Study on Health Assessment and Residual Useful Life Prediction of Wind Turbine

Abstract

Health management and maintenance problems are the bottlenecks that hinder the development of wind turbines. The complex multi-failure mode makes the management of complex equipment and life prediction difficult. In this paper, a framework of health assessment and residual life prediction based on multi-dimensional performance and multi-failure mode of wind turbines is proposed. It includes five parts: the nonlinear performance degradation model based on repeated error variation; the graph model of multi-performance and multi-fault mapping of wind turbines; the health exponential function model based on graph theory matrix and the TOPSIS partition of health state space; a two-tier residual life fusion strategy from “multi-performance feature $\rightarrow$ failure mode” to “multi-failure mode $\rightarrow$ wind turbine”; the decision-making model and optimization algorithm based on health assessment and residual life prediction.