couple of seconds to several years, while maintaining acceptable diagnosis. The tool also automate some of the cumbersome computation process such as remaining life, next inspection time, and boiler efficiency