Monitoring and Evaluation of Health Status of Large Complex Systems Based on Multi-source Information Fusion
- Chinese keyword
- Large complex systems, health status, real-time monitoring and evaluation, multi-source information fusion, artificial intelligence, big data analysis, Bayesian network
- English keyword
The word "health state" was first borrowed from the field of biology to describe the various working states that a system may have between "failure" and "normal". Initially, engineers only used "fault" and "normal" binary functions to judge the working state of the system, With the continuous development of related technologies, people found that it is incomplete to define the state of the system only by this binary function. Later, the word "health state" was quoted from the biological field to divide the working state of complex systems into multiple levels to describe the multi-value state of the system. Usually, under the condition that the system is not subjected to excessive or sudden stress, it is not a sudden change process from the normal state of the factory to the final fault state of the complex system from the microscopic level. The degradation of the health state of the system is a gradual process in which the performance of the equipment gradually decreases from normal to functional failure (failure) with the passage of the working time of the equipment.
Health status assessment refers to relying on advanced detection methods, Combining reliable and effective evaluation methods and complete operation data to judge the system, Its purpose is to manage and evaluate the data, and adopt active measures to monitor the health status of the equipment system and give an alarm to the aging of the equipment in time. It can effectively improve the maintenance support capability of the system, reduce maintenance costs, save spare parts and effectively prevent sudden failures. Health status assessment can generally be divided into three categories: health status assessment based on equipment online monitoring data; Health status assessment based on off-line preventive test data of equipment: Comprehensive health status assessment based on the fusion of on-line monitoring data and off-line preventive test data of equipment. However, it is an important research direction and development trend in this research field to fuse multi-source information of the system and use advanced means such as artificial intelligence and big data analysis to monitor and evaluate the health status of the system.
With the rapid development of modern science and technology, products or systems tend to be multifunctional, integrated, complicated and intelligent. The reliability, stability, fault diagnosis and prediction, maintenance support and other issues of the system have attracted more and more attention. Under normal circumstances, complex systems will experience a series of performance degradation states to varying degrees from normal to complete failure. Therefore, if the degree of performance degradation can be accurately monitored in the process of performance degradation of complex systems, then equipment maintenance plans can be formulated in a targeted way, thus effectively avoiding equipment failure due to failure. Health assessment of complex systems is an active maintenance technology based on this idea, which is quite different from traditional fault diagnosis. Fault diagnosis technology is a passive maintenance mode that repairs faults in time and effectively on the basis of finding faults. However, health assessment technology focuses on the analysis of the performance degradation trend of complex systems in the whole life cycle and is an active maintenance mode.
Condition monitoring of complex systems has always been a key research direction in industry and academia. It can realize the information perception, information transmission, calculation and evaluation, behavior prediction of the system operation state, and on this basis, fast fault maintenance, condition-based maintenance and various deep-level system decisions, thus helping to reduce the operation risks and maintenance costs of complex systems, and improving the systematic decision-making ability and management efficiency. With the further promotion of the "Industry 4.0" and "Made in China 2025" strategies, the ecosystem with industrial Internet and industrial big data technology as its core is rapidly becoming the key development direction and key technical support in the field of industrial manufacturing. As a result, in the fields of advanced equipment manufacturing such as aviation, aerospace, navigation, industrial manufacturing, nuclear, new energy, high-speed rail, etc., intelligent diagnosis and prediction technology for complex systems has rapidly become one of the cutting-edge technologies and hot research directions supporting the development of technical systems in various fields. At the same time, more and more research practices and application evaluations have further proved that system health management based on condition monitoring and testing has important value and significance for system condition evaluation, anomaly discovery, rapid diagnosis, accurate maintenance, and improvement of system operation reliability and safety.