State pre-warning and optimization for rotating-machinery maintenance

Maintenance of rotating machinery has significant practical implications for preserving the service condition and quality of products. Moreover, it directly affects the economic efficiency of enterprises. Although frequent maintenance can preserve the condition and quality of products, it can increase the cost of enterprises. Conversely, long intervals in maintenance can prove to be economical but would not ensure the desired condition and quality. This study presented a real-time maintenance strategy which was based on condition assessment using the fuzzy C-means method and the kurtosis index. Changes in the kurtosis index can be monitored to successfully capture the features of early faults. The performance condition was assessed using the fuzzy C-means method, and the result was considered as the reliability of the equipment. Enterprise-efficiency optimization was regarded as a proposed criterion to make a real-time maintenance recommendation. The result of analyzing data from a steel enterprise shows that this real-time maintenance strategy is more suitable for the management of on-site equipments. Moreover, it reduces the monitoring cost, thereby obtaining increased enterprise benefit.