Research On Geomagnetic Signal Acquisition And Calibration Based On TMR Sensors

Since the discovery of the role of the geomagnetic field,it has been used in navigation,geological exploration,earthquake forecasting,sensors,medical treatment and many other fields,and it has become closely connected with human life.Among them,underwater geomagnetic detection has become an important detection means for mobile area measurement and search due to its advantages such as high reliability and good concealment.Although there are many high performance magnetic measurement solutions,there are some drawbacks in this application,such as the need for extremely low operating temperature,high power consumption,hysteresis and offset,etc.In this paper,geomagnetic signal acquisition and calibration based on Tunneling Magnetoresistive(TMR)sensors are carried out to address these drawbacks of existing schemes,and the main research is as follows:Designing a geomagnetic acquisition circuit based on TMR sensors and analyzing the sensor error sources.Starting from the magnetic measurement principle,this paper selects the TMR sensor with small size,high sensitivity and low power consumption as a magnetic detection means,and then designs a geomagnetic signal acquisition circuit based on the TMR sensor and examines the resolution and power consumption of the circuit.According to the composition structure and operating environment of the three-axis TMR sensor,the acquisition error is divided into modelable and non-modelable errors.For the unmodelable errors,near-field magnetic field simulation is used for design optimization.Since the unmodelable error comes from the high-frequency magnetic field disturbance in the digital part of the circuit,this paper performs near-field magnetic field simulation of the acquisition circuit based on Ansys,and uses it to optimize the circuit design,reduce the radiated magnetic field strength,and move the area of the circuit with strong electromagnetic radiation away from the sensor location,which greatly reduces the impact of high-frequency magnetic field disturbance on the geomagnetic signal acquired by the sensor.For the modelable errors,they are mathematically modeled,and the singular value decomposition traceless Kalman filter(SVD-UKF)algorithm is selected by comparison to compensate for the correction of such errors.The singular value decomposition traceless Kalman compensation system is then optimized according to the physical context of the magnetic sensor calibration by selecting the geomagnetic scalar as the observation,optimizing the steps for solving the attitude matrix,and selecting the nine parameters in the matrix to be measured as the observation vectors.The first traceless transformation is also optimized to significantly reduce the workload of selecting the set of sampling points and updating the solution in each iteration,reducing the computational effort of the system and improving the accuracy at the theoretical level.To verify the computational accuracy of this compensation system,the classical optimal ellipsoidal and least-squares combination compensation schemes in the domain are selected for comparison.It is proved by simulation that the accuracy of SVDUKF algorithm is better than the optimal ellipsoidal algorithm.Finally,experiments are designed for physical verification of the SVD-UKF calibration scheme.By using a cesium light pump magnetometer to measure the magnetic field data at a point as the standard value,and then the TMR geomagnetic acquisition circuit to the same point for real data acquisition.The data is fed into the singular value decomposition traceless Kalman system compensation to obtain the measurement to be made,from which the vector and scalar quantities after geomagnetic compensation are calculated.The error is controlled to within 0.47%after the algorithmic system compensation.