The Application Of The Fiber Warning System In The Protection Of The Natural Gas Pipeline

The fiber optic warning system to collect identifying local oscillator signal the vibration source vibration type fiber optic warning system is able to achieve the key along the gas pipeline sabotage anticipation, but also has important theoretical and practical significance. Traditional fiber optic warning system can predict and sabotage targeting, and positioning accuracy and speed increasing. But now, for the type of optical fiber technology acquired vibration source vibration warning signal is still difficult to achieve accurate judgments. If the fiber-optic warning system can not effectively distinguish between threatening and non-threatening vibration source vibration source can cause unnecessary alarm and repair the situation, an enormous waste of human and material resources. Therefore, how to identify vibration sources, it has become a warning fiber optic technology in the development of an important issue.In this paper, fiber optic warning system collected using a local oscillator signal amplitude vibration analysis, short-time average energy law, energy law probability distribution of the signal spectrum analysis and the fundamental characteristics of property cycle method, and then create the corresponding maximum energy model, energy information entropy model, the fundamental frequency cycle model and signal stability analysis model line model features four kinds of different types of vibration sources classification and recognition. This paper mainly completed the work in the following four areas:First, the theory proposes the application of the time-domain and frequency-domain analysis features characteristic analysis of the vibration source vibration signals acquired analysis. Among them, the time-domain analysis features include signal amplitude vibration analysis, short-time average energy law and energy probability distribution method; frequency domain analysis includes a vibration signal spectrum analysis and the fundamental frequency cycle method.Second, to build a fiber-based Sagnac interferometry experiment site warning systems, mechanical excavation, gas leak, dug and vehicle vibration source through four kinds of simulation experiments and obtain the corresponding vibration signal data. Firstly, to obtain vibration signals based on the principle of optical fiber Sagnac interferometer principle early warning system will be described, and the experimental gas pipeline fiber optic warning system model based on this principle. Build a fiber optic warning experimental site in follow-up study, conducted four kinds of vibration source simulation. Experimental data acquired domain and frequency domain analysis method for analysis and identification of use, provide data support for the construction of the local oscillator signal identification repository.Third, to build four kinds of characteristic properties of the model based on the time domain and frequency domain analysis of vibration source vibration signals. Local oscillator signal characteristic vibration properties model paper has built the largest energy model, energy information entropy model, baseband signal stability cycle model and model row. Among them, the largest energy model used to identify mining machinery vibration source; energy information entropy model used to identify gas leak vibration source; signal stability model to identify dug vibration source; fundamental frequency cycle model used to identify the vehicle passes the vibration source. It can be achieved by the above four models of local oscillator vibration signal analysis, processing, and identify the type of the local oscillator signal being processed.Fourth, build characteristic properties of the model and decision tree classifier Combination, Construction type vibration source identification and classification model for local oscillator vibration signal classification and recognition system. Vibration source type recognition classifier design objective is distributed fiber optic warning system to monitor the vibration source vibration signal system and phased recognition. Categorize the first phase of the experimental phase of the monitoring to a known type of vibration source vibration signal, as the establishment of training data sample set:Based on the decision tree classifier to identify key process vibration signals in two stages. The second phase, characterized by the properties of the model and the sample set than the right, right in the practical application monitored vibration signal classification and recognition.In this paper, by constructing fiber optic warning system based on the principle of Sagnac obtain natural gas pipeline along the local oscillator signal of vibration data through the signal time domain and frequency domain analysis of the characteristics of the vibration signal is processed and then combined by constructing characteristic attributes.