A Mobile Robot Based Monitoring Platform for Pipeline Leakage Diagnosis Based on Cross-correlation Analysis

Nowadays, various pipelines are broadly used in the transportation of resources in the production process and life applications, such as natural gas, compressed air, water, oil and so on. Pipe leakages usually occur due to the improper manufacturing process, installation of valves or other devices, prolonged use, etc. Pipeline leakage will cause energy loss, pollution and other severe problems, which will lead to equipment damage and even casualties. Therefore, the detection of pipeline leakage is significantly important. Traditional pipeline leak detection requires visual inspection or many fixed detection sensors. This may cause missed inspection by human errors and also will greatly increase the cost on the expensive devices. Compared with visual inspection and specific testing devices, mobile robots have the advantages of portability, economic cost and wide application for effective and efficient pipeline leakage detection. Robots can be deployed on lines or near equipment prone to pipeline leaks which can be detected online according to the path set. This paper presents an effective pipeline leak detection method to monitor the leakage of a two-stage reciprocating compressor with a mobile robot. An android mobile phone is installed on the mobile robot platform. After the robot moves to the monitoring point, the microphone of the mobile phone starts to collect environmental sound. The cross-correlation analysis is then performed on the LABVIEW software platform to implement online detection of different leakage faults. The experimental results show that the extreme value of the correlation number is effective and efficient for real-time detection of the presence of leaks.