The Processing And Application Of Monitoring Data On Large Scale Equipment

Mobile power stations and pipelayers play an important role under oil and gas pipeline construction, monitoring is necessary due to the particularity of this kind of crawler engineering vehicles. Oil mass, oil pressure and device posture were monitored, which will be useful to know the real-time status of the devices remotely. As the data become more larger, it can be used to mining the operation rules of the device during the life cycle.The RTU(Remote Terminal Unit) of the engineering vehicles and the monitoring website combine a complete monitoring system, on the basis of which the data from RTU through the network is being processed, the whole process was explained in detail in the dissertation. The main contents of this dissertation:1)The research background, purpose and significance were expounded detailedly. The development status of device monitoring at home and abroad was researched in this dissertation, and the application status of monitoring technology applied on vehicles was researched and analyzed too. The status of data mining during the era of big data was researched. Data analysis, processing and mining conducted in the context of this issue can not only bring economic benefits, but also reflect some innovation.2)The method how to analyze data packets transmitted by RTU was described in this part, the non-physical data was translated to the real data via the polynomial function obtained through experiment. All the real data was stored in the corresponding data table of SQL Server database. And a network platform for monitoring was developed through the ASP.NET technology. Displaying monitoring parameters, managing device information, showing GPS location, alarming exception information and sending alarm messages were realized on the platform.3)Noise processing method such as smoothing means and the wavelet transform were used to analyze the data in the database. And appropriate denoising methods were chosen by comparing the results. Based on the data splicing rules, device parameters were screened and divided into different groups, parameters in each group was analyzed and denoised individually, including oil mass, oil pressure, rotate speed, water temperature, dip angle.4)The relation between oil consumption and time was acquired through the relation between oil mass and time. Neural network processing method was introduced to deal with data and build data model, the structure and network parameters of BP network were determined through trial and error. After building the network model about the oil consumption and other device parameters, the test sample were used to test the network and then the error analysis was given in the dissertation.