Waterwheel chamber fault diagnosis method based on knowledge graph

Abstract

The invention discloses a waterwheel chamber fault diagnosis method based on a knowledge graph, and relates to the technical field of hydraulic turbine set fault diagnosis. According to the method, awaterwheel chamber fault knowledge graph is constructed; the difficulty of establishing a knowledge base of an unstructured text in the past can be solved, the Bayes network is adopted to independently construct the network by learning historical data, network parameters are learned, uncertainty knowledge can be expressed and reasoned, the reasoning conclusion is accurate, and the application effect of waterwheel room fault diagnosis in engineering practice is greatly improved; according to the method, hidden dangers and problems in operation can be effectively reflected; judgment and early warning are carried out on the degradation trend of operation; according to the method, the abnormal change of the top cover water level can be warned and sensed in advance fundamentally, when the sensor breaks down or is abnormal, the fault can be diagnosed, a large number of uncertain factors of the waterwheel chamber water level change can be accurately described, consistent and coherent reasoning is achieved, the process is simple, and the diagnosis accuracy is extremely high.

Description

technical field [0001] The invention relates to the technical field of fault diagnosis of water turbine units, in particular to a fault diagnosis method for waterwheel chambers based on knowledge graphs. Background technique [0002] During the operation of the hydroelectric generating set, the safe operation of the waterwheel room is very important. The biggest threat to it is the accident of flooding the waterwheel room. Shutdown treatment will also cause the oil in the oil tank to overflow to the top cover and then be discharged to the tail water, causing environmental protection incidents and negative social impacts. The causes of such accidents mainly include the detachment of the top cover drainage pipe, the reduction of the efficiency of the top cover pump, and the failure of the vacuum breaker valve. It is necessary to diagnose the failure of the waterwheel chamber. [0003] At present, the fault diagnosis method of most hydropower plants for the waterwheel room is ...

AI Technical Summary

Problems solved by technology

[0004] 1) It is impossible to fundamentally warn in advance and sense the abnormal change of the water level of the top cover. When the sensor fails or is abnormal, the fault cannot be diagnosed

[0005] 2) The obtained information is relatively limited, which cannot effectively reflect the hidden dangers and problems in operation, let alone judge and warn the deterioration trend of operation

[0006] 3) The water level in the waterwheel room rises abnormally, and there are many influencing factors. The fault diagnosis needs to judge and analyze the cause of the abnormality from a large amount of collected data, but there are a lot of uncertain information in reality

[0007] 4) The traditional fault diagnosis modeling theory and method have the limitation of accurately describing a large number of uncertain factors in the water level change of the waterwheel chamber, lack of consistent and coherent reasoning, and the actual diagnosis results are far from the actual situation

Images