阅读说明：本技术 海洋电磁地层各向异性电阻率与发射源姿态联合反演方法 (Ocean electromagnetic formation anisotropic resistivity and emission source posture joint inversion method ) 是由 罗鸣 李予国 段双敏 李盼 吉芙蓉 吴俊良 于 2020-08-05 设计创作，主要内容包括：本发明提供一种海洋电磁地层各向异性电阻率与发射源姿态联合反演方法,包括：读取并转换参与反演的海洋可控源电磁场数据；设置联合反演执行参数；建立观测系统并设计联合反演初始模型；构建各向异性电阻率和发射源位置参数联合反演目标函数；求取电磁场关于各向异性电阻率和发射源位置参数雅各比矩阵和海森矩阵；基于反演参数特性自适应计算正则化因子；求取模型更新量；计算反演迭代模型的目标函数拟合差；判断是否满足反演要求；输出最终反演模型。该方法同时考虑海底介质电阻率各向异性和发射源的位置及姿态参数对反演结果的影响,能够有效解决发射源位置及姿态参数不准确给海洋可控源电磁资料解释带来的影响,适用性更广,容错性更强。(The invention provides a joint inversion method of anisotropic resistivity and emission source attitude of a marine electromagnetic stratum, which comprises the following steps: reading and converting ocean controllable source electromagnetic field data participating in inversion; setting a joint inversion execution parameter; establishing an observation system and designing a joint inversion initial model; constructing an anisotropic resistivity and emission source position parameter joint inversion target function; solving a jacobian matrix and a hessian matrix of parameters of the electromagnetic field, which are related to the anisotropic resistivity and the position of the transmitting source; adaptively calculating a regularization factor based on the inversion parameter characteristics; calculating the model updating quantity; calculating a target function fitting difference of the inversion iteration model; judging whether inversion requirements are met; and outputting the final inversion model. The method considers the influences of the resistivity anisotropy of the seabed medium and the position and attitude parameters of the emission source on the inversion result, can effectively solve the influence of the inaccurate position and attitude parameters of the emission source on the interpretation of the electromagnetic data of the marine controllable source, and has wider applicability and stronger fault tolerance.)

1. A joint inversion method for anisotropic resistivity and emission source attitude of a marine electromagnetic formation is characterized by mainly comprising the following steps:

s1, reading and converting ocean controllable source electromagnetic field data participating in inversion, wherein the data comprises real part and imaginary part data of the electromagnetic field, amplitude and phase data, and polarization ellipse major axis and minor axis parameters;

s2, setting joint inversion execution parameters, wherein the parameters comprise inversion maximum iteration times, target fitting difference, maximum iteration step length, step size proportionality coefficient, penalty function type and regularization attenuation coefficient;

s3, establishing an observation system and designing a joint inversion initial model;

s4, constructing an anisotropic resistivity and emission source attitude parameter joint inversion target function;

s5, solving a Jacobian matrix and a Hessian matrix of the electromagnetic field relative to anisotropic resistivity and emission source attitude parameters;

s6, adaptively calculating a regularization factor based on the inversion parameter characteristics;

s7, solving the anisotropic resistivity and the emission source attitude parameter updating quantity;

s8, calculating a target function fitting difference of the inversion iteration model;

s9, judging whether inversion requirements are met, if so, turning to S10, and if not, turning to S5;

and S10, outputting the final inversion model.

2. The method for joint inversion of anisotropic resistivity and emitter attitude of marine electromagnetic formations according to claim 1, wherein the joint inversion objective function of the parameters of anisotropic resistivity and emitter attitude is constructed by:

in the formula, phi is the target function of the inversion algorithm; m is a model inversion parameter vector including a sea floor anisotropic resistivity parameter m_ρA position parameter m of the emission source_PAnd the attitude parameter m of the emission source_TI.e. m ═ m_ρ+m_P+m_T；Is the gradient of the model parameter vector; i | · | | is a standard deviation operator; d is an observation data vector used for inversion; w_dWeighting the matrix for the data; w_mWeighting the model with a matrix; f (m) a forward response operator representing model m; mu.s_ρ、μ_pAnd mu_TRespectively a seabed anisotropic resistivity parameter m in an inversion model_ρA source position parameter m_PAnd an emission source position parameter m_TThe regularization factor of (1).

3. The method of joint inversion of anisotropic resistivity and emitter attitude of a marine electromagnetic formation according to claim 1, wherein the adaptive regularization factor is determined according to the following formula:

wherein i represents the ith iteration of inversion; mu.s_iIs a regularization factor; max | is the element with the maximum matrix absolute value; a is_mjIs a matrix product [ (W)_dJ)^T(W_dJ)]An element of (1); m is the matrix product [ (W)_dJ)^T(W_dJ)]Dimension (d); χ is the attenuation coefficient; lambda is the transverse resistivity rho of the submarine stratum of the inversion model_hResistivity p in the vertical direction_vWeighting factors for the emission source position parameters (x, y) and the emission source attitude parameters (Azm, Dip) are determined using the following formula

Wherein, alpha and beta are weight coefficients, m is an inversion model parameter, (x, y) are an abscissa and an ordinate of an emission source, and (Azm, Dip) are an azimuth angle and an inclination angle of the emission source.

4. The marine electromagnetic formation anisotropic resistivity and emitter attitude joint inversion method of claim 1, wherein the method for calculating the model parameter update amount based on the inversion parameters comprises:

wherein, Δ m is the model parameter updating amount of the next iteration, i is the ith inversion iteration, H_iIs Hessian matrix, g_iIs the gradient of the objective function.

5. The joint inversion method of sea floor anisotropic resistivity and source position according to claim 4, wherein the method of obtaining the minimum value of the objective function is:

where i is the ith iteration of inversion, J_iIs a jacobian matrix.

6. The joint inversion method of anisotropic resistivity at sea bottom and source position according to claim 1, wherein the jacobian matrix J is obtained_iThe method comprises the following steps:

wherein i is the ith inversion iteration; j. the design is a square_iA Jacobian matrix that is the forward response F (m); ρ ═ p (ρ)_h，ρ_v) For the anisotropic resistivity distribution of the formation, P ═ x, y is the emitter position parameter, and T ═ Azm, Dip is the emitter attitude parameter.

7. The joint inversion method of the sea floor anisotropic resistivity and the position of the transmitting source according to claim 1, wherein the method for obtaining the hessian matrix Hi comprises the following steps:

where i represents the number of inversion iterations i.

Technical Field

The invention relates to the technical field of marine geophysical joint inversion of multiple parameters, in particular to a joint inversion method of anisotropic resistivity and emission source posture of a marine electromagnetic stratum.

Background

Marine Controlled Source Electromagnetic Methods (CSEM) are a marine geophysical exploration method for exploring subsea hydrocarbon and mineral resources. With the increasing demands for resource development and deep exploration, the marine electromagnetic method is playing an increasingly important role in the investigation of mineral resources and the research of the structures of the marine crust and mantle. The frequency domain marine CSEM method generally uses a horizontal electric dipole source which is towed dozens of meters above the sea bottom as a transmitting source, transmits low-frequency electromagnetic signals with the range of 0.1-10Hz to an electromagnetic acquisition station in the towing process of a scientific investigation ship, and obtains the electrical distribution condition of the sea bottom medium by inverting the electromagnetic signals received by the sea bottom surface acquisition station. The strength and quality of the electromagnetic signals received by the electromagnetic acquisition station are not only related to the resistivity of the seabed medium, but also have direct relation to the relative position and attitude parameter precision of the emission source and the receiving station. In actual offshore operation, the observation ship is found to be influenced by factors such as sea motion, ship speed instability and the like in the process of dragging the emission source, so that the emission source cannot advance according to a preset path, the position and attitude parameters of the emission source can float in a large range, the variable quantity is difficult to monitor and record, the position and attitude parameters recorded by the emission source are deviated from the actual position and attitude parameters, and the processing and inversion explanation of marine electromagnetic data are greatly difficult due to inaccurate emission source position and attitude data. Therefore, the position and attitude parameters of the emission source need to be calibrated to improve the quality of the marine electromagnetic communication data.

Research shows that about 30% of oil and gas resources in the world exist in the electrically anisotropic stratum, the electrical anisotropy of the seabed medium is an important influence factor for obtaining the correct electrical distribution of the seabed medium, and if the resistivity anisotropy is neglected in explaining marine electromagnetic data, a reasonable seabed geoelectric model cannot be obtained. However, the inversion interpretation methods widely applied at present are still traditional geophysical inversion methods, which often assume that the seabed medium is electrically isotropic when inverting marine electromagnetic data acquired by marine resource exploration, and such inversion methods may provide inversion results with large errors for later data interpretation, which may also result in obtaining wrong interpretation results. For this reason, the properties of resistivity anisotropy of the seafloor medium must be considered when interpreting marine electromagnetic data.

In the existing inversion methods, the influence of the position and the attitude parameters of the electromagnetic emission source on the inversion result is rarely considered, most methods directly utilize the electromagnetic emission source parameters with errors to perform inversion, and the inversion result must also have errors. Therefore, if an inversion interpretation method for effectively correcting the position and the attitude parameters of the electromagnetic emission source can be provided, the accuracy of electromagnetic data interpretation can be effectively improved.

Disclosure of Invention

The invention aims to provide a joint inversion method for the anisotropic resistivity and the emitter attitude of a marine electromagnetic stratum, which can be applied to the accuracy of marine electromagnetic detection and inversion interpretation.

In order to achieve the above purpose, the invention provides the following technical scheme:

1. a joint inversion method for anisotropic resistivity and emission source attitude of a marine electromagnetic formation is characterized by mainly comprising the following steps:

s1, reading and converting ocean controllable source electromagnetic field data participating in inversion, wherein the data comprises real part and imaginary part data of the electromagnetic field, amplitude and phase data, and polarization ellipse major axis and minor axis parameters;

s2, setting joint inversion execution parameters, wherein the parameters comprise inversion maximum iteration times, target fitting difference, maximum iteration step length, step size proportionality coefficient, penalty function type and regularization attenuation coefficient;

s3, establishing an observation system and designing a joint inversion initial model;

s4, constructing an anisotropic resistivity and emission source attitude parameter joint inversion target function;

s5, solving a Jacobian matrix and a Hessian matrix of the electromagnetic field relative to anisotropic resistivity and emission source attitude parameters;

s6, adaptively calculating a regularization factor based on the inversion parameter characteristics;

s7, solving the anisotropic resistivity and the emission source attitude parameter updating quantity;

s8, calculating a target function fitting difference of the inversion iteration model;

s9, judging whether inversion requirements are met, if so, turning to S10, and if not, turning to S5;

and S10, outputting the final inversion model.

2. The method for joint inversion of anisotropic resistivity and emitter attitude of marine electromagnetic formations according to claim 1, wherein the joint inversion objective function of the parameters of anisotropic resistivity and emitter attitude is constructed by:

in the formula, phi is the target function of the inversion algorithm; m is a model inversion parameter vector including a sea floor anisotropic resistivity parameter m_ρA position parameter m of the emission source_PAnd the attitude parameter m of the emission source_TI.e. m ═ m_ρ+m_P+m_T；Is the gradient of the model parameter vector; i | · | | is a standard deviation operator; d is an observation data vector used for inversion; w_dWeighting the matrix for the data; w_mWeighting the model with a matrix; f (m) a forward response operator representing model m; mu.s_ρ、μ_pAnd mu_TRespectively a seabed anisotropic resistivity parameter m in an inversion model_ρA source position parameter m_PAnd an emission source position parameter m_TThe regularization factor of (1).

3. The method of joint inversion of anisotropic resistivity and emitter attitude of a marine electromagnetic formation according to claim 1, wherein the adaptive regularization factor is determined according to the following formula:

in the formula (I), the compound is shown in the specification,i represents the ith iteration of inversion; mu.s_iIs a regularization factor; max | is the element with the maximum matrix absolute value; a is_mjIs a matrix product [ (W)_dJ)^T(W_dJ)]An element of (1); m is the matrix product [ (W)_dJ)^T(W_dJ)]Dimension (d); χ is the attenuation coefficient; lambda is the transverse resistivity rho of the submarine stratum of the inversion model_hResistivity p in the vertical direction_vWeighting factors for the emission source position parameters (x, y) and the emission source attitude parameters (Azm, Dip) are determined using the following formula

Wherein, alpha and beta are weight coefficients, m is an inversion model parameter, (x, y) are an abscissa and an ordinate of an emission source, and (Azm, Dip) are an azimuth angle and an inclination angle of the emission source.

4. The marine electromagnetic formation anisotropic resistivity and emitter attitude joint inversion method of claim 1, wherein the method for calculating the model parameter update amount based on the inversion parameters comprises:

wherein, Δ m is the model parameter updating amount of the next iteration, i is the ith inversion iteration, H_iIs Hessian matrix, g_iIs the gradient of the objective function.

5. The joint inversion method of sea floor anisotropic resistivity and source position according to claim 4, wherein the method of obtaining the minimum value of the objective function is:

where i is the ith iteration of inversion, J_iIs a jacobian matrix.

6. The joint inversion method of anisotropic resistivity at sea floor and source location of claim 1, wherein jacobian is foundRatio matrix J_iThe method comprises the following steps:

wherein i is the ith inversion iteration; j. the design is a square_iA Jacobian matrix that is the forward response F (m); ρ ═ p (ρ)_h，ρ_v) For the anisotropic resistivity distribution of the formation, P ═ x, y is the emitter position parameter, and T ═ Azm, Dip is the emitter attitude parameter.

7. The joint inversion method of sea floor anisotropic resistivity and source position according to claim 1, wherein a hessian matrix H is obtained_iThe method comprises the following steps:

where i represents the number of inversion iterations i.

Compared with the prior art, the method provided by some embodiments of the invention has the beneficial effects that:

the invention mainly aims at the problem of seabed medium anisotropy widely existing in the detection of an ocean controllable source and the problem of low data quality caused by the deviation of position and attitude parameters recorded by an electromagnetic emission source of the ocean controllable source from actual position and attitude parameters. The inversion method not only provides an effective inversion method for the interpretation of complex seabed medium resistivity anisotropic electromagnetic data, but also provides a feasible technical means for the data processing problem under the condition that the position and the attitude parameters of the electromagnetic emission source have errors. Compared with the traditional inversion algorithm, the method simultaneously considers the influences of the resistivity anisotropy of the seabed medium and the position and attitude parameters of the emission source on the inversion result, combines the resistivity anisotropy of the seabed medium and the position and attitude parameters of the emission source as the inversion parameters to perform inversion simultaneously, and can effectively solve the influence of the inaccurate position and attitude parameters of the emission source on the interpretation of the electromagnetic data of the marine controllable source; the joint inversion method has wider applicability and can be suitable for inversion interpretation of resistivity anisotropy data; the fault tolerance is stronger, and the inversion problem when certain errors exist in the position and the attitude parameters of the electromagnetic emission source can be processed.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the embodiments or the prior art descriptions will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive exercise.

FIG. 1 is a block flow diagram of the process of the present invention;

FIG. 2 is a structural diagram of a one-dimensional resistivity anisotropy model;

FIG. 3 is a graph showing the variation of the target fitting difference of the inversion iteration model and the inversion iteration times;

FIG. 4 is a diagram illustrating the inversion of the emission source position parameters (x, y);

FIG. 5 is a diagram illustrating the inversion results of emission source attitude parameters (Azm, Dip);

FIG. 6 is a schematic diagram of inversion of the sea floor anisotropic resistivity.

Detailed Description

In order to make the technical problems, technical solutions and advantageous effects to be solved by the present invention more clearly apparent, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

The invention provides a combined inversion method of ocean electromagnetic stratum anisotropic resistivity and an emission source attitude, which considers the influence of the ocean bottom medium resistivity anisotropy and the position parameter of the emission source on an inversion result at the same time, combines the ocean bottom medium resistivity anisotropy and the position parameter of the emission source as an inversion parameter to perform inversion at the same time, and can effectively solve the problem of low quality of ocean controllable source electromagnetic data caused by inaccurate emission source position and attitude parameters.

Referring to fig. 1, a joint inversion method of anisotropic resistivity and emitter attitude of a marine electromagnetic formation mainly comprises the following steps:

s1, reading and converting ocean controllable source electromagnetic field data participating in inversion, wherein the data comprises real part data and imaginary part data of an electromagnetic field, amplitude and phase data, and polarization ellipse long axis and short axis parameters.

According to user requirements, converting the marine controllable source electromagnetic data files participating in the joint inversion into data files in a specific format, converting input electromagnetic field data into specific parameters (parameters of a real part and an imaginary part of an electromagnetic field, an amplitude and a phase, a long axis and a short axis of a polarization ellipse and the like) according to user requirements, completing reading of inversion data and starting the joint inversion.

S2, setting joint inversion execution parameters, wherein the parameters comprise inversion maximum iteration times, target fitting difference, maximum iteration step length, step size proportionality coefficient, penalty function type and regularization attenuation coefficient.

The joint inversion process is a complex and huge operation system, and a large number of parameters need to be applied in the process of realizing inversion iteration until a final joint inversion model is obtained. Different inversion execution parameters can obtain different inversion effects, and users can set the inversion execution parameters according to the characteristics of the joint inversion data and the specific requirements needed to be used. The inversion parameters mainly involved include: the method comprises the following steps of inversion maximum iteration times, target fitting difference, maximum iteration step length, step length proportionality coefficient, penalty function type, data components participating in inversion, regularized attenuation coefficient and the like.

S21, inversion maximum iteration times: maximum number of inversion iterations. The parameters are set, so that the joint inversion program can be exited after the joint inversion process reaches the maximum iteration number, continuous cycle operation of the inversion program is avoided, and the program calculation efficiency is improved.

S22, target fitting difference: and the target value of the fitting degree of the inversion model and the real model. The final goal of the joint inversion is to make the inversion model approach the real model continuously, and the poor data fitting is to determine the fitting standard of the inversion model and the real model. Setting a target fitting difference by a user according to the data quality, and exiting joint inversion when the fitting difference of the inversion model reaches or is lower than the target fitting difference; if the fitting difference of the inversion model does not reach the target fitting difference, the joint inversion procedure continues.

S23, maximum iteration step length: and setting the maximum value of the model parameter variation in the joint inversion process by a user. In the joint inversion, because the magnitude span of the marine controllable source electromagnetic field (or other parameters) is large, the marine controllable source electromagnetic field usually takes logarithms and then participates in the joint inversion operation. If the iterative model variation is too large, the variation of the result is very large, for example, the variation of the model parameter (after taking the logarithm) is 1.0, that is, the true value is increased by one order of magnitude, and such variation is usually unreasonable, so the parameter is designed to limit the variation of the model parameter.

S24, step length proportion coefficient: and setting a proportionality coefficient of the model parameter transformation quantity in the joint inversion process by a user. In the inversion iteration process, the variation degrees of different inversion parameters are different, and in order to enable the overall model variation to be in a reasonable range, the model parameter variation can be multiplied by a ratio coefficient (namely a step-size ratio coefficient) so as to ensure the steady convergence of the inversion model.

S25, penalty function type: the manner of constraints between the model parameters. In the joint inversion process, all model parameters participate in inversion at the same time, certain relation exists among different inversion parameters, restriction modes among the model parameters can be realized by selecting different penalty functions, and the model can be smoother or the curvature of the model can be increased.

S26, regularization attenuation coefficient: the attenuation coefficient of the regularization factor is adaptively selected. In the inversion iteration process, in order to ensure the ratio relationship between different parameters, the regularization factor needs to be changed correspondingly with the change of the inversion times, and the parameter is the attenuation coefficient of the regularization factor along with the iteration times.

And S3, setting parameters of a joint inversion initial model, wherein the parameters comprise a background layer resistivity parameter, a thickness parameter and an observation system parameter.

The inversion initial model is an initial value of a parameter of the joint inversion model, and usually needs to be set according to an observation system of input data, known water depth data and other data.

And S4, constructing an anisotropic resistivity, a transmitting source position and an attitude parameter joint inversion target function.

The joint inversion of the ocean controllable source electromagnetic seabed anisotropic resistivity, the position of an electromagnetic emission source and the attitude parameters relates to multi-attribute inversion parameters, and the relationship among the parameters is complex. In view of different resistivity anisotropy parameters of the submarine medium and differences between the position and the attitude parameters of the emission source, the joint inversion process of the electromagnetic submarine anisotropy resistivity and the position parameters of the electromagnetic emission source of the ocean controllable source is stabilized by utilizing regularization constraint. The objective function used was:

in the formula, phi is the target function of the inversion algorithm; m is a model inversion parameter vector including a sea floor anisotropic resistivity parameter m_ρA position parameter m of the emission source_PAnd the attitude parameter m of the emission source_TI.e. m = m_ρ+m_P+m_T；Is the gradient of the model parameter vector; i | · | | is a standard deviation operator; d is an observation data vector used for inversion; w_dWeighting the matrix for the data; w_mWeighting the model with a matrix; f (m) a forward response operator representing model m; mu.s_ρ、μ_pAnd mu_TRespectively a seabed anisotropic resistivity parameter m in an inversion model_ρAnd emittingSource position parameter m_PAnd an emission source position parameter m_TThe regularization factor of (1).

Inversion parameter m of seabed resistivity of inversion model_ρHas the following forms:

in the formula, ρ_hAnd ρ_vRespectively, the transverse resistivity and the vertical resistivity of the seabed stratum, and M is the layer number of the seabed stratum.

Source location parameter m_pHas the following forms:

m_P＝[l0g₁₀x₁…log₁₀x_t，log₁₀y₁…log₁₀y_t]^T；

attitude inversion parameter m of inclined electric dipole source_THas the following forms:

m_T＝[Azm₁…Azm_t，Dip₁…Dip_t]^T；

in the formula, (x, y) is a position parameter (abscissa, ordinate) of the emission source, (Azm, Dip) is an attitude parameter (azimuth, inclination) of the emission source, and t is the number of the electric dipole sources.

S5, solving a Jacobian matrix and a Hessian matrix of the electromagnetic field relative to the anisotropic resistivity and the attitude parameters of the emission source.

S51. Jacobian matrix J_iIs a matrix of partial derivatives of the electromagnetic field with respect to inversion parameters, different inversion parameters having different forms:

wherein i is the ith inversion iteration; j. the design is a square_iA Jacobian matrix that is the forward response F (m); ρ ═ p (ρ)_h，ρ_v) For the anisotropic resistivity distribution of the formation, P ═ x, y is the emitter position parameter, and T ═ Azm, Dip is the emitter attitude parameter.

If the resistivity tensor is expressedRho is the inverse model transverse resistivity rho_hAnd vertical resistivity ρ_vI.e. ρ ═ f (ρ)_h，ρ_v) Thus, the Jacobian matrix of the electromagnetic field with respect to the resistivity tensor ρ has the form:

similarly, the position inversion parameter m of the electromagnetic field relative to the inclined electric dipole source can be obtained_pAnd an attitude parameter m_TThe jacobian matrix of (a) is:

in the formula, (x, y) is a position parameter (abscissa, ordinate) of the emission source, (Azm, Dip) is an attitude parameter (azimuth, inclination) of the emission source, and t is the number of the electric dipole sources.

S52. Hessian matrix H_iFor the second derivative of the objective function to the inversion parameters, neglecting the second derivative terms and the asymmetry terms, Hessian matrix H_iCan be simplified as follows:

where i is the ith iteration of inversion, J_iA Jacobian matrix that is the forward response F (m); m is_ρ、m_PAnd m_TRespectively a seabed resistivity parameter, a position parameter and an attitude parameter of an emission source;

is the gradient of the model parameter vector; i | · | | is a standard deviation operator; d is an observation data vector used for inversion; w_dWeighting the matrix for the data; w_mWeighting the model with a matrix; mu.s_ρ、μ_pAnd mu_TAre respectively reversedSea-bottom anisotropy resistivity parameter m in model_ρA source position parameter m_PAnd an emission source position parameter m_TThe regularization factor of (1).

S6, adaptively calculating a regularization factor based on the inversion parameter characteristics.

The selection mode of the regularization factor mu is the key for solving the problem of the joint inversion of the ocean controllable source electromagnetic seabed anisotropic resistivity and the electromagnetic emission source parameter. In geophysical inversion methods, many regularization factor selection methods have emerged. When the ocean controllable source electromagnetic seabed anisotropic resistivity and electromagnetic emission source parameter joint inversion is carried out, different regularization factors are selected according to different inversion parameters. Firstly, selecting characteristic parameters related to a Jacobian matrix in the inversion process as a base number, and then regulating the regularization factors of different inversion parameters by combining the characteristics of the inversion parameters and the relationship among the parameters, thereby realizing the self-adaptive selection of the regularization factors.

Regularization factor mu_iThe following can be written:

wherein i represents the ith iteration of inversion; mu.s_iIs a regularization factor; max | is the element with the maximum matrix absolute value; a is_mjIs a matrix product [ (W)_dJ)^T(W_dJ)]An element of (1); m is the matrix product [ (W)_dJ)^T(W_dJ)]Dimension (d); χ is the attenuation coefficient; lambda is the inverse model transverse resistivity rho_hResistivity p in the vertical direction_vThe weighting factors for the emission source position parameters (x, y) and the emission source attitude parameters (Azm, Dip) are determined using the following equations:

wherein alpha and beta are weight coefficients, m is an inversion model parameter, and an initial value lambda of the weight vector₁Is a unit vector. When inverting the modeWhen the resistivity anisotropy exists in the type stratum, the regularization factor can be adjusted in a self-adaptive mode according to the resistivity anisotropy rate, so that the weight among elements in the target function is adjusted, and the effect of adjusting the inversion process in a self-adaptive mode is achieved.

And S7, obtaining the model updating quantity.

By finding the minimum of the objective function, the model update can be calculated, i.e. the gradient g of the objective function is calculated_i0, can be determined by the following formula

Where i is the ith iteration of inversion, J_iA Jacobian matrix that is the forward response F (m); phi is the target function of the inversion algorithm; m is_ρ、m_PAnd m_TRespectively a seabed resistivity parameter, a position parameter and an attitude parameter of an emission source; (ii) aIs the gradient of the model parameter vector; i | · | | is a standard deviation operator; d is an observation data vector used for inversion; w_dWeighting the matrix for the data; w_mWeighting the model with a matrix; f (m) a forward response operator representing model m; mu.s_ρ、μ_pAnd mu_TRespectively a seabed anisotropic resistivity parameter m in an inversion model_ρA source position parameter m_PAnd an emission source position parameter m_TThe regularization factor of (1).

The model parameter update quantity of the next iteration is calculated to be Δ m, which can be expressed as:

wherein, Δ m is the model parameter updating amount of the next iteration, i is the ith inversion iteration, H_iIs Hessian matrix, g_iIs the gradient of the objective function.

And S8, calculating the target function fitting difference of the inversion iteration model.

In the inversion iteration process, the inversion iteration model gradually converges to the real model, and the target function fitting difference psi gradually converges to 1.0. The objective function fitting difference ψ has the following form:

wherein i is the inversion iteration number, and d is an observation data vector used for inversion; f (m) a forward response operator representing model m; n is the number of inversion data; k is the number of the inversion data;_kis the standard deviation of the kth data.

S9, judging whether inversion requirements are met, if yes, turning to S9, and if not, turning to S5.

There are two criteria for determining whether to exit the joint inversion: 1) whether a joint inversion target fitting difference is met; 2) whether the inversion iteration number reaches the maximum iteration number or not. If not, turning to S5, and continuing to iteratively solve the model updating quantity; if so, the routine proceeds to S10.

And S10, outputting the final inversion model.

Referring to fig. 2, a schematic diagram of a one-dimensional typical earth electric model is shown. In order to verify the effectiveness of the joint inversion of the sea controllable source electromagnetic sea bottom anisotropic resistivity and the electromagnetic emission source parameters, the one-dimensional resistivity anisotropic model shown in fig. 2 is taken as an example, and the joint inversion of the sea bottom anisotropic resistivity and the emission source parameters is carried out by utilizing synthetic data. Assuming that 26 electromagnetic emission sources are arranged in the range of 500m-13000m of measuring line at equal intervals and are all positioned 50m right above the sea bottom, the postures of all the emission sources are the same, the azimuth angles are 90 degrees, and the inclination angles are 5 degrees; the 1 receiving station is arranged at the position of the sea bottom surface (0, 0, 1000). The emission current was set to 1 amp. The inversion data consists of the real part and the imaginary part of each component of the electromagnetic field, and each data is added with 2% of random Gaussian noise. In the inversion example, the initial model of inversion is air, seawater and uniform half space with resistivity of 1 Ω m, and the resistivity and depth of the air layer and the seawater are fixed. The position of the line of the electromagnetic emission source in the inversion initial model is set to be crossed with the position of the real line (an included angle of about 45 degrees exists).

Referring to fig. 3, a schematic diagram of a relationship between target fitting difference of an inversion iteration model and inversion iteration number is shown. It can be seen from the figure that, in the joint inversion process, the fitting difference of the target function is continuously reduced and gradually approaches to the real mode, and after 20 times of inversion iteration, the initial large fitting difference 17 finally converges to 3.1.

Referring to fig. 4, the result of inversion of the source location parameters is shown. As can be seen from fig. 4, the ordinate and the abscissa in the graph are respectively the position of (x, y), the hollow triangle is the initial position of the electromagnetic emission source, the solid triangle is the actual position of the electromagnetic emission source, and the solid circular dots are the position of the electromagnetic emission source obtained by joint inversion, so that even if the input position of the emission source is greatly different from the actual position (the position difference reaches 100m), relatively accurate position information can still be obtained by joint inversion (the distance between the position of the electromagnetic emission source obtained by inversion and the actual position does not exceed 25m), thereby indicating that the position of the emission source obtained by joint inversion is accurate.

Referring to FIG. 5, the inversion result of the emission source attitude parameters is shown. The horizontal coordinate is a y coordinate of the emission source, the left vertical coordinate is an azimuth angle of the electromagnetic emission source, the right vertical coordinate is an inclination angle of the electromagnetic emission source, the square and the hexagonal hollow line are initial values of the azimuth angle and the inclination angle of the emission source respectively, the square and the hexagonal solid line are the azimuth angle and the inclination angle of the electromagnetic emission source obtained by joint inversion, and the deviations of the azimuth angle and the inclination angle of the emission source obtained by the joint inversion and the actual values are respectively not more than 5 degrees and 1 degree, so that the posture parameters (the azimuth angle and the inclination angle) of the electromagnetic emission source obtained by the inversion are accurate;

referring to fig. 6, a diagram of inversion results of the sea floor anisotropic resistivity is shown, where light gray and dark gray line segments are real sea floor medium lateral and vertical resistivity distribution conditions, and light gray and dark gray step curves are sea floor medium lateral resistivity and vertical resistivity curves obtained by inversion, it can be seen from the diagram that the anisotropic resistivity of the jointly inverted surrounding rock is consistent with the real conditions, and the buried depth, thickness and vertical resistivity values of the high-resistivity layer are also accurately recovered. In conclusion, the combined inversion method of the ocean controllable source electromagnetic seabed anisotropic resistivity and the electromagnetic emission source parameters, which is provided by the invention, can invert the anisotropic resistivity information of the seabed medium and the position state parameters of the electromagnetic emission source. It is also shown that the algorithm proposed by the present invention is effective.

The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.