A Research Of Diagnostic Technique Based On Noise Source In Complex System Of Mechanical Vibration

The good working condition of mechanical equipment is an important guarantee for safety. With its continuous emphasis to the extent of strengthen, the research of equipment fault diagnosis is increasingly enhanced. In recent years, there are a lot of new technologies of diagnostic equipment to show up, and the feature-based information integration technology for the field of expert systems has become an important research direction.This paper is based on the vibration radiation noise monitoring of mechanical equipment, and studies the fault diagnosis method based on vibration and noise characteristics of information. Due to the many characteristics of information in the diagnosis process can improve the accuracy of diagnostic results, this paper studies a variety of signal feature extraction method theoretically, including time domain analysis, amplitude domain analysis, frequency domain analysis and time-frequency analysis, and using the above methods to deal with a single-stage centrifugal pump vibration data, which further illustrates the advantages and disadvantages of each method. In addition, to meet the demand of intelligent diagnosis, this paper presents fault diagnosis method which can be characteristic of many information integration and combination of Fuzzy Inference System. The paper initially does researches on the integration of information technology methods and hierarchical structure, analyzes the advantages of fuzzy logic in the field of information integration; at the same time, introduces a wide range of fuzzy inference system; focuses on two aspects of Mamdani Fuzzy Inference System from theory and application; through specific steps to complete the system design; and combines with simulation to further explain the system in equipment fault diagnosis. Finally field tests in Songhua Lake is carried out, and the paper sets up the various membership function based on the characteristics of diesel engine when each feature changes; gives inference rule and chooses a suitable computing methods and synthesis of fuzzy rules to determine the diagnostic system of fuzzy output finally.Through the experimental data deal, signal processing methods can be used to extract equipment running in the course of many characteristics of information is verified; and, researchers can make the job a state of more comprehensive and concrete analysis based on many characteristics; at the same time, the fuzzy diagnosis system based on many characteristics of information integration does decision-making classification of faults accurately; at the absence of a priori knowledge, the system can also fulfill the simple diagnosis of the equipment.