【作者】 范天姣； 朱凯光； 景春阳； 彭聪； 杨洋；

【Author】 Fan Tianjiao;Zhu Kaiguang;Jing Chunyang;Peng Cong;Yang Yang;The College of Instrumentation and Electrical Engineering,Jilin University;Key Laboratory of Geo-Exploration Instrumentation,Ministry of Education,Jilin University;

【机构】 吉林大学仪器科学与电气工程学院； 地球信息探测仪器教育部重点实验室；

【摘要】 系统辨识方法是以伪随机编码为发射源,从瞬变电磁数据中提取大地脉冲响应的一种主要方法。该方法同时记录高频率发射电流信号和全时航空电磁数据,然后对源信号和场信号进行相关运算,以此获取高分辨率的浅层地质信息。航空电磁探测首次应用伪随机编码作为发射源,进行大面积浅层勘探。常规的系统辨识方法会存在一定的系统误差,本文通过去除该误差对系统辨识进行优化。系统辨识方法主要基于维纳霍夫方程,构建伪随机编码源,航空电磁数据与大地脉冲响应的数学关系。航空电磁数据在数值计算上为伪随机编码源与大地脉冲的卷积,对该等式两边分别进行相关运算,将航空电磁数据与伪随机编码源的互相关结果与伪随机编码源的自相关结果矩阵化后,通过峰值匹配可辨识得到与单周期伪随机编码源长度相同的大地脉冲响应。但辨识得到的大地脉冲响应长度远远小于实际的大地脉冲响应长度,则系统误差的主要来源为未参与系统辨识的大地脉冲响应。该系统误差随时间序列增长,无噪声的情况下辨识结果与理论大地脉冲响应的相对误差最大可达到100%。定义系统辨识方法可辨识得到的大地脉冲响应为辨识值,没有参与辨识的部分为系统误差。因大地脉冲响应是呈e指数下降,可将辨识值与其时间序列分别取对数,通过最小二乘法对该数据进行一阶拟合。拟合函数延长一倍时间长度,计算出延长后的大地脉冲响应可视为系统误差,并通过卷积的方法从航空电磁数据中去除。通过数值计算,首次去除系统误差,可将辨识结果与理论大地脉冲响应相对误差的最大值下降至40%。本文通过对基于维纳霍夫方程的系统辨识方法进行了讨论,推导了系统辨识方法在矩阵求解中存在的系统误差,并通过最小二乘的方法减小了该系统误差。通过对河北张北伪随机源电磁探测试验数据进行了处理,通过实际案例证明了系统辨识优化方法的可靠性与优势。更多还原

【Abstract】 The system identification method is a main method to extract the earth impulse response from the transient electromagnetic data using pseudo-random code as the emission source.This method records high frequency emission current signal and full-time airborne electromagnetic data at the same time,and then performs correlation operation on source signal and field signal to obtain high-resolution shallow geological information.For the first time,pseudo-random code is used as the emission source in airborne electromagnetic exploration to conduct large-area shallow exploration.There is a certain system error in the conventional system identification method.This paper optimizes the system identification by removing the error.The system identification method is mainly based on Wiener Hoff equation to construct the mathematical relationship between pseudo-random code source,airborne electromagnetic data and ground impulse response.Airborne electromagnetic data is convolution of pseudo-random code source and ground pulse in numerical calculation.The correlation operation is performed on both sides of the equation.After the cross-correlation result of airborne electromagnetic data and pseudo-random code source and the autocorrelation result of pseudo-random code source are matrixed,the ground pulse response with the same length as single period pseudo-randor code source can be identified by peak matching.But the length of the identified earth impulse response is far less than the actual earth impulse response,so the main source of the system error is the earth impulse response not involved in the system identification.The system error increases with the time series,and the relative error between the identification result and the theoretical impulse response can reach 100% without noise.The earth impulse response identified by the system identification method is defined as the identification value,and the part not involved in the identification is the system error.Because the earth impulse response is e-exponential decline,the identification value and its time series can be logarithm,and the data can be fitted by the least square method.When the fitting function is extended by one time,the extended impulse response can be regarded as a systematic error,which is removed from the airborne electromagnetic data by convolution.Through numerical calculation,the maximum relative error between the identification result and the theoretical ground impulse response can be reduced to40% by removing the system error for the first time.In this paper,the system identification method based on Wiener Hoff equation is discussed,and the system error of the system identification method in matrix solution is deduced,and the system error is reduced by the least square method.By processing the electromagnetic detection test data of Hebei Zhangbei pseudo-random source,the reliability and advantages of the system identification optimization method are proved through a practical case.更多还原