Research On Pipeline Parameter Measurement Method Based On Double Coil Near-field Eddy Current Technology

The demand for oil can often reflect the economic situation of a country.In recent years,China's economic situation has been good,and the exploitation of oil has gradually increased.Oil casing plays an indispensable role in oil extraction.However,due to the complex composition of petroleum crude oil,the petroleum casing is inevitably corroded.In order to ensure the normal exploitation of oil,it is necessary to detect the degree of corrosion of the oil casing.Electromagnetic thickness gauge is a tool for measuring the damage of petroleum casing by electromagnetic signals.It consists of a near-field and a far-field connection.The near-field portion mainly measures the inner diameter and permeability of the casing,and the far-field portion combines the permeability measured in the near-field to measure the wall thickness of the casing.Determine the degree of damage to the casing.In this thesis,the near field part is mainly discussed.Near-field eddy current detection model is derived by Green's function method.And an axisymmetric near-field eddy current simulation model is established in the simulation software COMSOL.In this model,three frequency mixed signals are transmitted by the transmitting coil.After the signal is received by the receiving coil,the phase difference of each frequency is calculated by the all-phase FFT spectrum analysis method.On the hardware circuit,a three-frequency square wave generated by the CPLD is used in the transmitting circuit.It is then filtered into a sine wave and finally coupled directly to the transmitting coil.The design of dual coil receiving is adopted by the receiving circuit.The signals received by the two receiving coils are analog-to-digital converted,and the converted digital signals are sent to the DSP.In the DSP,the phase difference will be calculated by the all-phase FFT spectrum analysis method and then uploaded to the central control board.By changing the physical information of the casing in the simulation model,the phase differences of the three emission frequencies in different casings can be obtained.Based on this,a near-field inversion simulation library is established accordingly.After the phase difference of each frequency is measured in the actual measurement,the physical information of the casing can be calculated using an inversion algorithm.In the thesis,the prediction capabilities of the GA-BP neural network algorithm and GA-RBF neural network algorithm will be analyzed and compared from multiple aspects.Finally,GA-RBF neural network algorithm is used as the inversion algorithm of casing physical properties.Finally,it is planned to use an electromagnetic thickness gauge to measure the casing in the laboratory to verify the measurement capability of the near-field dual-coil measurement module and the predictive capability of the neural network algorithm.Then use the electric thickness gauge to measure the experimental wells and calibration wells to verify the instrument's ability to measure and repeat in complex environments.Finally,an on-site operating well is measured with an electromagnetic thickness gauge to verify the actual engineering capability of the instrument.