The Application Of Voiceprint Recognition Technology In Gas Pipeline Leak Monitoring RTU

Nowadays, with the continuous development and upgrading of China industrial, gas pipeline transportation is increasingly important in industrial. And it is including steam pipeline transportation, natural gas pipeline transportation, toxic gas pipeline transportation and so on. In industrial, security is undoubtedly the most important aspect. Because the gas pipeline usually transports explosive or toxic gas, if gas leakage occurs during transport, it should cause enterprises and national economic losses. Therefore, a reliable gas pipeline leak detection system for the safe transport of the gas pipeline is especially important. However, the pipe material, ambient temperature, internal pressure, transmission media, and other factors often result in that the leak signal is submerged by noise, and ultimately make the loss of leak detection accuracy.In order to achieve the gas pipeline leak detection under strong interference, this paper presents a leakage signal detection method based on voice identification technology. Firstly, based on empirical mode decomposition (EMD) theory, it achieves the separation of noise and leak signal, then gets the best reconstructing signal by adaptive reconstructing the dynamic signal. Finally, it makes the signal feature extraction and leakage detection with the best reconstructing signal to achieve the purpose. At the same time, we design a RTU system with Linux operating system in an embedded computer to achieve wireless network communication, GPS time acquisition and real-time data collection and storage. The RTU system makes real-time leakage detection for the gas pipeline, if it has detected a leak signal, it will send the leak signal and alarm signal to the monitor system at the first time. The real test signal reconstruction results and leakage alarm signal results show that the leakage signal detection method of this paper is feasible and reliable.