A method for extracting iron and polymetallic ore-induced anomalies based on the bridge-massart strategy of biorthogonal wavelet base

Abstract

The application discloses a kind of iron polymetallic ore ore-induced anomaly extraction methods under the double-orthogonal wavelet base Brige-Massart strategy, it is related to geological exploration technical field.The main content includes: using the orthogonal wavelet base of multi-resolution analysis structure, complete orthogonal wavelet transform, complete decomposition, prediction and update;For the high-frequency strong interference noise caused by volcanic rock, intrusive rock and near-surface uneven body in gravity and magnetic data, using double-orthogonal wavelet base Brige-Massart strategy eliminates interference noise and keeps the original information of the anomaly caused by iron ore body;Test the actual effect of different wavelet base and denoising method on the anomaly caused by iron polymetallic ore body of different depth.The application further improves the signal extraction efficiency and precision of the anomaly caused by iron mine, linearly shifts the original signal by linear phase filtering, improves the symmetry of orthogonal wavelet, so as to reduce the signal distortion caused by volcanic rock and other high-frequency geological bodies.

Description

Technical Field

[0001] This invention relates to the field of geological exploration technology, and more specifically to a method for extracting mineralized anomalies in iron polymetallic ores under the Brige-Massart strategy with biorthogonal wavelet bases. Background Technology

[0002] Wavelet analysis, a novel filtering technique developed in recent years, is a product of multidisciplinary collaboration and the culmination of research by mathematicians, physicists, and engineers. This technique is widely applied in research across numerous disciplines and related fields, including signal processing, image processing, pattern recognition, speech recognition, computer vision, aerospace technology, and geophysical exploration. As a novel time-frequency analysis method, wavelet analysis possesses multi-resolution analysis capabilities, representing local signal information in both the time and frequency domains. It can characterize transient components in normal signals and reveal their frequency components, thus earning the nickname "mathematical microscope" and playing a crucial role in time-frequency analysis. These characteristics make wavelet analysis an effective tool for geophysical numerical analysis.

[0003] Looking back at the development of wavelet analysis, its theoretical and applied developments are closely intertwined. Wavelet analysis originated from Fourier analysis, but Fourier analysis is only effective for stationary signals. To analyze non-stationary signals, numerous scholars continuously researched and ultimately discovered wavelet analysis. French geophysicist J. Morlet pioneered the use of wavelet methods to analyze and process seismic data and proposed the concept of wavelet analysis. Wavelet analysis began to be applied in the field of geophysics. Subsequently, J. Morlet and A. Grossman studied the continuous wavelet transform and its inverse transform, applying the theory to seismic wave signal analysis for oil exploration. Afterwards, a large number of scientists conducted in-depth research and analysis of wavelet theory. Among them, Y. Meyer and S. Mallat proposed the idea of ​​multi-resolution analysis, and S. Mallat also provided a numerical algorithm for discrete wavelets, namely the Mallat tower algorithm. Daubechies, starting from the discrete filter iterative method, constructed orthogonal wavelet bases with finite support and symmetric bioorthogonal wavelets, setting the basic framework for the construction of orthogonal wavelets.

[0004] However, traditional wavelet transform (first-generation wavelet) employs convolution operations, which are complex and computationally intensive, hindering real-time processing and hardware implementation. Starting in the mid-1990s, lifting wavelets emerged, a derivative of first-generation wavelets. Lifting wavelets do not require Fourier transform; instead, they utilize conjugate filters for decomposition and reconstruction, conforming to the properties of first-generation wavelet-generated wavelet bases. Sweldens conducted in-depth research on second-generation wavelets and their construction methods. This new wavelet construction method—the lifting method—is a novel approach for constructing quasi-orthogonal (semi-orthogonal) wavelets. This method does not rely on traditional Fourier transforms, providing a new and effective tool for wavelet analysis. However, while lifting wavelets simplify the computation process to some extent, further optimization is still needed in certain aspects. For example, ensuring effective denoising while preserving important signal characteristics remains a challenge when selecting appropriate wavelet bases and denoising methods. Furthermore, although lifting wavelets do not depend on traditional Fourier transforms, their integration and compatibility with other techniques must be considered in practical applications. Therefore, this invention proposes a method for extracting gravity and magnetic anomalies under the Brige-Massart strategy with biorthogonal wavelet bases. Summary of the Invention

[0005] In view of this, the present invention provides a method for extracting mineral-induced anomalies in iron polymetallic ores under the Brige-Massart strategy with biorthogonal wavelet basis, aiming to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A method for extracting mineral-induced anomalies in iron polymetallic ores using the Brige-Massart strategy with biorthogonal wavelet bases includes the following steps: Step 1: Construct orthogonal wavelet bases using multiresolution analysis; Step 2: Complete the orthogonal wavelet transform; Step 3: Complete decomposition, prediction, and update; Step 4: For interference anomalies or noise caused by superimposed volcanic rocks, shallow intrusive bodies and near-surface inhomogeneities in gravity and magnetic anomalies caused by iron polymetallic ore, the Brige-Massart strategy of biorthogonal wavelet basis is used to eliminate interference anomalies or noise, while maintaining the mineral anomalies caused by iron polymetallic ore bodies. Step 5: Examine the actual extraction effect of different wavelet bases and denoising methods on anomalies caused by iron polymetallic ore bodies at different depths.

[0007] Optionally, the specific content of step 1 is as follows: Clarify the concept of multi-resolution: Based on the stepwise partitioning of the function space, square-integrable functions... It is considered as a limit approximated stepwise; a low-pass smoothing function is used. Smoothing is performed to obtain an approximation for each level, while the smoothing function... The process of approximating stepwise is scaled up and down stepwise to form multiple resolutions; Constructing orthogonal wavelet bases: space Sequences of function spaces satisfying the conditions of monotonicity, approximation, scaling, and translation invariance constitute multiresolution analysis; closure approximation of all spaces. Furthermore, the functions in each space constitute a standardized orthogonal basis for the current space; From the orthogonal decomposition theory in functional spaces, we have:

[0008] in, It is the adjacent wavelet space; It is wavelet space; It is an orthogonal decomposition; It is an operation; according to: ,

[0009] get:

[0010] Where j is the scale; It is a wavelet space with scale j; It is composed of the straight sums of infinitely many orthogonal complementary spaces, ultimately yielding... The orthonormal basis of space is: ; in, It is a wavelet function; t is a variable; ; Biorthogonal wavelet basis performs linear shift on the original signal through linear phase filtering, which improves the symmetry of orthogonal wavelets and reduces signal distortion. For gravity and magnetic data of iron polymetallic mining areas, anomalies generated by high-frequency geological bodies of volcanic rocks and intrusive rocks overlap with ore body anomalies in the frequency band. Using a more symmetrical biorthogonal wavelet basis can reduce phase distortion during the filtering process, thereby distinguishing between mineral-induced anomalies and surrounding rock interference.

[0011] Optionally, step 2 includes the following: Based on the Mallat algorithm, any function Based on a resolution of 2 -N The low-frequency part of the time signal and The high-frequency components are completely reconstructed. Multiresolution analysis improves frequency resolution by continuously decomposing the low-frequency components. Multi-resolution analysis decomposes the low-frequency part of each layer of signal. The decomposition process has a specific relationship: S=A3+D3+D2+D1, where S represents the original signal, A and D represent the low-frequency and high-frequency parts respectively, and the numbers represent the decomposition levels. The original signal S contains mineral anomalies caused by iron polymetallic ore bodies or mineralization and interference caused by volcanic rocks. The two can be effectively separated through multi-scale decomposition.

[0012] Optionally, step 3 includes the following: Decomposition: Decompose the input signal S i Divided into two smaller subsets S i-1 and d i-1 The decomposition process is represented as F(S) i )=(S i-1 ,d i-1 ), where F(S) i () represents the decomposition process; In gravity and magnetic data processing closely related to iron polymetallic deposits, the decomposition process requires the rational selection of the number of decomposition layers based on the planar distribution scale and vertical burial depth of the ore body, so that the ore body signal is concentrated in the low-frequency subset, while the high-frequency interference of volcanic rocks enters the detail subset.

[0013] Optionally, the specific content of the prediction in step 3 is as follows: based on the correlation of the original data, use the even-numbered sequence S i-1 The predicted value P(S) i-1 To predict odd sequence d i-1 The filter P is applied to the even signal and used as the predicted value of the odd signal. The actual value of the odd signal is subtracted from the predicted value to obtain the residual signal. The decomposition and prediction process is repeated, and the original signal set is obtained after n steps.

[0014] Optionally, the specific content of the update in step 3 is: using the already calculated d i-1 For S i-1 The update is performed so that the global characteristics of the original signal set are reflected in the subset S. i-1 Continue to maintain this, that is, have Update S by constructing an operator U. i-1 The definition is as follows:

[0015] When the in-situ boost filter is reused, interleaved wavelet transform coefficients are obtained.

[0016] Optionally, step 5 involves: combining magnetotelluric sounding inversion profiles, seismic data, and drilling data, adjusting core parameter settings for typical iron polymetallic deposits to ensure that gravity and magnetic data processing achieves the desired effect; and verifying the accuracy of the method by comparing the processing results under different methods and parameter settings.

[0017] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a method for extracting mineral-induced anomalies in iron polymetallic ores under the Brige-Massart strategy with biorthogonal wavelet bases, the beneficial effects of which are: 1. Improve signal extraction accuracy: The orthogonal wavelet basis constructed by multi-resolution analysis can approximate the function step by step, thereby extracting the mineral-induced anomaly signal more accurately. The introduction of the biorthogonal wavelet basis further improves the accuracy of signal extraction. The original signal is linearly shifted by linear phase filtering, which improves the symmetry of the orthogonal wavelet and thus reduces signal distortion. 2. Enhanced anti-interference capability: The wavelet threshold determined by the Brige-Massart strategy is used for denoising, which effectively eliminates interference noise while maintaining the original information of the abnormal signal. Compared with traditional denoising methods, this method performs better in denoising, significantly improves the signal-to-noise ratio, and reduces reconstruction error. 3. Optimize computational efficiency: The wavelet transform is implemented using the lifting algorithm, which avoids the complex convolution operation in the Mallat algorithm, reduces computational complexity, reduces storage space requirements, and further improves computational efficiency. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0019] Figure 1 A schematic diagram illustrating the decomposition and reconstruction of the algorithm; Figure 2 This is a three-layer multiresolution analysis tree structure diagram; Figure 3 This is a diagram of a three-sphere model; Figure 4 This is the original anomaly curve; Figure 5 The curve after adding noise; Figure 6 To improve the noise reduction effect of Fast Fourier Transform; Figure 7 The noise reduction effect under three thresholds; Figure 8 To simultaneously identify volcanic structures and mineralized anomalies in the western Tianshan metallogenic belt; Figure 9 Map for remote sensing identification of volcanic structures in the western Tianshan metallogenic belt; Figure 10 To identify the Songhu iron mine and iron ore anomalies in the western Tianshan metallogenic belt. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Detailed explanation of wavelet basis selection criteria for volcanic rock interference: I. 1. Spectral characteristics of volcanic rock interference Volcanic rocks are common geological formations in polymetallic iron ore deposits, mainly including basalt, andesite, rhyolite, and tuff. These volcanic rocks exhibit significant characteristics in gravity and magnetic anomalies, and a correct understanding of their spectral features is crucial for selecting appropriate wavelet bases.

[0022] From a frequency perspective, volcanic rock interference is predominantly high-frequency, although some volcanic rock bodies may exhibit mid-to-high frequency characteristics. Volcanic rocks typically form in shallow geological environments, often in layered, lenticular, or stock formations, with a relatively small spatial distribution. Consequently, the resulting gravity and magnetic anomalies have shorter wavelengths, usually between several hundred meters and two kilometers. In contrast, the anomalies of deep ore bodies typically range from 1 to 3 kilometers in wavelength. While there are some differences in the frequency domain between the two, they often overlap.

[0023] Spatially, volcanic rock interference is characterized by shallow origin, localized distribution, and often sheet-like or banded patterns. When the volcanic caprock covers the ore body, the interference signal and the ore body signal are spatially superimposed perpendicularly. In areas with well-developed volcanic structures, the interference signal is often distributed in a ring or radial pattern. Volcanic rock interlayers exhibit layered characteristics and clear directionality. These different spatial distribution characteristics cause volcanic rock interference to show varying energy distribution patterns at different wavelet transform scales.

[0024] In terms of amplitude intensity, volcanic rock disturbances exhibit dramatic changes and large gradients. Volcanic rock masses often have high magnetic susceptibility and density, and show significant differences in physical properties from the surrounding rocks. Therefore, the resulting gravity and magnetic anomalies often have high amplitudes and rapid gradient changes, which are characterized by concentrated high-frequency component energy and large amplitudes in the wavelet coefficient domain.

[0025] From the perspective of their relationship with ore bodies, volcanic rock interference often coexists with or overlaps with polymetallic iron ore bodies. In volcanic rock-type polymetallic iron deposits, the ore body and volcanic rock mass share the same origin, and the anomalous signals highly overlap in frequency bands, making separation extremely difficult. In skarn-type iron ore areas, volcanic rock caprocks cover the ore body, and the interference signals and ore body signals are spatially vertically superimposed, requiring consideration of their positional relationship during denoising. In sedimentary metamorphic iron ore areas, volcanic rock interlayers are adjacent to the ore body strata, and the interference signals exhibit directional and periodic characteristics, easily leading to misjudgment.

[0026] 2. Advantages of bioorthogonal wavelet bases

[0027] Biorthogonal wavelet bases are an ideal choice for extracting mineral-induced anomalies from volcanic rock interference. Their advantages in many aspects enable them to effectively overcome the limitations of orthogonal wavelet bases and achieve the optimal balance between interference suppression and signal preservation.

[0028] Regarding linear phase characteristics, the filter coefficients of the biorthogonal wavelet basis exhibit strict symmetry, ensuring no phase shift occurs during signal transformation. This characteristic is particularly important in areas where volcanic rock interference and ore body anomalies are spatially adjacent. When volcanic rocks overlie the ore body or when volcanic rock interlayers are adjacent to the ore body strata, any phase distortion can lead to a relative shift in the spatial location of the anomaly, affecting the accuracy of ore body localization. The biorthogonal wavelet basis, through linear phase filtering, ensures that the spatial relationship between the ore body and volcanic rock interference remains consistent before and after denoising, providing a reliable spatial reference for subsequent geological interpretation.

[0029] In terms of symmetry, both the wavelet functions and scaling functions of the bioorthogonal wavelet basis are symmetric, giving it a stronger ability to characterize complex waveforms. The anomalous morphology of volcanic rock disturbances typically exhibits asymmetric and rapidly changing characteristics. A wavelet basis with good symmetry can more accurately capture these waveform features, achieving effective separation of the disturbance and the target signal in the wavelet coefficient domain. Simultaneously, symmetry also helps preserve the morphological integrity of the ore body anomaly, avoiding waveform distortion caused by wavelet basis asymmetry.

[0030] Regarding the reconstruction characteristics after anomaly decomposition, the biorthogonal system ensures that the decomposition filter and the reconstruction filter are dual, achieving distortion-free reconstruction of gravity and magnetic anomaly signals. This is particularly important for anomalies in deep, concealed ore bodies, as the signal amplitude displayed by these anomalies is weak and easily weakened or lost during denoising. Through its reconstruction characteristics, the biorthogonal wavelet basis, while suppressing high-frequency interferences such as volcanic rocks and structures, ensures the complete preservation of low-frequency signals from polymetallic ore bodies or mineralization, as well as mineralization-related magnetic geological bodies. This results in a reconstructed anomaly curve that removes interference while maintaining the true form of the original gravity and magnetic anomaly.

[0031] Regarding the controllability of the vanishing moment, the bioorthogonal wavelet basis can flexibly select an appropriate order of vanishing moment based on the statistical characteristics of volcanic rock interference. The higher the order of the vanishing moment, the stronger the wavelet basis's ability to suppress higher-order polynomial components in the signal. Considering the rich high-frequency components and complex waveforms in volcanic rock interference, selecting a wavelet basis with a high vanishing moment (such as Bior 4.4) can more effectively suppress the interference signal; while for ore body edge regions where detailed information needs to be preserved, a wavelet basis with a lower vanishing moment (such as Bior 3.5) can be selected to achieve a balance between interference suppression and information preservation.

[0032] In terms of the adaptability of the number of decomposition layers, the bioorthogonal wavelet basis exhibits good scale decomposition characteristics, enabling the reasonable setting of the number of decomposition layers based on the frequency range of different volcanic rock interferences. For volcanic cap layer type interference, which has a high interference frequency, the number of decomposition layers can be set to 5-6 layers to disperse the volcanic rock interference into the first three detail subsets. For volcanic interlayer type interference, which has a medium interference frequency, the number of decomposition layers can be set to four to five layers to achieve anomalous separation between the interference signal and the ore body signal at different scales. For interference caused by volcanic structures, which has a wide interference bandwidth, the optimal number of decomposition layers needs to be adaptively determined based on the interference characteristics.

[0033] 3. The suitability of the Brige-Massart strategy for suppressing volcanic rock disturbances.

[0034] The Brige-Massart strategy is an adaptive threshold selection method based on the energy distribution of wavelet coefficients. Its core idea is to retain the coefficients with the highest energy at each decomposition level, while setting the remaining coefficients to zero or attenuating them. This strategy demonstrates good adaptability for suppressing high-frequency and mid-to-high-frequency interference such as volcanic rock, mainly in the following four aspects.

[0035] First, in terms of energy difference identification, the Brige-Massart strategy fully leverages the fundamental differences in wavelet coefficient energy distribution between volcanic rock interference and orebody anomalies. While volcanic rock interference has a larger amplitude, its energy distribution is relatively dispersed, manifested in multiple wavelet coefficients having moderate amplitudes; whereas the energy of orebody anomalies is relatively concentrated, manifested in a few wavelet coefficients having larger amplitudes, while the remaining coefficients have smaller amplitudes. The Brige-Massart strategy, by retaining the coefficients with the highest energy in each layer, effectively preserves the energy-concentrated orebody signal components while suppressing the energy-dispersed volcanic rock interference. This energy difference-based identification mechanism enables effective separation of the two even when the amplitude of volcanic rock interference is comparable to or even higher than that of the orebody signal.

[0036] Secondly, regarding multi-scale adaptation, the Brige-Massart strategy selects thresholds for different decomposition scales, adapting to the scale-dependent distribution characteristics of volcanic rock interference. Volcanic rock interference exhibits different energy distribution patterns at different scales of wavelet decomposition; in the shallow detail subset, volcanic rock interference energy is dominant, while in the deep approximation subset, ore body signal energy is dominant. This strategy, by independently setting the retention coefficient ratio for each layer, allows for a smaller retention ratio in the shallow detail subset to suppress volcanic rock interference, and a larger retention ratio in the deep approximation subset to preserve the ore body signal, achieving optimal denoising results at each scale. This multi-scale adaptive processing overcomes the limitation of global thresholding methods in simultaneously considering signal characteristics at different scales.

[0037] Third, regarding edge preservation, the Brige-Massart strategy employs coefficient energy sorting rather than global threshold truncation, effectively preserving edge information of orebody anomalies. Volcanic rock disturbances exhibit dramatic gradient changes, easily generating large wavelet coefficients at anomaly boundaries. If a global thresholding method is used, these boundary coefficients may be misjudged as disturbances and suppressed, leading to blurred orebody edges. The Brige-Massart strategy retains the coefficients with the highest energy; as long as the energy of the orebody boundary coefficient is greater than the volcanic rock disturbance coefficient in the same layer, it can be effectively preserved. This characteristic results in clear and intact orebody anomaly boundaries after denoising, providing reliable geometric parameters for subsequent orebody location and resource estimation.

[0038] Fourth, regarding the adaptability to different interference types, the Brige-Massart strategy can flexibly adjust the retention coefficient ratio according to the different characteristics of volcanic rock interference, achieving targeted interference suppression. For volcanic cap layer interference, the interference signal has concentrated energy in shallow detail subsets, so a lower retention coefficient ratio (e.g., 8-20%) can be set to strengthen the suppression of volcanic cap layer interference. For volcanic interlayer interference, the interference signal is dominant in wavelet coefficients in a specific direction, so the retention coefficients of different directional sub-bands can be adjusted in combination with directional information to achieve selective suppression of directional interference. For volcanic structure interference, the interference bandwidth is wide and the energy distribution is complex, so the number of decomposition layers can be increased and combined with the multi-scale adaptive characteristics of the Brige-Massart strategy to disperse the volcanic structure interference into multiple detail layers for layer-by-layer suppression. This flexible adaptability to different interference types allows the Brige-Massart strategy to be optimized for the characteristics of volcanic rock interference in different geological scenarios, achieving the best denoising effect.

[0039] 4. A comprehensive approach to wavelet basis selection and strategy application

[0040] Based on the above analysis of the spectral characteristics of volcanic rock interference, the advantages of biorthogonal wavelet bases, and the adaptability of the Brige-Massart strategy, the following comprehensive method is proposed for the volcanic rock interference characteristics of different types of iron polymetallic deposits.

[0041] For skarn-type iron ore deposits, the contact zone is exceptionally complex, and there is a certain separation between the interference from the volcanic caprock and the frequency band of the ore body signal. In the application example, the Bior 3.5 wavelet basis is used. This wavelet basis combines symmetry and tight support characteristics, effectively suppressing high-frequency interference generated by the volcanic caprock while preserving the complex and anomalous morphology of the contact zone. When combined with the Brige-Massart strategy, the number of decomposition layers is set to four to five, and a conservative strategy of retaining coefficients of 10%-15% is adopted to preserve weak mineralization information near the contact zone.

[0042] For volcanic polymetallic iron deposits, the volcanic rock bodies themselves generate strong interference, which highly overlaps with the frequency band of the ore body signal. In this application example, the Bior 4.4 wavelet basis is used. This wavelet basis has a higher vanishing moment, which can enhance the high-frequency suppression capability. When combined with the Brige-Massart strategy, the number of decomposition layers is set to 5-6 layers, dispersing the volcanic rock interference into the first three detail subsets. The retention coefficient ratio is set to 8%-12%, achieving ore body signal extraction under strong interference conditions.

[0043] For sedimentary metamorphic iron ore deposits, the volcanic rock interlayers in these deposits are layered, and the interference signals exhibit directional and periodic characteristics. In this application example, the Bior 2.8 wavelet basis is used. This wavelet basis employs a long filter design, providing more refined frequency band segmentation capabilities. When combined with the Brige-Massart strategy, the number of decomposition layers is set to 5-6. Different retention coefficient ratios are applied to different directional sub-bands based on directional information, with the overall retention coefficient ratio set at 5%-10% to effectively distinguish layered interference from ore body strata.

[0044] II. Application Effects of Typical Examples

[0045] 1. Metallogenic background and disturbance characteristics of typical iron polymetallic deposits

[0046] Xinjiang's iron-polymetallic deposits are mainly distributed in the Tianshan metallogenic belt, the Altai metallogenic belt, and the West Kunlun metallogenic belt, making it one of my country's important iron-polymetallic mineral resource bases. Representative deposits include the Beizhan iron mine and the Chagangnuoer iron mine in the western section of the Tianshan Mountains, the Mengku iron mine in the Altai region, and the Zankan iron mine in the West Kunlun region. The region boasts diverse types of iron-polymetallic deposits, primarily skarn-type and marine volcanic rock-type deposits. The ore bodies are mainly stratiform, lenticular, and layered, with varying depths, ranging from surface outcrops to depths exceeding one thousand meters.

[0047] The main sources of interference in Xinjiang's iron-polymetallic mining areas are volcanic strata and volcanic bodies. In the Tianshan metallogenic belt, Carboniferous-Permian volcanic rocks are widely developed, including basalt, andesite, rhyolite, and pyroclastic rocks. These volcanic rocks often coexist with or overlie ore bodies, generating high-frequency, strong interference. Volcanic bodies themselves have high magnetic susceptibility and density, resulting in large amplitude and dramatic gradient changes in gravity and magnetic anomalies, which highly overlap with ore body anomaly signals in the frequency band, making separation extremely difficult. Especially in marine volcanic iron-polymetallic mining areas, ore bodies and volcanic rocks share the same origin and similar anomaly characteristics, making it difficult for traditional denoising methods to effectively distinguish between ore body signals and volcanic rock interference.

[0048] In the Altai metallogenic belt, Devonian-Carboniferous volcanic rocks are widely distributed. Volcanic rock interlayers are adjacent to ore bodies, and interference signals are distributed in layers, exhibiting clear directional and periodic characteristics. The anomalous wavelengths generated by volcanic rock interlayers are typically between 500 and 1500 meters, overlapping with the anomalous frequency bands of ore bodies. This easily leads to misinterpretation of volcanic rock interlayer anomalies as ore body anomalies, or vice versa. In the West Kunlun metallogenic belt, the volcanic rock cover is thick, exceeding 500 meters in some areas. The high-frequency interference generated by the volcanic rock cover severely suppresses anomalous signals from deep ore bodies, making the identification of deep, concealed ore bodies extremely difficult.

[0049] 2. Classification and identification markers of volcanic rock disturbances

[0050] Volcanic rock interference can be classified into three main types based on its origin and spatial occurrence. Volcanic cap layer interference refers to interference formed when volcanic rock strata overlie the ore body; it is widely developed in the Tianshan metallogenic belt of Xinjiang and the Handan-Xingtai region of Hebei. Its gravity and magnetic characteristics are characterized by high frequency, medium amplitude, and wide distribution. Identifying features include consistency with the distribution range of volcanic rock strata, rapid changes in the anomalous gradient, and dominance of high-frequency components in power spectral analysis. The estimated source depth is typically less than 500 meters.

[0051] Volcanic structural disturbances are disturbances caused by volcanic structures such as volcanic conduits and subvolcanic rock masses, and are relatively common in the Altai metallogenic belt and the West Kunlun metallogenic belt in Xinjiang. Their gravity and magnetic characteristics are characterized by high frequency, high amplitude, and a ring-like or radial distribution. Identifying features are closely related to the location of the crater, often exhibiting ring-like anomalies, and appearing as high-gradient rings on horizontal gradient maps, consistent with the morphology of volcanic structures.

[0052] Volcanic interlayer interference refers to volcanic rocks occurring in interlayers within sedimentary metamorphic rock systems, most typically found in sedimentary metamorphic iron ore areas in eastern Hebei Province. Its gravity and magnetic characteristics are mid-to-high frequency, layered, and exhibit significant directionality. The identifying feature is its alignment with the stratigraphic trend. It often displays multi-layered superposition characteristics, and its dominant direction can be identified through directional filtering and spectral analysis, aligning with the direction of regional tectonic lines.

[0053] 3. Applicability assessment of methods under different metallogenic geological conditions

[0054] (1) Separation of volcanic structures and extraction of iron ore anomalies

[0055] 1) Data processing challenges

[0056] In the volcanic rock-covered area and surrounding gravity and magnetic anomaly zone represented by Chagangnuoer in the Tianshan metallogenic belt of Xinjiang, the geological conditions are characterized by the widespread development of Carboniferous-Permian volcanic rocks, with thicknesses ranging from tens to hundreds of meters, covering the ore bodies. The ore body types include volcanic rock-type iron polymetallic deposits and skarn-type iron deposits. The data processing challenges of this type of metallogenic geological condition are the strong co-origin of interference between the ore bodies and volcanic rocks, the high degree of frequency band overlap, and the strong magnetic properties of the volcanic rocks themselves, resulting in large interference amplitudes.

[0057] 2) Countermeasures of the Invention

[0058] In response to this mineralization geological condition, the following strategies are proposed in this invention: select the wavelet basis of bior4.4 to enhance the high-frequency suppression capability by utilizing its high vanishing moment characteristics; set the number of decomposition layers to 5 to disperse volcanic rock interference into the first three detail subsets; and adopt the Brige-Massart strategy with a retention coefficient ratio of 8-12%.

[0059] 3) Application effect

[0060] ① Ground verification effect

[0061] The reconstruction error can be controlled within 0.003×10⁻⁶ m / s. 2 Below, the signal-to-noise ratio is improved by more than 15 decibels, and known ore bodies and newly discovered deep concealed ore bodies are clearly distinguishable. Figure 8 This effectively overcomes the suppression of deep ore body signals by strong interference from volcanic rocks, while preserving iron ore mineralization anomalies, thus effectively realizing mineralization anomaly ( Figure 8 (Middle pink area) and interfering volcanic rock anomalies ( Figure 8 The separation of the light yellow area (in the middle). The newly discovered deep concealed ore body during the project's surface reconnaissance, and the discovery of iron and copper mineralization clues during the surface reconnaissance, further verify the data extraction effect of this patent.

[0062] ② Comparison of Gravity and Magnetic Processing and Remote Sensing Extraction Effects

[0063] The following figure shows the method of the present invention ( Figure 8 ) and remote sensing data ( Figure 9The extracted volcanic structure. As can be seen from the two renderings, the method of this invention can intelligently and automatically extract volcanic structures and mineral-induced anomalies. The location of the volcanic structure corresponds well with the location of the volcanic structure identified by remote sensing in the following figure. However, this invention utilizes the characteristics of gravity and magnetic fields to extract concealed volcanic rock masses, thus the extracted volcanic structure is closer to the actual geological situation. In addition, the remote sensing influence only reflects the volcanic structure and fails to extract the range of mineral-induced anomalies at the same time. The method of this invention is faster, more efficient, and more multifunctional.

[0064] (2) The Songhu Iron Mine and the surrounding newly extracted iron ore mines caused abnormalities.

[0065] 1) Data processing challenges

[0066] The volcanic rock interference geological background, exemplified by the Songhu iron deposit in the Awulale metallogenic belt of Xinjiang, is characterized by Devonian volcanic rocks adjacent to the ore body strata, a disordered distribution of volcanic rocks exhibiting directional and periodic features, and a marine volcanic rock type of iron polymetallic deposit. The main challenge of this geological background is the overlap of interference signals from volcanic rock interlayers with the anomalous frequency bands of the ore body, which is highly directional and prone to misjudgment.

[0067] 2) Countermeasures of the Invention

[0068] In response to this mineralization geological condition, the following strategies are recommended: select the Bior 2.8 wavelet basis and utilize its long filter characteristics to achieve fine frequency band segmentation; set the number of decomposition layers to 5-6 layers to identify anomalous differences between different layers; combine the Brige-Massart strategy with directional information and use differentiated retention coefficients for different directional sub-bands, with the overall retention coefficient ratio set to 5-10%.

[0069] 3) Application effect

[0070] ① Mineral-induced anomalies and typical iron ore extraction effects

[0071] Extracting mid-to-low frequency and deep-source features is the core objective of using the Brige-Massart strategy to extract orogenic anomalies in iron-polymetallic deposits in the Songhu area. Using the bior2.8 wavelet basis, a 6-layer wavelet decomposition was performed to identify volcanic interlayers, the Dahalajunshan Formation tuff, and copper and other polymetallic mineralization layers within the new area surrounding the Songhu iron mine.

[0072] ② Air-ground coordinated guidance for drilling deployment

[0073] Regarding expected application results, the Brige-Massart strategy can successfully distinguish between volcanic rock interlayers and ore body strata, reducing the misjudgment rate by approximately 40%, effectively guiding borehole layout based on mineralized anomalies. For the identification of the Songhu iron deposit and iron ore mineralized anomalies in the Western Tianshan metallogenic belt using the method of this invention, see [link to relevant documentation]. Figure 10As shown, borehole AW01 was drilled at the center of the iron ore anomaly in the new area. Currently, it has encountered a 21.6m thick pyrite mineralization layer, a 2m thick layer of massive, industrial-grade magnetite tuff, and a 12.1m thick layer of medium- to low-grade magnetite tuff. Surface verification results indicate that the Brige-Massart strategy can successfully locate the ore body strata.

[0074] 4. Application Recommendations

[0075] Given the diverse geological backgrounds of Xinjiang's polymetallic iron deposits, the following points should be noted during the data acquisition phase: In the volcanic rock-covered areas of the Xinjiang Tianshan metallogenic belt, sampling density needs to be increased, with a recommended spacing of 50-100 meters between measurement points to ensure effective sampling against high-frequency interference; in the volcanic rock interlayer areas of the Xinjiang Altai metallogenic belt, measurement lines should be laid out along the direction perpendicular to the regional structural lines to collect directional anomaly characteristics.

[0076] During the data processing phase, parameter tuning should be conducted in conjunction with known borehole data, using cross-validation to determine the optimal wavelet basis and number of decomposition layers. For the strongly disturbed areas of the Xinjiang Tianshan metallogenic belt, the bior4.4 wavelet basis is preferred, with 5-6 decomposition layers. Threshold optimization should dynamically adjust the Brige-Massart retention coefficient based on the intensity of volcanic rock interference; the stronger the interference, the lower the retention coefficient should be. The retention coefficient for the Xinjiang Tianshan metallogenic belt can be set to 8-12%.

[0077] During the results interpretation phase, the denoised anomalies should be overlaid with volcanic rock distribution maps and regional geological maps. Combined with surface reconnaissance results and drilling verification results, the interference suppression effect should be comprehensively evaluated. Based on the geological profiles of typical Xinjiang deposits, the extracted mineralized anomalies should be correlated with favorable mineralization locations to ensure the reliability and geological rationality of the interpretation results.

[0078] This invention provides a method for extracting mineral-induced anomalies in iron polymetallic ores using the Brige-Massart strategy with biorthogonal wavelet bases, comprising the following steps: 1) First, clarify the concept of multi-resolution.

[0079] Multiresolution is based on a stepwise partitioning of the function space, dividing square-integrable functions... This can be viewed as a limit case that is approximated stepwise. In this process, a low-pass smoothing function is used. right Smoothing is performed to obtain an approximation for each level, while the smoothing function... The process of approximating stepwise involves scaling up and down at each step, resulting in the concept of multi-resolution. This process involves using different resolutions to approximate a smooth function stepwise. .

[0080] In short, multiresolution analysis involves constructing a set of function spaces, each with a uniform structure, where the closure of all spaces approximates a given value. In each space, all functions form a normalized orthogonal basis for that space, and all functions within the space closure of a function form a... The normalized orthogonal basis. Therefore, when decomposing a signal in this type of space, mutually orthogonal time-frequency characteristics can be obtained, thus revealing the signal characteristics at certain locations.

[0081] Secondly, orthogonal wavelet bases are constructed using multi-resolution analysis.

[0082] The definition of multi-resolution is: spatial Multiresolution analysis in [the context of the text] refers to... A spatial sequence that satisfies the conditions of monotonicity, approximation, scaling, and translation invariance. The set of functions constituting these conditions forms a multiresolution analysis, and the function that generates such a multiresolution analysis is called the scaling function. For example, if... To generate a multiresolution analysis, then It is called a scaling function.

[0083] From the orthogonal decomposition theory in functional spaces, we have:

[0084] in, It is the adjacent wavelet space; It is wavelet space; It is an orthogonal decomposition; It is an operation; according to: ,

[0085] get:

[0086] Where j is the scale; It is a wavelet space with scale j; It is composed of the straight sums of infinitely many orthogonal complementary spaces, ultimately yielding... The orthonormal basis of space is: ; in, It is a wavelet function; t is a variable; ; It is worth noting that orthogonal wavelets are a special case of bioorthogonal wavelets, which in turn are a generalization of orthogonal wavelets after removing certain degrees of orthogonality. Although there may be some data redundancy when using bioorthogonal wavelets, they perform a linear shift on the original signal through linear phase filtering, which, in this sense, improves the symmetry of orthogonal wavelets.

[0087] 2) Then, perform the orthogonal wavelet transform.

[0088] Based on multiresolution theory, Mallat proposed a fast wavelet decomposition and reconstruction algorithm, known as the Mallat algorithm. This algorithm unifies all previous methods for constructing orthogonal wavelet bases, providing a method for constructing orthogonal wavelets and a fast algorithm for orthogonal wavelet transform. The Mallat algorithm holds a position in wavelet analysis equivalent to that of the Fast Fourier Transform algorithm in classical Fourier analysis.

[0089] Any function All can be based on a resolution of 2 -N The low-frequency part of the time signal and The high-frequency components are completely reconstructed. The core of multiresolution analysis lies in the fact that the ultimate goal of decomposition is to construct a frequency-high approximation. Orthogonal wavelet bases in space. These orthogonal wavelet bases with different frequency resolutions are equivalent to bandpass filters with varying bandwidths. By continuously decomposing the low-frequency components, multiresolution analysis can improve the frequency resolution.

[0090] The above equation shows that the wavelet functions of bioorthogonal wavelets are orthogonal under different scale scaling, but the wavelet function systems obtained by translation within the same scale are not orthogonal. Therefore, the wavelet functions used in the decomposition and reconstruction processes are not the same, and the corresponding filters are generated by these two functions respectively.

[0091] Orthogonal wavelets (except for Haar wavelets) generally lack symmetry, meaning that modifying wavelet coefficients can cause a shift in the signal spectrum, altering the inherent properties of the original signal. In contrast, bioorthogonal wavelets can be constructed as symmetric filters with linear phase. Thus, when the original signal is transformed to the wavelet domain, any manipulation of the coefficients only changes the signal amplitude, not its frequency distribution. In applications of both types of wavelets, the wavelet bases of orthogonal wavelets are mutually orthogonal, so the wavelet decomposition coefficients are also independent, maximizing the removal of noise information, but they generally do not possess generalized linear phase.

[0092] Multiresolution analysis decomposes the low-frequency component of each signal layer, ignoring the high-frequency component. The decomposition process follows a specific relationship: S = A³ + D³ + D² + D¹, where S represents the original signal, A and D represent the low-frequency and high-frequency components respectively, and the numbers indicate the decomposition levels. For any given layer, wavelet decomposition divides it into a low-frequency component D. i and high-frequency part A i Then D i Part of it is further decomposed into D at the next lower level. i+1 and A i+1 This process is repeated layer by layer, such as... Figure 2 As shown.

[0093] Based on the characteristics of biorthogonal wavelets, their linear filters can preserve some redundant information between phase shifts in a signal, which plays an important role in signal reconstruction. Biorthogonal wavelets have an advantage that orthogonal wavelets lack: they can accurately reconstruct signals using their own filters. Although tightly supported orthogonal wavelet bases lack symmetry, they possess excellent time-frequency analysis capabilities. In the abrupt changes of a signal, some wavelet components exhibit large amplitudes, which contrasts sharply with the uniform behavior of noise in the high-frequency range. Therefore, utilizing the characteristics of orthogonal wavelets can effectively distinguish between abrupt changes and noise in a signal.

[0094] 3) Next, complete the split, prediction, and update.

[0095] 1. Decomposition: Input signal S i Divided into two smaller subsets S i-1 and d i-1 d i-1 Also known as wavelet subsets. The simplest decomposition method is based on the input signal S. i The wavelet generated by dividing the wavelet based on its parity is called a lazy wavelet, and the decomposition process is represented as F(S). i )=(S i-1 ,d i-1 ), where F(S) i ) is the decomposition process ( Figure 1 ).

[0096] 2. Prediction: Based on the correlation of the original data, use the even sequence S i-1 The predicted value P(S) i-1 To predict (or interpolate) odd sequence d i-1 The filter P is applied to the even-numbered signal to obtain the predicted value of the odd-numbered signal. The actual value of the odd-numbered signal is subtracted from the predicted value to obtain the residual signal. In practice, although it is impossible to obtain the residual signal from the subset S... i-1 Accurately predict subset di-1 However, P(S) i-1 It is very close to d. i-1 Therefore, P(S) can be used. i-1 ) and d i-1 The difference is used to replace the original d. i-1 d generated in this way i-1 Compared to the original d i-1 Containing less information, we get d i-1 =d i-1 -P(S) i-1 Here, a smaller subset S can already be used. i-1 and wavelet subset d i-1 To replace the original signal set S i The decomposition and prediction process is repeated, and after n steps, the original signal set is obtained. Figure 1 ).

[0097] 3. Update: Figure 1 In order to make certain global properties of the original signal set appear in its subset S i-1 To maintain this, updates are necessary. To achieve this, a better subset S needs to be found. i-1 This ensures that it retains a certain scalar characteristic Q(x) of the original signal (such as the mean, vanishing moment, etc.), i.e., Q(S) remains unchanged. i-1 )=Q(S i This can be achieved by utilizing the already computed wavelet subset d. i-1 For S i-1 This is achieved by updating S, thus preserving its property Q(x). Specifically, an operator U can be constructed to update S. i-1 The definition is as follows: S i-1 =S i-1 +U(d i-1 ) The diagram illustrating the decomposition and reconstruction of the algorithm is as follows: Figure 1 As shown. Throughout the process, the lifting method implements in-situ computation, meaning the algorithm does not require data other than the output of the previous lifting step; at each point, the old data stream can be replaced with the new data stream. When the in-situ lifting filter is reused, interleaved wavelet transform coefficients are obtained.

[0098] The lifting method, based on known wavelet filter coefficients and perfect reconstruction conditions, accelerates wavelet transform without consuming excessive memory, and can be considered a more efficient algorithm for wavelet transform. The Mallat algorithm, based on convolutional discrete wavelets, is computationally intensive, computationally complex, and requires significant storage space, making it unsuitable for hardware implementation. Compared to the Mallat algorithm, the lifting algorithm is a more efficient wavelet transform implementation, hailed as the second-generation wavelet transform. It inherits the multi-resolution characteristics of wavelets, boasts fast computation speed, and requires no additional storage during computation, effectively saving computational storage space.

[0099] 4) Then, the Brige-Massart strategy of biorthogonal wavelet basis is used to eliminate interference noise and preserve the original information of the abnormal signal; By employing the Brige-Massart strategy with biorthogonal wavelet bases, data interference noise (including model data or measured data) is eliminated, while retaining the effective information in the signal.

[0100] 5) Finally, the core parameters were adjusted and the effects of gravity and magnetic data processing were verified; By combining magnetotelluric sounding inversion profiles, seismic data, and drilling data, and adjusting the settings of core parameters (wavelet basis, threshold, and selection of high-frequency and low-frequency components), the desired effect of gravity and magnetic data processing can be achieved, thereby providing scientific and technical support for basic geological research and energy resource prediction in key iron polymetallic mineral zones.

[0101] To examine the actual effectiveness of different wavelet bases and denoising methods on anomalies of different depths; Three spheres at different burial depths were used to simulate a stable anomalous signal. Figure 3 Gaussian white noise was added to the model to obtain the abnormal curves before and after noise addition. Figure 4 Before adding noise, Figure 5 (After adding noise). Figure 6 The result of denoising using Fourier transform shows a maximum error of 0.0045 × 10⁻⁶ before and after reconstruction. -6 m / s 2 This error falls between that of the first-generation wavelet and its lifting wavelet. Compared to the first-generation Bior 3.5 wavelet, after threshold selection, the standard deviation of the Fourier transform is slightly larger, the signal-to-noise ratio improvement is also slightly smaller, and the anomaly curve after the inverse transform has small fluctuations at the troughs, resulting in a poorer image fitting effect than the Bior 3.5 wavelet. Using orthogonal wavelets for denoising, the results show that orthogonal wavelets can remove a large amount of noise, and the reconstructed signal is smoother. Figure 7 ).

[0102] By comparing the denoising effects of different methods, we can conclude that Fourier transform denoising is characterized by a small error between the reconstructed signal and the original signal, and a significant improvement in the signal-to-noise ratio (SNR). The denoising effect is best when using a wavelet threshold determined by the Brige-Massart strategy. Based on this strategy, the wavelet reconstruction error is the smallest among all methods, the SNR improvement is the largest, and the standard deviation before and after wavelet reconstruction is the smallest. This strategy helps improve the denoising effect while ensuring that important signal features are not destroyed.

[0103] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0104] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for extracting mineral-induced anomalies in iron polymetallic ores using the Brige-Massart strategy with biorthogonal wavelet bases, characterized in that, Includes the following steps: Step 1: Construct orthogonal wavelet bases using multiresolution analysis; Step 2: Complete the orthogonal wavelet transform; Step 3: Complete decomposition, prediction, and update; Step 4: For interference anomalies or noise caused by superimposed volcanic rocks, shallow intrusive bodies and near-surface inhomogeneities in gravity and magnetic anomalies caused by iron polymetallic ore, the Brige-Massart strategy of biorthogonal wavelet basis is used to eliminate interference anomalies or noise, while maintaining the mineral anomalies caused by iron polymetallic ore bodies. Step 5: Examine the actual extraction effect of different wavelet bases and denoising methods on anomalies caused by iron polymetallic ore bodies at different depths.

2. The method for extracting mineral-induced anomalies in iron polymetallic ores using the Brige-Massart strategy based on biorthogonal wavelet bases according to claim 1, characterized in that, The specific content of step 1 is as follows: Clarify the concept of multi-resolution: Based on the stepwise partitioning of the function space, square-integrable functions... It is considered as a limit approximated stepwise; a low-pass smoothing function is used. Smoothing is performed to obtain an approximation for each level, while the smoothing function... The process of approximating stepwise is scaled up and down stepwise to form multiple resolutions; Constructing orthogonal wavelet bases: space Sequences of function spaces satisfying the conditions of monotonicity, approximation, scaling, and translation invariance constitute multiresolution analysis; closure approximation of all spaces. Furthermore, the functions in each space constitute a standardized orthogonal basis for the current space; From the orthogonal decomposition theory in functional spaces, we have: in, It is the adjacent wavelet space; It is wavelet space; It is an orthogonal decomposition; It is an operation; according to: , get: Where j is the scale; It is a wavelet space with scale j; It is composed of the straight sums of infinitely many orthogonal complementary spaces, ultimately yielding... The orthonormal basis of space is: ； in, It is a wavelet function; t is a variable; ; Biorthogonal wavelet basis performs linear shift on the original signal through linear phase filtering, which improves the symmetry of orthogonal wavelets and reduces signal distortion. For gravity and magnetic data of iron polymetallic mining areas, anomalies generated by high-frequency geological bodies of volcanic rocks and intrusive rocks overlap with ore body anomalies in the frequency band. Using a more symmetrical biorthogonal wavelet basis can reduce phase distortion during the filtering process, thereby distinguishing between mineral-induced anomalies and surrounding rock interference.

3. The method for extracting mineral-induced anomalies in polymetallic iron ore under the Brige-Massart strategy based on biorthogonal wavelet bases according to claim 1, characterized in that, The specific content of step 2 is as follows: Based on the Mallat algorithm, any function Based on a resolution of 2 -N The low-frequency part of the time signal and The high-frequency components are completely reconstructed. ; Multiresolution analysis improves frequency resolution by continuously decomposing the low-frequency components. Multi-resolution analysis decomposes the low-frequency part of each layer of signal. The decomposition process has a specific relationship: S=A3+D3+D2+D1, where S represents the original signal, A and D represent the low-frequency and high-frequency parts respectively, and the numbers represent the decomposition levels. The original signal S contains mineral anomalies caused by iron polymetallic ore bodies or mineralization and interference caused by volcanic rocks. The two can be effectively separated through multi-scale decomposition.

4. The method for extracting mineral-induced anomalies in polymetallic iron ore under the Brige-Massart strategy with biorthogonal wavelet bases according to claim 1, characterized in that, The specific content of step 3 is as follows: Decomposition: Decompose the input signal S i Divided into two smaller subsets S i-1 and d i-1 The decomposition process is represented as F(S) i )=(S i-1 ,d i-1 ), where F(S) i () represents the decomposition process; In gravity and magnetic data processing closely related to iron polymetallic deposits, the decomposition process requires the rational selection of the number of decomposition layers based on the planar distribution scale and vertical burial depth of the ore body, so that the ore body signal is concentrated in the low-frequency subset, while the high-frequency interference of volcanic rocks enters the detail subset.

5. The method for extracting mineral-induced anomalies in polymetallic iron ore under the Brige-Massart strategy based on biorthogonal wavelet bases according to claim 1, characterized in that, The specific content of the prediction in step 3 is as follows: Based on the correlation of the original data, use the even-numbered sequence S i-1 The predicted value P(S) i-1 To predict odd sequence d i-1 The filter P is applied to the even signal and used as the predicted value of the odd signal. The actual value of the odd signal is subtracted from the predicted value to obtain the residual signal. The decomposition and prediction process is repeated, and the original signal set is obtained after n steps.

6. The method for extracting mineral-induced anomalies in polymetallic iron ore under the Brige-Massart strategy based on biorthogonal wavelet bases according to claim 1, characterized in that, The specific content of the update in step 3 is: using the already calculated d i-1 For S i-1 The update is performed so that the global characteristics of the original signal set are reflected in the subset S. i-1 Continue to maintain this, that is, have Update S by constructing an operator U. i-1 The definition is as follows: When the in-situ boost filter is reused, interleaved wavelet transform coefficients are obtained.

7. The method for extracting mineral-induced anomalies in polymetallic iron ore under the Brige-Massart strategy with biorthogonal wavelet bases according to claim 1, characterized in that, Step 5 involves: combining magnetotelluric sounding inversion profiles, seismic data, and drilling data; adjusting core parameter settings for typical iron polymetallic deposits to ensure that gravity and magnetic data processing achieves the desired results; and verifying the accuracy of the method by comparing the processing results under different methods and parameter settings.

Images