| Transformers,as the core equipment of power transmission,and distribution,the accurate assessment of its status and timely warning of failure risks are closely related to the safety of the power grid.In order to overcome the shortcomings of traditional state assessment methods,this paper adds meteorological factors to build new feature quantities based on Dissolved Gas Analysis(DGA).Based on the intelligent algorithm Extreme Learning Machine(ELM),this paper builds a transformer fault risk early warning model based on meteorological and condition detection data.Firstly,this article introduces the types of faults and their mechanisms,and explains the relationship between meteorological,condition detection characteristic quantities and transformer faults theoretically,and then selects meteorological factors:humidity,ambient temperature,precipitation,solar radiation,natural wind speed and state detection: transformer oil temperature,dissolved gas,and transformer load rate to establish a transformer fault diagnosis and risk warning index system.Then,MATLAB is used as a simulation platform,and a transformer fault diagnosis and risk early warning model is established based on the ELM algorithm.Finally,the actual historical data is used as input,and the model is substituted for simulation analysis.The performance difference between DGA and the model in this paper is compared,and the accuracy of ELM algorithm is verified by comparing with other intelligent algorithms.The simulation results show that compared with the traditional DGA method,the diagnosis accuracy of this method is increased by 10%,and the diagnosis and prediction results accuracy of the model are higher than other intelligent algorithms.It is finally verified that the method proposed in this paper can effectively warn transformer failure risks. |
