Velocity prediction of a pipeline inspection gauge (PIG) based on differential pressure and acceleration with artificial neural networks
Pipeline Inspection Gauge (PIG), Artificial Neural Networks, Soft Sensors, Raspberry Pi.
In the oil and gas industry, a device known as the Pipeline Inspection Gauge (PIG) runs through oil and gas pipelines performing various maintenance operations. The efficiency of these operations can be increased by employing a closed loop speed control system. To get the speed, it is usually resorted to the use of odometers. Although such a method is relatively simple, it can cause certain measurement problems resulting from slippage between the odometer wheel and the duct. In order to contribute to the solution of these problems, the objective of this work is to develop a soft sensor (virtual sensor) to measure the velocity of PIGs from the pressure difference to which the device is submitted in the duct. A soft sensor is basically made up of two elements: a mathematical model of the system and sensors that measure the physical variables required by the model. To obtain the model, it is intended to use artificial neural networks. This model will be shipped in a Raspberry Pi to be installed in the PIG, which will also be responsible for obtaining the sensor data. The SIP testing pipeline from the Petroleum Assessment and Measurement Laboratory (LAMP / UFRN) will be used to evaluate the results. The proposed system is expected to be able to complement the use of odometers, thereby increasing the reliability of speed measurement.