The Research Of Transformer Fault Diagnosis Expert System Based On Case-based Reasoning

Artificial Intelligence has people’s widespread concern and is used in all aspects of lifesince birth. With the deepening research of Artificial Intelligence, Expert System becomes themost important branch, and continuous development as time goes. Artificial Intelligencethoughts emerge from out of the Dartmouth Conference on Artificial Intelligence and developrapidly with an immeasurable scale. Artificial Intelligence is designed to allow computers tohandle and solve the specialized and complex problems on all levels as same as the humanbrain. These also contribute to the emergence of Expert System in Artificial Intelligence. TheExpert System allows computers to solve problems like humans, which is the most importantand prominent in Artificial Intelligence in recent years.Since the appearance of transformer, it is difficult to ensure the long-term steady state.The cause of this problem is multifaceted. First of all, transformer has a complex structure asan important part of power system and real-time monitoring of transformer is difficult toreach. Secondly, factors causing the problem is complex, have not too precise logicalrelationship with time and environment, which increases the difficulty of solving problems.Thirdly, on the current power system level, if the transformer appears abnormal conditions, itis very possible that the entire power system fails, and even security incidents. So far,transformer equipment in most power systems still rely on manual sampling, regularmaintenance or solving after problem appearance. In such a way that is susceptible tointerference and influence of the external environment, and also vulnerable to be affected bysubjective factors of human experts, resulting in that problems are not adequately addressed.Solution of fault diagnosis of the modern transformer requires us to not deal with failure in aperiodic review or temporary repairs as the traditional way. We need to achieve all aspects ofmonitoring the state of the transformer fault, ensure that the transformer stable and efficientwork.We combine transformer fault diagnosis with Expert System, according to theprofessional knowledge and valuable experience to build a case-based reasoning mechanismsof transformer fault diagnosis Expert System, which we can deal with transformer faultdiagnosis and the process as past experience, which makes transformer fault handling will be more highly efficient and practical.This paper describes the characteristics and logical reasoning of Expert System based oncase, using SSH2frame structures technology. There is design and implement of atransformer fault diagnosis Expert System for case-based reasoning. The main contentsinclude the following aspects:1. Describes research background, situation, content and meaning of a transformer faultdiagnosis combining with Expert System;2. Introduces the background, stage of development and classification features ofExpert System and describes its advantages in transformer fault diagnosis;3. Describes the main inference principles of case-based reasoning Expert System, anddesigns Expert System inference module algorithm according to the theory;4. Introduces system design database design and functional design of the transformerfault diagnosis Expert System. The Expert System developed under the framework of SSH,SSH framework can greatly improve the development efficiency of the system.In addition, many technical aspects of the study in this paper is not very mature, we stillneed further improvement and perfection in the future study and research.