Thread Content
Condition-based maintenance is also known as predictive maintenance (Condition Based Maintenance, CBM; Predictive Maintenance, abbreviated as PdM). It is the different names we use in Chinese translation for the same type of maintenance work. Predictive maintenance integrates equipment condition monitoring, fault diagnosis, fault (condition) prediction, maintenance decision support, and maintenance activities; it is an emerging maintenance strategy that offers the best economic benefits (hereafter referred to collectively as predictive maintenance). Predictive maintenance differs not only in its name but also in the connotation and extension of its concept; therefore, there are two notions of it: narrow and broad predictive maintenance. Predictive maintenance in a narrow sense is based on condition monitoring; it emphasizes \"fault diagnosis\" and involves periodically or continuously monitoring the condition of equipment. Based on the results of this monitoring, it is determined whether there are any abnormalities in the equipment’s condition or signs of impending failure, allowing for timely maintenance to be scheduled. In a narrow sense, predictive maintenance does not fix a predetermined maintenance schedule; instead, it arranges maintenance activities at appropriate times based on the results of monitoring and diagnosis. It emphasizes a integrated process of monitoring, diagnosis, and maintenance, and this approach is widely applicable to process industries and large-scale production methods. In a broad sense, predictive maintenance integrates condition monitoring, fault diagnosis, condition prediction, and maintenance decision-making. Condition monitoring and fault diagnosis serve as the foundation; condition prediction is the key aspect; and maintenance decision-making determines the final requirements for maintenance activities. Broadly defined, predictive maintenance is a systematic process that incorporates maintenance management into the framework of predictive maintenance, taking into account the entire maintenance process as a whole in order to derive information related to maintenance activities. Break-down maintenance is a \"failure-based\" approach to maintenance; it involves repairing equipment only when it is damaged or no longer functional. It restores the equipment to its original state only after it has suffered partial or total damage, meaning repairs are carried out after the equipment has been used up, and thus it falls under the category of unplanned maintenance. Preventive maintenance, also known as scheduled maintenance, is a type of maintenance that is based on time. It involves performing shutdown inspections, disassembly, and replacement of components at regular intervals, as determined by production plans and experience, in order to prevent damage, secondary damage, and production losses. This method of maintenance is what is currently commonly used, namely scheduled maintenance or regular maintenance, such as annual, semi-annual, quarterly, or monthly upkeep. Predictive maintenance technology framework: By now, predictive maintenance has developed its own technical framework, as shown in the diagram. Early detection of equipment failures: The P-F curve shows that point P represents the point at which defects in the equipment can be detected, while point F denotes the point of functional failure. The time elapsed from the emergence of defect-related information to the occurrence of functional failure is referred to as the P-F interval. To prevent functional failures, maintenance should be carried out before point F; whereas, to make the most of the effective service life of the equipment or components, maintenance should be done after point P. This means that a suitable point should be found between points P and F for repairs; this is the basic idea behind \"predictive–condition-based maintenance\". 1. Condition monitoring technology: With the development of condition monitoring technology, various monitoring methods have been developed for different engineering fields. These monitoring methods can be classified into many types depending on the means used for condition detection. Common methods include online & offline vibration meter monitoring, ultrasonic noise monitoring, offline temperature monitoring using thermal imaging & infrared guns, as well as online temperature sensor monitoring; online & offline pressure monitoring, oil analysis monitoring, acoustic emission monitoring, and others. 2. Fault diagnosis technology – Focusing on “fault diagnosis” alone, it is a science that has developed recently and is receiving increasing attention; especially in continuous production systems, fault diagnosis holds great significance. Based on the principles of diagnostic methods, fault diagnosis can be divided into: time-frequency diagnosis, statistical diagnosis, information theory analysis, as well as other artificial intelligence approaches (such as expert system diagnosis and artificial neural network diagnosis), fuzzy diagnosis, grey system theory diagnosis, and integrated diagnosis (such as fuzzy expert system fault diagnosis, neural network expert system fault diagnosis, fuzzy neural network diagnosis, etc.). 3. State prediction technology: State prediction involves using the operational information that can be monitored of a device to assess the current state of its components and predict their future state. The commonly used methods include time-series model prediction, grey model prediction, and neural network prediction. There are generally three basic approaches to the development of prediction methods: physical models, knowledge systems, and statistical models. In practical applications, the three approaches can be combined to create a mixed fault prediction technique that integrates traditional physical models with intelligent analysis methods, and is capable of handling both digital and symbolic information, thereby making it more effective for implementing predictive maintenance.