The Dga-line Devices, Automatic Control And Signal Processing Algorithms

The analysis of the dissolved gas in oil of transformer is an important kind of technology to diagnose internal faulty of transformer. As the chromatographic analysis has the advantage of high level of separation, high speed of analysis, high detection sensitivity and minor dosage of sample, it is popular to the power sector at home and abroad. The research of on-line chromatographic detection to the dissolved gas in oil of transformer used as a complement and development of off-line detection can realize the timing,on-line and intelligent monitor and diagnosis of filed chromatographic analysis of the gas dissolved in oil, which has very important practical significance and value. This paper firstly analyzes the inexorable trend of developing on-line detecting of transformer, introduces the development of the on-line detector of dissolve gas in oil in and aboard and the main factories of on-line DGA equipment and their equipments at home. Qualitative analysis and quantitative analysis to the gas dissolved in oil is an effective means to assess the work status of the transformer. Extracting membrane used to extract the oil and gas is the source of the sample gas and a first important part to realize the on-line detecting technology. Gas chromatograph is the core of the DGA technology. A detailed analysis about the components and key parts of the chromatograph is made. Right choice of the key parts is important to make the chromatograph work reliably. Auto injection is achieved by vacuum pumping to make pressure difference which can push the sample gas into the loop. The system is achieved intelligent operation by PC 104 control board. And the main processing unit use the Linux operating system. The process and method to realize data acquisition and system control based on multithreading technology are introduced. Chromatographic column is used to separate the mixed gas, and the zirconia oxygen sensor is used to make quantitative analysis to the fuel gas which can be analyzed indirectly by experiments. A logarithm-response area method is brought up according to the Nernst equation for quantitative analysis in this paper. A discussion about how the temperature influences the chromatographic column and gas detector is made. And in order to make the chromatographic column and zirconia oxygen sensor detector work at fitting temperature circumstance, a double temperature control system,which is analyzed from both hardware and software, base on intelligent PID arithmetic is designed. A research about smooth filtering method and automatic peak searching arithmetic based on increasing tagent search are made to process the chromatographic signal automatically. At last the source of error is analyzed, and the accuracy of the system is tested by contrasting the on-line and off-line detection. The result shows that the DGA system we designed can provide real time and reliable data source to the status analysis of the transformer.