Research On Pipeline Leak Detection Technology Based On Intelligent Algorithm

Given the actual status of oil pipeline in our country, the paper mainly researches on leak detection method based on negative pressure wave. The artificial immune algorithm from intelligent algorithms is introduced to improve on the traditional method based on negative pressure wave so that the localization precision is increased. The detail work is as follows:Firstly, the wavelet transform is used to get the featured inflection point of negative pressure wave, the wavelet transform are utilized to deal with filtering of negative pressure wave, which also can detect the signal edge under a strong noise background effectively, and the thesis puts forward several methods to eliminate the misinformation in leak detection based on negative pressure wave.Secondly, Using density regulating mechanism, individual diversity retaining strategy, and immune memory function, a kind of adaptive immune algorithm is proposed. The algorithm takes advantage of the adaptive selection strategy that is based on density and fitness. At the beginning of the algorithm, selection probability mainly bases on the selection probability of antibody density, so it is good for keeping diversity of antibodies, and the searching space can be enlarged. With the increasing number of iteration, selection probability gradually bases on fitness. In this way the speed of convergence can be accelerated, so that the speed of the algorithm can be accelerated. It shows that the adaptive immune algorithm is superior to both genetic algorithm and normal immune algorithm at solving optimal problems.Finally, by analyzing the velocity change of negative pressure wave which is generated because of pipeline leak during the process of transmission, the formula of wave velocity changing with temperature is deduced. Also the problem of the pipeline leak positioning is transformed into function optimization problem. By using the three above algorithms as well as the regular method in the localization problem of leak point and calculating the actual example, the results show that this method put forward in this thesis is more effective than its counterparts.