基于蒙特卡罗-压缩感知的瞬发脉冲中子测井信号自适应重构方法研究

Prompt neutron logging has become the core frontier method of uranium/ thorium, oil and gas, coal and other mineral exploration and mining. The signal acquisition and processing of prompt pulsed neutron is the basis of the analysis and processing of the subsequent data. This project is aimed at the short period of prompt pulsed neutron signal, the huge amount of information, the low sampling rate of hardware and the difficulty of signal reconstruction. Taking the prompt pulsed neutron logging signal as the research object, using Monte Carlo logging characteristics simulation and compressed sensing theory, by changing the pulsed neutron signal acquisition for sparse signal processing, a new method for pulsed neutron undersampling and high precision signal reconstruction based on Monte Carlo-Compression perception framework, that includes the under Nyquist acquisition, Monte Carlo observation matrix design and High precision dynamic reconfiguration. A software and hardware fusion framework system for pulsed neutron logging signal acquisition and processing based on compression perception theory. This project can provide more advanced and effective signal processing methods for solving the problem of logging high-precision data acquisition and real-time data transmission, improving the accuracy and efficiency of nuclear logging tool, breakthroughs in the method of ore finding and prospecting. The results of the project will lay a solid foundation for further improving the method theory of nuclear logging signal processing and the development of new logging system.

瞬发脉冲中子测井已成为铀/钍、油气、煤炭等矿产勘探与开采的核心前沿方法，而瞬发脉冲中子的信号采集与处理是后续数据分析处理之根本。本项目针对瞬发脉冲中子信号周期极短、信息量巨大，硬件采样率低、信号重构困难等问题。以瞬发脉冲中子测井信号为研究对象，运用蒙特卡罗测井特性仿真和压缩感知理论，通过变脉冲中子信号采集为稀疏信号处理，从欠Nyquist采集、蒙特卡罗观测矩阵设计、高精度动态重构3个方面开展基于蒙特卡罗-压缩感知框架的脉冲中子欠采样与高精度信号重构的新方法研究，建立基于压缩感知理论的脉冲中子测井信号采集与处理的软硬件融合处理框架体系。为有效解决测井高精度数据采集与实时数据传输难题，提升核测井仪器精度与效率，实现寻矿找矿方法上的突破，其成果将为完善核测井信号处理理论体系及研制新型测井系统奠定坚实基础。

本项目的研究课题来源于瞬发脉冲中子信号采集、处理及高分辨率重构等关键问题。四年来，课题组成员在深入分析脉冲中子测井信号的基础上，以自适应压缩感知框架为主线方法，通过横向比较与纵向研究相结合方法，开展了一系列改进完善工作，提出了若干个优化改进算法与模型，对中子测井信号克服硬件采集瓶颈、提升信号质量具有现实的意义与实际应用价值，也将对国内同行进一步的研究工作具有一定的参考作用。同时，也为项目组成员将来的研究工作打下了坚实的基础。