Research On Adaptive Amplification Based On Peak Envelope Of Magnetic Resonance Signal And Lattice Filter Algorithm

At present,about 40 percent of the population in China are facing the problem of water shortage,especially in the northern arid,semi-arid areas and karst areas of the southwest of China.In order to solve the shortage of fresh water resources in these areas,it is necessary to rationally develop and utilize underground fresh water resources.Magnetic resonance sounding(MRS)technology is a direct and effective way to detect underground water resources.Because of its high resolution,high efficiency,abundant information and uniqueness of interpretation,it has attracted great attention from hydrogeological researchers at home and abroad.Although MRS technology has many advantages as above,the saturation phenomenon often occurs in the magnetic resonance detection system under the noise interference of strong power line.The traditional solution is to artificially fix a appropriate magnification,based on several experiments to estimate the noise level,to prevent the saturation of the system.This method does not fundamentally solve this problem,because as time goes on,the characteristics of the noise will change.The fixed amplification factor can not guarantee that the system will not appear saturation.In view of the above problems,an adaptive amplification system based on the peak envelope of magnetic resonance signal is designed in this paper.The peak envelope circuit is used to track the trend of the magnetic resonance signal in real time.The coding circuit will compare and code the envelope amplitude,and the coding value determines the gain of the programmable amplifier AD526.In the controller FPGA module,the gain code and the coded value are converted,and all the order codes(magnification)are also stored.The program control amplifier adjusts the gain based on the gain code at the frequency of 50 kHz.In addition,since the gain isnot uniform,it is also necessary to normalize and interpolate the obtained raw data and amplification to obtain an effective magnetic resonance signal.The results of indoor and outdoor experiments show that the adaptive amplification system can neither detect saturation in the outdoor environment with 100 nV noise,but can also detect effective magnetic resonance signals and can perform field magnetic resonance water search experiments.On the other hand,although the adaptive amplification system can play a role in anti-saturation,the nanovolt level magnetic resonance signal is still very weak,compared with the harmonic noise.In order to obtain high-quality data,noise cancellation is also required.So,this paper proposes a reference-based adaptive lattice joint filtering algorithm.The modularized lattice structure in the filter allows the algorithm to have the ability to converge quickly,which can improve the signal-to-noise ratio(SNR)better and adapt to low signal-to-noise ratio signals.The simulation experiments show that the adaptive lattice joint filtering algorithm has stronger ability to eliminate power frequency noise,compared with the typical variable step size LMS algorithm.Especially when the SNR of original data is below-20 dB,the SNR of adaptive lattice joint filtering algorithm is about 2dB higher than that of variable step LMS algorithm.And the effectiveness of this algorithm is verified by the processing results of field measured data.