What does an expert system look like in transformer fault diagnosis?
For the transformer diagnosis system, expert diagnosis is not really about finding experts in real life to diagnose the transformer. This is definitely unrealistic. The expert system here is specifically designed to solve some faults that cannot be solved by functions or mathematical models. At this time, this expert system is needed to complete it. Simply put, it can provide reasonable supplements in situations where information is incomplete and uncertain. Therefore, this method is also the direction that domestic scholars and manufacturers have been researching. They have invested a lot of manpower and resources, so it has been widely applied in real life.
Now let's give a simple example of the practical application of this system: using the combination of online obtained oil and gas volume and partial discharge signals for fault diagnosis; Using fuzzy reasoning to construct an expert system for reproducing fault diagnosis of power transformers; The above are some simple applications in various directions to facilitate a better understanding of the concept of this system. However, there are still many shortcomings in transformer fault diagnosis. Even though we lack a lot of research results, we still need to truly recognize these issues, such as:
1. The lack of sufficient supplementation in terms of expertise and breadth of acquisition directly leads to the incompleteness of the knowledge reserve system, as it still requires technical proficiency to fully transplant professional knowledge to computers.
2. As we mentioned last time in fuzzy theory, its accuracy cannot be guaranteed in the face of some uncertain faults.
3. It can be considered a problem in the field of computer science, which is the difficulty in maintaining knowledge because the algorithms are too complex!
The above are the projects that oil immersed transformer manufacturers need to research and develop, hoping to become more advanced and provide better social value.
