Digital twins of ball valve’s pressure oil system for efficient operation and maintenance

Abstract

As a crucial auxiliary system of hydropower stations with long diversion pipelines, the pressure oil system (POS) undertakes the control of ball valve opening and closing, which directly determines the safe and stable operation of hydropower units. To meet the industrial demand of intelligent operation and maintenance (iOM), this paper constructs a dedicated digital twin (DT) model for the POS, consisting of three sub-models: digital simulation sub-model, state evaluation and fault diagnosis sub-model, and maintenance decision support sub-model. A DT iOM demonstration system (DT_iOMS) is developed based on the model framework to realize full-condition simulation of the POS. Combined with the analytic hierarchy process and grey relational analysis, rapid evaluation of equipment operating state is realized, and maintenance schemes are automatically generated by integrating three-dimensional virtual models and intelligent algorithms. Test results show that the established DT system can realize fault early warning 15 min in advance with an average prediction error of only 0.94%, forming a complete technical framework for iOM of the POS. The proposed method can provide feasible technical solutions for similar auxiliary equipment such as water supply, drainage and gas supply systems in hydropower stations, and has prominent engineering application value.