Speaker
Chen-yang Lan (藍振洋), Assistant Professor, National Taiwan University of Science and Technology
Chen-yang Lan (藍振洋), Assistant Professor, National Taiwan University of Science and Technology
Topic
Model-Based Analysis for the Predictive Maintenance of Rotating Machines
轉動設備預知保養使用的模型式分析方法
Abstract
Due to the ID4, IoT and Big Data, the high-tech intelligent facility has become a key competence for many companies. In many advanced functions required by this intelligent facility, how to effectively and intelligently operate and maintain rotating machine is among one of the focuses. Through most rotary machines, steady supply is provided for production. As a result, rotary machines are the most common and popular equipment at the facility and also consume a major portion of energy in production cost. Using technologies for condition monitoring, the current condition of each machine is better understood for effective operation and maintenance, not to mention the improvement in reliability for automated system. In addition, upkeeping the machine for better performance can also save energy. And connecting the measured data for real time analysis can help in avoiding downside associated with the route inspection.
Driven by above motivations, the technology for condition monitoring has drawn much attention again. Traditionally, vibration analysis is the main stream in condition monitoring for rotary machine. However, the high initial investment cost to install such technique, due to the cost of high bandwidth sensor, instrumentation and expert system, and the knowledge and training required for staffs prevent the propagation and deployment of this technology to most rotary machines. As a result, an alternative method could be considered using the motor electrical signals. This technology is described as in ISO-20958 and published in many references. It could detect both electrical and mechanical faults and analyzes the frequency peaks appeared in motor current signal owing to mechanical vibration for mechanical faults. This method requires only the measurement of electrical signal from motor and makes it easier to implement with better coverage for condition monitoring. Therefore, it is truly the effective way to make facility smart.
因為工業4.0、工業物聯網及大數據分析等應用的蓬勃發展及誘因，智慧化高科技廠房設施已然成為各廠爭相競逐的目標之一。此高科技廠房在諸多的設定目標中，如何有效益的智慧化轉動設備之運轉與維護，將是重要的課題。轉動設備主要提供穩定的基礎供給給生產線，在廠務諸多設備中是最普遍，其能耗也是生產成本中的一個大項。而透過設備狀態監診技術，將可有效瞭解各設備之現況，適時進行運轉調配與保養措施，進而提高自動化系統可靠度。而有效維持轉動設備運轉狀態，也可達到節能的效益。對各設備的即時量測資料連線與自動分析，更可避免巡檢判斷上的缺點。
在此前提下，轉動設備的狀態監診技術又再度引起關注。傳統上，振動訊號量測與分析是最主要的工具。然而，昂貴的感測器與儀器設備及專家系統建置費用，始終讓人卻步。另外振動訊號量測與分析方法的先天限制，也讓振動訊號狀態監診技術應用普及始終難以提高。因此，可以考慮以馬達電訊號為主的狀態監診技術，如ISO-20958中的電訊號分析及模型式電訊號分析的技術。此技術利用振動訊號所引起的電流頻率峰值，檢知設備異常狀態。可用於檢知轉動設備的機械與電氣異常。其建置使用上，也只需量測分析馬達的電訊號，實為智慧化高科技廠房經濟有效的實施方法。
Model-Based Analysis for the Predictive Maintenance of Rotating Machines
轉動設備預知保養使用的模型式分析方法
Abstract
Due to the ID4, IoT and Big Data, the high-tech intelligent facility has become a key competence for many companies. In many advanced functions required by this intelligent facility, how to effectively and intelligently operate and maintain rotating machine is among one of the focuses. Through most rotary machines, steady supply is provided for production. As a result, rotary machines are the most common and popular equipment at the facility and also consume a major portion of energy in production cost. Using technologies for condition monitoring, the current condition of each machine is better understood for effective operation and maintenance, not to mention the improvement in reliability for automated system. In addition, upkeeping the machine for better performance can also save energy. And connecting the measured data for real time analysis can help in avoiding downside associated with the route inspection.
Driven by above motivations, the technology for condition monitoring has drawn much attention again. Traditionally, vibration analysis is the main stream in condition monitoring for rotary machine. However, the high initial investment cost to install such technique, due to the cost of high bandwidth sensor, instrumentation and expert system, and the knowledge and training required for staffs prevent the propagation and deployment of this technology to most rotary machines. As a result, an alternative method could be considered using the motor electrical signals. This technology is described as in ISO-20958 and published in many references. It could detect both electrical and mechanical faults and analyzes the frequency peaks appeared in motor current signal owing to mechanical vibration for mechanical faults. This method requires only the measurement of electrical signal from motor and makes it easier to implement with better coverage for condition monitoring. Therefore, it is truly the effective way to make facility smart.
因為工業4.0、工業物聯網及大數據分析等應用的蓬勃發展及誘因，智慧化高科技廠房設施已然成為各廠爭相競逐的目標之一。此高科技廠房在諸多的設定目標中，如何有效益的智慧化轉動設備之運轉與維護，將是重要的課題。轉動設備主要提供穩定的基礎供給給生產線，在廠務諸多設備中是最普遍，其能耗也是生產成本中的一個大項。而透過設備狀態監診技術，將可有效瞭解各設備之現況，適時進行運轉調配與保養措施，進而提高自動化系統可靠度。而有效維持轉動設備運轉狀態，也可達到節能的效益。對各設備的即時量測資料連線與自動分析，更可避免巡檢判斷上的缺點。
在此前提下，轉動設備的狀態監診技術又再度引起關注。傳統上，振動訊號量測與分析是最主要的工具。然而，昂貴的感測器與儀器設備及專家系統建置費用，始終讓人卻步。另外振動訊號量測與分析方法的先天限制，也讓振動訊號狀態監診技術應用普及始終難以提高。因此，可以考慮以馬達電訊號為主的狀態監診技術，如ISO-20958中的電訊號分析及模型式電訊號分析的技術。此技術利用振動訊號所引起的電流頻率峰值，檢知設備異常狀態。可用於檢知轉動設備的機械與電氣異常。其建置使用上，也只需量測分析馬達的電訊號，實為智慧化高科技廠房經濟有效的實施方法。