Research On Improvement Of Condition Monitoring And Fault Diagnosis For Electrical Equipment In Power Plant

It is crucial for the development of equipment fault diagnosis technology for daily safe production in power plants, because it not only concerns to the arrangement of daily production schedule and equipment maintenance, but also find potential trouble in equipment and clear it in time, thus avoiding sudden failure in important equipment which would affect the safety and stability of the whole power grid. The premise of fault diagnosis is monitoring and analysis of quantity of equipment state, while have been widely applied in the present monitoring system. For example, DCS system has been utilized in new units and built units with more than 100 mW. Although the DCS system is very powerful in monitoring, it has some disadvantages in the aspect of data storage and analysis. To solve this problem, some improvements have been made by some power plants. For the data storage improvement, own self-developed database have been used in some power plants, while mature commercial database products (such as PI system) have also been used in other power plants for data storage of equipment state quantity every year. For the data analysis improvement, many power plants rely on making forms, curves, graphs artificially, which methods are slow and inefficient, and cannot be consecutive. Although it is difficult to analysis the huge amount of storage data in power plants, in recent years, the new developed data mining method combining with the machines can solve this issue perfectly.In this paper, we firstly presented the current situation of equipment condition monitoring and fault diagnosis, the condition monitoring and fault diagnosis technology for various equipment, and their future development. Secondly, we did some analysis on various state monitoring methods for high voltage circuit breaker mechanically and electrically, and discussed the fault diagnosis technology based on fuzzy theory. The fuzzy diagnosis is to build a space of all signs that may appear in a certain fault, and build a space of all reasons that accounts for a certain fault, which corresponds to the signs in the space. Generally, the degree is described as membership, from which the diagnosis results can be obtained.High voltage circuit breaker fault includes mechanical and electrical fault, and most researches on the condition monitoring and fault diagnosis of the high voltage circuit breaker are focused on the mechanical vibration signals, while other researches are focused on the electrical part. In this thesis, the author will do comprehensive analysis on the two aspects. Mechanical vibration signals of high voltage circuit breaker will be analyzed by using wavelet decomposition and modulus maximum value. The wavelet decomposition is used to detect the signal breaking point, which is due to that the coefficient after wavelet transformation has a modulus maximum value when there is signal saltation. Therefore, the time of the fault happens can be determined, and the modulus value is considered to be characteristic parameter of fault diagnosis. Besides, the vibration signal amplitude and the period of vibration can be extracted as a characteristic parameter for fault diagnosis. Meanwhile, criterion for electrical life of high voltage circuit breaker has also been studied. According to the extracted mechanical and electrical characteristic parameters, associated or not associated single parameter performance indicator will be evaluated, which can turn to comparable and quantifiable amount after normalization. Therefore, the research on condition monitoring and diagnosis of high voltage circuit breaker can be realized by using synthetic judgement method to comprehensively evaluate service life of high voltage circuit breaker.