Hydropower unit health assessment based on a combination weighting and improved fuzzy comprehensive evaluation method

Abstract

The health state of hydroelectric power generation units is of great significance to ensure the stability and economy of safe operation of the power grid. In order to address the challenges in existing assessment methods of the insufficient reliability of the evaluation of multi-complex systems and the inability to reflect anomalies of a single index. A state evaluation model based on combination weighting and improved fuzzy comprehensive evaluation method is accordingly proposed. First, a hierarchical analysis system is constructed based on actual monitoring indicator data from the hydropower unit. Optimal comprehensive and indicator weights are subsequently obtained for each indicator level using a combination of the improved hierarchical analysis and CRICIT method through game theory. Next, the industry guidelines and regulations are difficult to effectively determine the limit values of each index of the unit, and they do not fully take into account the actual situation of the unit itself and the huge amount of accumulated historical health data. To address this issue. The Gaussian threshold method was proposed to determine the limit values of the monitoring data for each indicator, which more accurately determines the indicator thresholds as well as their standard values. The degradation degree of the hydroelectric unit can be calculated by comparing the real-time monitoring data with these limits. Finally, the combined weights of dynamic change and the fuzzy evaluation matrix are used to obtain the state evaluation matrix reflecting the condition of the turbine. The proposed approach is validated using the actual monitoring data and operating conditions for case study hydroelectric station, The results show that the improved evaluation method has an optimal evaluation effect.