Study On The Warning And Assessment System Of Substation Equipment

Warning and assessment system of substation equipment can monitor the operating status of substation equipment online, give early warning signals of the equipment which state trend to serious deterioration, give its reason to the fault equipment. The establish of the system will promote the shift of substation equipment from regular maintenance to condition-based maintenance, improve the overall automation level of substation, so the research has important practical significance. Based on the research of preventive test methods and on-line monitoring technology of substation equipment, explore new methods of condition assessment and fault diagnosis of power transformer. Advanced applications designed to improve system.Study the preventive test methods and online monitoring technology of transformers,high voltage bushings, circuit breakers and surge arresters. Sorting out the indicators of the equipments and established a scientific and objective state information system to the substation equipment, and given the solutions of the realization of substation equipment online monitoring information integration.Proposed a new transformer condition assessment method which based on gray clustering and evidence synthesis. The hierarchical model of a transformer condition assessment is established. Transformer condition index system is built in this paper. Using gray clustering method to cluster single indexes to get the state of complex indexes. Rules of conflict evidence synthesis of D-S evidence theory is improved and the overall state of transformer is then got by using the clustering coefficient as a function of the basic probability assignment. Practical case study shows that the model established is effective.Proposed a new transformer fault diagnosis method based on the nuclear extreme learning machine, established a transformer fault classification model. On the basis of in-depth study of extreme learning machine and kernel function, combined them and use it to transformer fault diagnosis. Gives a method for obtaining eigenvectors based on the information of DGA. Transformer fault forms are classified in this paper. Select RBF as the kernel of KELM. Using One-Versus-Rest model act as the main of the learning machine.Large number of samples are collected, the superiority of the proposed method is verified by the simulation.The implementation of warning and assessment system hardware are shown. The terminal of the warning and assessment system are given. Showcases the architecture,operating environment, features of the warning and assessment system.