A novel prognostic model of performance degradation trend for power machinery maintenance

Power machinery has two types of fault modes. The first type leads equipment to stop working, and the second one results in performance degradation. The second type should not be ignored, because of its safe, economic and environmental consequence. Aiming at the second type of fault modes, current prognostic model for the remaining useful life of equipment is usually based on the historical data of the equipment fault or malfunction, which can provide evidence for maintenance. However, this model just depends on the time based fault data, without taking the operation state into consideration. In this paper, a novel prognostic model of performance degradation trend is developed, which is based on current prognostic models for the remaining useful life. It combines the historical fault data and monitoring data in operation. This model can be used for maintenance optimization. Maintenance activities, according to the result of this model, actually combine the viewpoint of Time Based Maintenance and Condition Based Maintenance. Finally, compressor washing of a gas turbine engine is cited as an instance to validate this model. Maintenance strategies based on the new model infers that it not only keeps the reliability of equipment, but also reduces the maintenance cost. •The relationship between forecasting prognosis and statistical prognosis is established.•A performance degradation trend prediction model for power machinery is built.•The new prognostic model is applied to gas turbine compressor washing.•The new prognostic model is better than regression model in 7 real cases.•The new prognostic model is used for maintenance optimization.