Data Processing And Fault Diagnosis Based On Analyzing The Evenness Of The Diesel Engine By The Fluctuations Of The Instantaneous Speed

With the rapid development of modern science and technology, machinery equipment come towards the direction of large-scale, complicated, comprehensive and high grade. So the machinery equipments are increasingly complicated. Diesel engine is one of the widely used machinery equipment. As a representative of reciprocating machine, their frame is very complicated, so, their faults are bound to be complexity and diversity. Once the fault happened in a certain part or aspect, it will cause the paralysis of the entire piece of the equipment and enormous economic losses and casualties. The traditional diagnostic methods of demolition-style and repairing after the application are not only time-consuming and labor-intensive, but may not be able to achieve the desired results. They can no longer meet the requirements of modern production. If we test the equipment during the normal operation of the device, find the potential fault factors, take measures as soon as possible and prevent the unexpected emergence, we will not only avoid catastrophic equipment faults, but also avoid the lack of maintenance and excess repair, it has significant economic benefits.The test during the operation of the device demands us to design an online fault diagnosis system, analyze the transmission data sent by the diesel engine on real-time. Using signal processing technology and artificial intelligence technology for fault analysis and diagnosis.Engine fault includes mechanical fault and performance fault, the later one is 90% of them while the former one is 10%. The diesel engine's run evenness is very critical in the experience diesel engine flow. So the study of the diesel engine's run evenness has important and practical significance.In response to these circumstances, the main job of the article is:1. Designing the human-machine interface of Diesel Engine Fault Diagnosis System. It mainly includes the log interface, the main interface, data acquisition interface and fault diagnosis interface. The design of the log interface realize the memory input using VB to call Access database, It can make the input now to the Access database automatically and saved them in the current interface automatically in the next text. Data acquisition interface complete the data communications, testing, receiving, and drawing the instantaneous speed waveform using the data collected after the treatment in order to fault diagnosis. The fault diagnosis interface completes the calculation of feature points, histogram display and fault diagnosis.2. Data-processing. The main study is the design of digital filters and the filtering treatment on the instantaneous speed of the original waveform. The digital filter includes FIR and IIR, to the fault diagnosis system, in VB programming, the article uses the window function to design the filter. It is relatively simple, easy to achieve and meet the requirement of the design. However, the design did not take into account other parameters such as cut-off frequency to the filter effect. In order to better observe the effect of various parameters on the filter, At the same time, compare the various filtering method, This article studies simultaneously the design of FIR and IIR digital filter for waveform processing in MATLAB. The design of FIR filter still uses the window function and analyze the effect of cut-off frequency to the filter. the design of the IIR uses the Butterworth and the Chebyshev function, This paper analyzes the design steps and the waveform after filtering.3. Feature Point Extraction by Wavelet Decomposition. The paper analyze the wave by wavelet toolbox, Including the signal extension, One-dimensional stationary wavelet transform, One-dimensional stationary wavelet denoising and using db1 wavelet decomposition to identify the most points.4. Fault Diagnosis. The article analyze the speed waveform in both time and frequency domain. From the time domain of the Instantaneous speed waveform, we Identify the feature points and do the error analysis, then make a histogram. Analyze it According to the fuzzy pattern recognition domain. From the amplitude-frequency characteristic map of the instantaneous speed, we can see the basic operating frequency and firing frequency as well as the non-dimensional factor calculated by them , we regarded them as parameters of diesel engine fault diagnosis...