Research On Running State And Risk Assessment Method For Induced Draft Fan In Power Station

With the reform of the power industry system,power generation companies are facing increasing competition pressure.In order to maximize the economic benefits of power plants,the demand of improving the reliability of equipment and reducing maintenance costs has become increasingly urgent.The implementation of condition-based maintenance has become an inevitable trend for the development of electric power enterprises.As the main research focus of state maintenance,running state and risk assessment technology is of great significance to the safe and stable operation of equipment and the implementation of condition-based maintenance.The accumulation of massive data resources in power plants also lays a solid research foundation for condition assessment research.Therefore,this paper carries out the research on running state and risk assessment method for induced draft fan in power station.First,in order to cooperate with the implementation of equipment state maintenance,based on the historical operation data of the SIS system in power station,the upper and lower thresholds of the operating parameters are calculated by cluster analysis to determine the operating threshold and the allowable threshold under different load conditions.Besides,a fuzzy state assessment model based on cluster analysis is established.Taking the induction fan motor of thermal power plant as an example,the model give a reasonable judgment on the state assessment results.Secondly,in view of the insufficiency of the above-mentioned method in the application of fixed thresholds due to the change of operating conditions to the uncertainty of the evaluation results,this paper also proposes a state assessment method using an auto encoder model combined with fuzzy analysis to solve the shortcomings of the above schemes in practical applications.Taking the induced draft fan of thermal power plant as the research object,after selecting the key monitoring parameters,the auto encoder model between parameters is established.The model is trained by the historical operation data of the induced draft fan,and the distribution characteristics of the training errors of each parameter are counted.Combined with fuzzy state assessment,the membership degree of each working condition is determined in real time,achieving accurate state assessment and avoiding the uncertainty of parameter threshold setting.Finally,based on the actual requirements of state maintenance,this paper establishes a risk assessment model for induced draft fan in power station using equipment state assessment results,and designs the surge risk assessment indicators in terms of comprehensive operation status and safety performance.By formulating a short-term risk assessment plan that considers the operational state,risk assessment results are linked with the maintenance strategy,achieving the leap from equipment operation status and operation risk assessment to optimal maintenance,enhancing the practicability of risk analysis management,and providing auxiliary decision-making for maintenance planning.