Study On Internal Combustion Engine Fault Diagnosis And GUI Design Based On Time Series Analysis

Engine operating state is crucial to the normal operation of mechanical systems. With the development of science and technology, the complexity of mechanical equipment is improved gradually. The diagnosis of equipment fault attracks more and more attention, and it has become an important technology branch of industry area. The technology is a subject with which the equipment abnormalies are discovered, the reasons of these abnormalies are analyzed, and the improvement measures are put forward through understanding and mastering the running status of equipment. Industrial production and running of enterprises are both inseparable from the mechanical fault diagnosis whose importance is self-evident. The establishment of the state monitoring and fault diagnosis system for the performance of internal combustion engines helps to improve the system reliability and maintainability. In addition, it facilitates timely maintenance, costs reduction, and it helps to extend the the mechanical equipment life as well as improve production efficiency.Time series analysis is a method through which the system dynamic performance can be directly acquired. The system dynamic model established by the sequence analysis method can describe the change or the movement of the system, which reveals the developing trend of the state performance and running condition, and forecasts its lifetime. However, the accuracy of the prediction model established by the time series analysis method is not high, which results to make the fitting performance too poor to reflect the actual situation of the system, especially in some mutation hop points, the model can not respond to the changes in time. Therefore, to improve the prediction accuracy of the model, the Kalman filter is introduced for time series analysis and processing, and the processing method is used to forecast the state of the inlet guide vane of the internal combustion engine. Finally, the residual analysis method is used based on the Kalman filter to realize the fault diagnosis. The related contents are as follows:(1) The modeling methods and modeling steps of time analyses are studied systematically including judgment of the time series stationarity, processing from nonstationarity to stationarity, identification of the time series model, the method to model, and the model parameter calculation. Then the advantages and disadvantages of timing analysis method are analyzed. The temperature, angle, pressure and other sensor data near the inlet guide vane of a new internal combustion engine NG60 in a company are acquired, and they are modeled and analyzed by the time series analysis method.(2) The principle of the Kalman filter and common techniques of Kalman filtering are introduced, then the state-space model is introduced for estimating the performance state of the internal combustion engine. The conversion between the state-space models and time series analysis methods are studed, and the state-space model is established on the basis of time analysis. Compared to time series analysis methods, the Kalman filter has higher accuracy, smaller error, and smaller computation complexity.(3) Engine fault diagnosis method of the internal combustion based on the ARIMA model is put forward. And then the residuals of Kalman filter to be used for the fault detection are analyzed, and the method is applied to the sensor data such as the temperature,angle, pressure near the inlet guide vane of internal combustion engine, to calculate the filter residuals and residuals fault threshold for fault detection.It is proved that the method has good application value.(4) The time analysis GUI of an internal combustion engine is designed, including the analysis of time series in frequency domain, and the correlations property, which facilitates a more comprehensive and direct monitoring of the performance of the internal combustion engine, and makes it easy to design the internal combustion engine for the technical staff, and access the sensor data of the internal combustion engine in time.