A method and system for achieving precision prediction of a metal detector

Abstract

This invention relates to the field of metal detectors, specifically a method for predicting the accuracy of a metal detector. The method includes: setting the frequency of the electromagnetic field between the probe coils of the metal detector to a preset low-frequency frequency; obtaining the first resistance change and the first reactance change of the probe coils when a product sample passes through the electromagnetic field; calculating the dielectric loss factor and capacitive coupling factor of the product sample; calculating the predicted resistance change and predicted reactance change values ​​of multiple metal spheres; obtaining the principal characteristic vector of the product sample, and obtaining resistance change thresholds and reactance change thresholds based on the principal characteristic vectors; and determining the minimum diameter among all metal spheres whose predicted resistance change value is greater than the resistance change threshold or whose predicted reactance change value is greater than the reactance change threshold as the accuracy of the metal detector. This invention avoids the prediction bias caused by neglecting product effects in traditional methods.

Description

Technical Field

[0001] This invention generally relates to the field of metal detection. More specifically, this invention relates to a method and system for predicting the accuracy of a metal detector. Background Technology

[0002] Metal detectors use a transmitting coil inside the probe to generate an alternating electromagnetic field within a window. When an object (product or metallic foreign object) passes through the window, it causes a change in the electromagnetic field. The receiving coil detects this change (the change in coil impedance) and converts it into an electrical signal (such as resistive and reactive components), thereby determining whether a metallic foreign object is present. The definition of detection accuracy refers to the smallest size (diameter) of a metal sphere that the device can detect. For example, "accuracy FE1.0" means it can detect an iron sphere with a diameter of 1.0 mm.

[0003] Some metal detectors now have accuracy prediction capabilities. After setting the parameters, they can predict the diameter of a single metal test ball that the device can detect under the current conditions.

[0004] However, existing prediction methods only predict the presence of a single metal sphere (i.e., only the metal sphere passes through the air). In actual production, however, metal foreign objects are always mixed with products (such as meat and food) and pass through the probe together. The metal signal and the product signal are not simply linearly added. When the metal (primarily generating eddy currents, altering reactance and resistance) and the product (primarily generating dielectric effects) mix, complex electromagnetic interactions occur (such as resonant frequency shifts and Q-value changes), leading to nonlinear signal superposition. Because the product effect is ignored, existing predictive metal detectors may lack accuracy and cannot provide users with a truly effective reference for quality control when inspecting specific products. Summary of the Invention

[0005] To address the aforementioned technical problem of potentially inaccurate predictions of metal detector accuracy, this invention provides a method for predicting the accuracy of metal detectors.

[0006] A method for predicting the accuracy of a metal detector includes: setting the frequency of the electromagnetic field between the probe coils of the metal detector to a preset low-frequency frequency, and obtaining the first resistance change and the first reactance change of the probe coils when a product sample passes through the electromagnetic field; calculating the dielectric loss factor and capacitive coupling factor of the product sample, wherein the dielectric loss factor is proportional to the first resistance change and the capacitive coupling factor is proportional to the first reactance change; calculating the predicted resistance change and the predicted reactance change of multiple metal spheres, wherein any two metal spheres have different diameters; wherein calculating the predicted resistance change and the predicted reactance change of any one metal sphere includes: determining the frequency of the electromagnetic field to a preset high-frequency frequency, and obtaining the first resistance change and the first reactance change of the probe coils when a product sample passes through the electromagnetic field. The changes in the second resistance and the second reactance of the probe coil during operation are used to calculate the eddy current resistivity and inductive reactance attenuation coefficient of the metal. The eddy current resistivity is proportional to the change in the second resistance, and the inductive reactance attenuation coefficient is proportional to the change in the second reactance. A preset operating frequency is obtained, and the predicted values ​​of the resistance and reactance changes of the metal spheres are calculated based on the operating frequency, dielectric loss factor, capacitive coupling factor, eddy current resistivity, and inductive reactance attenuation coefficient. The principal feature vector of the product sample is obtained, and the threshold values ​​of the resistance and reactance changes are obtained based on the principal feature vector. The minimum diameter among all metal spheres whose predicted resistance or reactance changes are greater than the threshold values ​​is determined as the accuracy of the metal detector.

[0007] In one embodiment, the formula for calculating the eddy current resistance coefficient of the metal is: ,in k 1 represents the eddy current resistivity of the metal. This is the change in the second resistance. f H The high frequency is referred to as "the high frequency".

[0008] In one embodiment, the formula for calculating the metallic resistivity attenuation coefficient is: ,in k 1 represents the eddy current resistivity of the metal. This represents the change in the second reactance.

[0009] In one embodiment, the formula for calculating the dielectric loss factor of the product is: ,in A The dielectric loss factor of the product. This represents the change in the first resistance. f L The low frequency is referred to as the low frequency.

[0010] In one embodiment, the formula for calculating the capacitive coupling factor of the product is: ,in B For the product's capacitance coupling factor, This represents the change in the first reactance.

[0011] In one embodiment, calculating the predicted resistance change and the predicted reactance change of the metal sphere includes: calculating the predicted resistance change of the metal sphere, wherein the calculation formula is:

[0012] ;

[0013] This is the predicted value of the resistance change. f M Given the operating frequency; calculate the predicted value of the reactance change of the metal sphere, where the calculation formula is:

[0014] ;

[0015] This is the predicted value of the resistance change;

[0016] The predicted vector of the metal sphere is determined as follows: .

[0017] In one embodiment, obtaining the main feature vector of a product sample includes: setting the electromagnetic field frequency between the probe coils of the metal detector to the operating frequency, and then obtaining the target feature vector at multiple consecutive moments during the process of the product sample passing through the electromagnetic field, wherein the first... j The target feature vector at each time step is During the process of the product sample passing through the electromagnetic field, For the probe coil at the first j The change in resistance at each moment. For the probe coil at the first j The change in probe coil reactance at each moment. j The index represents the target feature vector with the largest modulus among all target feature vectors at all times, which is the main feature vector of the product sample.

[0018] In one embodiment, obtaining the resistance change threshold and reactance change threshold based on the principal feature vector includes: calculating the resistance change threshold using a preset first margin coefficient, wherein the calculation formula is: , This represents the resistance change corresponding to the principal characteristic vector. k x This is the first margin coefficient, and k x >1; The reactance change threshold is calculated using a preset second margin coefficient, where the calculation formula is: , is the change in reactance corresponding to the feature vector. k y It is the second margin coefficient, and ky >1.

[0019] In one embodiment, obtaining the first resistance change and the first reactance change of the probe coil when the product sample passes through an electromagnetic field includes: setting the frequency of the electromagnetic field between the probe coils of the metal detector to a preset low-frequency frequency, and then obtaining the first characteristic vector at multiple consecutive moments during the process of the product sample passing through the electromagnetic field, wherein the first characteristic vector is... s The first feature vector at time is During the process of the metal sphere passing through the electromagnetic field, For the probe coil at the first s The change in resistance at each moment. For the probe coil at the first s The change in probe coil reactance at each moment. s The index represents the first characteristic vector with the largest modulus; the resistance change corresponding to the first characteristic vector with the largest modulus is determined as the first resistance change; the reactance change corresponding to the first characteristic vector with the largest modulus is determined as the first reactance change.

[0020] In one embodiment, obtaining the second resistance change and the second reactance change of the probe coil when the metal ball passes through the electromagnetic field includes: obtaining a second characteristic vector at multiple consecutive moments during the process of the metal ball passing through the electromagnetic field, wherein the... s The second feature vector at each time step is During the process of the metal sphere passing through the electromagnetic field, For the probe coil at the first s The change in resistance at each moment. For the probe coil at the first s The change in reactance of the probe coil at each moment; the change in resistance corresponding to the second characteristic vector with the largest modulus length is determined as the second resistance change; the change in reactance corresponding to the second characteristic vector with the largest modulus length is determined as the second reactance change.

[0021] The beneficial effects of this invention are as follows:

[0022] This invention utilizes the physical properties of low-frequency dominance over product effects and high-frequency dominance over metal effects to accurately extract parameters corresponding to both the product and the metal. Then, by substituting these separated parameters into a unified model at the actual operating frequency for virtual synthesis, it accurately simulates the total impedance change after the metal and product are mixed. This ensures that the prediction results truly reflect the detection capabilities on the production line, avoiding prediction biases caused by neglecting product effects in traditional methods. Attached Figure Description

[0023] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent from the following detailed description taken in conjunction with the accompanying drawings. In the drawings, several embodiments of the invention are illustrated by way of example and not limitation, and like or corresponding reference numerals denote like or corresponding parts, wherein:

[0024] Figure 1 This is a flowchart illustrating the steps of a method for predicting the accuracy of a metal detector according to an embodiment of the present invention. Detailed Implementation

[0025] 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, not all, of the embodiments of the present invention. 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.

[0026] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0027] Figure 1 This is a flowchart illustrating the steps of a method for predicting the accuracy of a metal detector according to an embodiment of the present invention.

[0028] like Figure 1 As shown, a method for predicting the accuracy of a metal detector includes steps S1 to S5.

[0029] Step S1: After setting the electromagnetic field frequency between the probe coils of the metal detector to a preset low frequency, obtain the first resistance change and the first reactance change of the probe coil when the product sample passes through the electromagnetic field.

[0030] In one embodiment, obtaining the first resistance change and the first reactance change of the probe coil when the product sample passes through an electromagnetic field includes: setting the frequency of the electromagnetic field between the probe coils of the metal detector to a preset low-frequency frequency, and then obtaining the first characteristic vector at multiple consecutive moments during the process of the product sample passing through the electromagnetic field, wherein the first characteristic vector is... s The first feature vector at time is During the process of the metal sphere passing through the electromagnetic field, For the probe coil at the first s The change in resistance at each moment. For the probe coil at the first s The change in probe coil reactance at each moment. s The index represents the first characteristic vector with the largest modulus; the resistance change corresponding to the first characteristic vector with the largest modulus is determined as the first resistance change; the reactance change corresponding to the first characteristic vector with the largest modulus is determined as the first reactance change.

[0031] It should be noted that the electromagnetic field frequency is set to a preset low frequency because, in a low-frequency environment, the dielectric effect of non-metallic products (especially foods containing water or salt) is dominant, while the eddy current effect of metals is relatively weak. The changes in resistance and reactance obtained at this time mainly reflect the dielectric loss characteristics and capacitive coupling characteristics of the product. Ignoring the influence of metal parameters allows for the extraction of the product's characteristic parameters with extremely high accuracy, providing accurate basic data for subsequent decoupling calculations.

[0032] Step S2: Calculate the dielectric loss factor and capacitive coupling factor of the product sample.

[0033] The dielectric loss factor is proportional to the change in the first resistance, and the capacitive coupling factor is proportional to the change in the first reactance.

[0034] In one embodiment, the formula for calculating the dielectric loss factor of the product is: ,in A The dielectric loss factor of the product. This represents the change in the first resistance. f L The low frequency is referred to as the low frequency.

[0035] In one embodiment, the formula for calculating the capacitive coupling factor of the product is: ,in B For the product's capacitance coupling factor, This represents the change in the first reactance.

[0036] It should be noted that the dielectric loss factor A characterizes the energy loss of the product in an alternating electromagnetic field due to molecular polarization friction and other factors, and it mainly affects the resistive component of the coil; the capacitive coupling factor B characterizes how the product, as a dielectric, alters the distributed capacitance between the probe coil and the object, and it mainly affects the reactive component of the coil. The calculation logic in the formula is based on a simplified physical model, that is, assuming that the influence of eddy currents in the metal approaches zero at low frequencies, thus simplifying the complex coupling equations into a linear relationship that can be solved directly.

[0037] Step S3: Calculate the predicted values ​​of resistance change and reactance change for multiple metal spheres.

[0038] Any two metal spheres have different diameters. It should be noted that all metal spheres are made of the same material.

[0039] In one embodiment, the diameters of the multiple metal spheres can form an arithmetic sequence, i.e., the nth... q The diameters of the metal spheres satisfy the formula: , S q For the first qThe diameter of the metal sphere, K The preset scaling parameters, b The preset intercept, q Represents an index.

[0040] The calculation of the predicted resistance change and the predicted reactance change of any metal ball includes steps S301 to S302.

[0041] Step S301: After determining that the frequency of the electromagnetic field is a preset high-frequency frequency, obtain the changes in the second resistance and the second reactance of the probe coil when the metal ball passes through the electromagnetic field, and calculate the metal eddy current resistance coefficient and the metal inductive reactance attenuation coefficient. The metal eddy current resistance coefficient is proportional to the change in the second resistance, and the metal inductive reactance attenuation coefficient is proportional to the change in the second reactance.

[0042] In one embodiment, obtaining the second resistance change and the second reactance change of the probe coil when the metal ball passes through the electromagnetic field includes: obtaining a second characteristic vector at multiple consecutive moments during the process of the metal ball passing through the electromagnetic field, wherein the... s The second feature vector at each time step is During the process of the metal sphere passing through the electromagnetic field, For the probe coil at the first s The change in resistance at each moment. For the probe coil at the first s The change in reactance of the probe coil at each moment; the change in resistance corresponding to the second characteristic vector with the largest modulus length is determined as the second resistance change; the change in reactance corresponding to the second characteristic vector with the largest modulus length is determined as the second reactance change.

[0043] In one embodiment, the formula for calculating the eddy current resistance coefficient of the metal is: ,in k 1 represents the eddy current resistivity of the metal. This is the change in the second resistance. f H The high frequency is referred to as "the high frequency".

[0044] In one embodiment, the formula for calculating the metallic resistivity attenuation coefficient is: ,in k 1 represents the eddy current resistivity of the metal. This represents the change in the second reactance.

[0045] It should be noted that the electromagnetic field frequency is set to a preset high-frequency frequency because in a high-frequency environment, the eddy current effect of metal is significantly enhanced and becomes dominant, while the dielectric effect of the product is relatively weak and can be ignored.

[0046] Before actual production, a database of metal ball parameters is pre-established, containing parameters for various materials (such as iron, copper, and stainless steel) and sizes (such as diameters from 0.5mm to 4.0mm). When predicting accuracy, the system directly calls these pre-stored metal eddy current resistivity coefficients and metal inductive reactance attenuation coefficients corresponding to different diameters, without requiring users to perform real-time metal ball tests.

[0047] Step S302: Obtain the preset operating frequency, and calculate the predicted values ​​of the resistance change and reactance change of the metal ball based on the operating frequency, dielectric loss factor, capacitive coupling factor, eddy current resistance coefficient and inductive reactance attenuation coefficient.

[0048] In one embodiment, calculating the predicted resistance change and the predicted reactance change of the metal sphere includes: calculating the predicted resistance change of the metal sphere, wherein the calculation formula is:

[0049] .

[0050] This is the predicted value of the resistance change. f M The operating frequency is [the specified frequency].

[0051] Calculate the predicted change in reactance of the metal sphere, using the following formula:

[0052] .

[0053] The predicted value of the resistance change is given; the predicted vector of the metal sphere is determined as follows. .

[0054] It should be noted that the preset operating frequency f M It is the intermediate frequency used by metal detectors when detecting foreign objects on an actual production line, i.e. f L < f M < f H Using the product parameters obtained in steps S1 to S2 ( A and B ) and the metal parameters obtained in step S301 ( k 1 and k 2) Simulate the working frequency using a mathematical model. f M The formula is used to calculate the total impedance change of the probe coil when a metal ball is mixed into the product. Based on a unified electromagnetic framework, it considers the square root growth of the metallic term with frequency (skin effect) and the linear or inverse relationship between the non-metallic term and frequency, thus achieving accurate prediction of nonlinear mixed signals.

[0055] Step S4: Obtain the principal feature vector of the product sample, and obtain the resistance change threshold and reactance change threshold based on the principal feature vector.

[0056] In one embodiment, obtaining the main feature vector of a product sample includes: setting the electromagnetic field frequency between the probe coils of the metal detector to the operating frequency, and then obtaining the target feature vector at multiple consecutive moments during the process of the product sample passing through the electromagnetic field, wherein the first... j The target feature vector at each time step is During the process of the product sample passing through the electromagnetic field, For the probe coil at the first j The change in resistance at each moment. For the probe coil at the first j The change in probe coil reactance at each moment. j The index represents the target feature vector with the largest modulus among all target feature vectors at all times, which is the main feature vector of the product sample.

[0057] In one embodiment, obtaining the resistance change threshold and reactance change threshold based on the principal feature vector includes: calculating the resistance change threshold using a preset first margin coefficient, wherein the calculation formula is: , This represents the resistance change corresponding to the principal characteristic vector. k x This is the first margin coefficient, and k x >1; The reactance change threshold is calculated using a preset second margin coefficient, where the calculation formula is: , is the change in reactance corresponding to the feature vector. k y It is the second margin coefficient, and k y >1.

[0058] Step S5: Determine the minimum diameter among all metal spheres whose predicted resistance change value is greater than the resistance change threshold or whose predicted reactance change value is greater than the reactance change threshold as the accuracy of the metal detector.

[0059] It should be noted that the threshold rectangle (enclosed by the resistance change threshold and the reactance change threshold) represents the fluctuation range of the product signal after considering the safety margin, i.e., the "safe zone". The accuracy of a metal detector is defined as the smallest metal ball diameter that can trigger an alarm. Therefore, the specific judgment logic for this step is usually as follows: traverse all metal balls with preset diameters and calculate their predicted vectors. If the predicted resistance change value of a metal ball is greater than the resistance change threshold, or the predicted reactance change value is greater than the reactance change threshold (i.e., the predicted vector exceeds the range of the threshold rectangle), then the metal ball is determined to be detectable (alarm triggered). Among all detectable metal balls, the one with the smallest diameter is determined as the metal detector accuracy for the current product and current parameters.

[0060] It should be noted that, in order to handle the situation where the product and metal are superimposed, it is necessary to analyze the electromagnetic properties of both. In one embodiment, the electromagnetic properties of water and iron balls are analyzed. Iron balls exhibit eddy current effects and permeability effects: an alternating magnetic field induces eddy currents in the metal, which generate a reverse magnetic field, leading to a decrease in the equivalent inductance and an increase in the resistance of the probe coil (manifested as an impedance change); materials such as iron and nickel also exhibit high permeability (… m This alters the magnetic field distribution of the probe coil, further affecting the signal. The signal source for water is capacitive coupling caused by its dielectric constant (ε): the dielectric constant (ε) of non-metallic materials affects the distributed capacitance between the probe coil and the object, thus changing the resonant frequency or impedance phase. Water's ε≈80, oil's ε≈2~5, and meat's ε≈40~60 (varying depending on water content); conductivity ( s A minor contribution: Foods containing electrolytes (such as salt water and sauces) have weak conductivity and may produce eddy current-like effects, but the intensity is much lower than that of metals. Although the eddy current effect of metals and the dielectric effect of nonmetals manifest differently, they are essentially energy interaction mechanisms of materials in electromagnetic fields and can be understood through a unified electromagnetic framework. The equivalent impedance of the probe coil... Z The comprehensive representation is as follows: .in This is the equivalent resistance of the probe coil. This is the equivalent reactance of the probe coil. The equivalent capacitance of the probe coil

[0061] Eddy current losses increase the equivalent resistance of the probe coil, and the reverse magnetic field generated by eddy current losses reduces the inductance; a high dielectric constant increases capacitive coupling; based on this, the above equation can be expressed as follows:

[0062] .

[0063] according to Z The resonant frequency of the probe coil can be obtained from the expression. f 0, where the resonant frequency of the probe coil is 0.f 0 satisfies the formula: .

[0064] The quality factor Q of the probe coil satisfies the formula .

[0065] It can be seen that metals primarily affect the accuracy prediction of metal detectors. L Achieving accuracy prediction for metal detectors and achieving accuracy prediction for metal detectors R (Eddy currents); Non-metallic substances mainly affect the accuracy prediction of metal detectors. C (Dielectric constant). Its essence is that metals and non-metals pass through each other via free charges ( s ) and bound charge ( e This enables energy interaction, thereby affecting the electromagnetic field within the probe coil.

[0066] From the perspective of a resonant system, there are two factors that cause nonlinear superposition of signals: the competition of resonant frequency shifts and the complex changes in Q value.

[0067] Competition for resonant frequency shift: The metal causes L to decrease, which in turn affects the resonant frequency. f An increase in 0 causes water to increase C, which in turn leads to f 0 decreases. When both coexist, the resonant frequency may stagnate at the intermediate value (not a simple arithmetic mean), or even exhibit double-peak resonance due to system nonlinearity;

[0068] The Q value exhibits complex variations: metals increase R, leading to a decrease in Q; water may increase dielectric loss (dielectric relaxation loss), potentially causing a further decrease in Q. Total loss may also show synergistic enhancement (e.g., the loss from salt water + metal is much higher than the sum of the losses of either).

[0069] As can be seen, the existence of the above factors makes the processing of mixed metallic and non-metallic signals very difficult. This invention employs an engineering-acceptable method to approximate this nonlinear relationship.

[0070] This invention employs an engineering-acceptable method to approximate this nonlinear relationship, namely, the unified model for probe coil variation, wherein the unified model for probe coil variation is:

[0071] .

[0072] in, oh The frequency of the electromagnetic field, k 1 represents the metal eddy current resistivity (also known as the metal eddy current loss coefficient). k 2 represents the eddy current resistivity of the metal (also known as the inductive reactance attenuation coefficient of the metal). k 3 represents the dielectric loss coefficient of non-metallic materials. k4 represents the non-metallic capacitive reactance adjustment coefficient. e 1 represents the real part of the dielectric constant of a preset size. e 2 represents the preset dielectric loss factor. j Represents the imaginary unit. This represents the change in impedance.

[0073] Metal eddy current resistivity k The dimension of 1 is Ω· s 1 / 2 Metal eddy current resistivity k The dimension of 2 is Ω· s Non-metallic dielectric loss coefficient k 3. Dimensionless, non-metallic capacitive reactance adjustment coefficient k The dimension of 4 is Ω· s .

[0074] It should be noted that the formula Represents the change in resistance ( ),formula Represents the change in reactance ( ).

[0075] Will ,make , The dielectric loss factor A and capacitive coupling factor B of the product are obtained.

[0076] Obtain the impedance and capacitive reactance of the probe coil at high angular frequencies; obtain the low-frequency points and their corresponding impedance and capacitive reactance. Construct a system of equations and apply them to A, B, k 1 and k 2. Solve the system of equations. , f L The low frequency, f H The high frequency refers to the frequency of the electromagnetic field when the frequency is a low frequency. f L hour, This represents the change in resistance of the probe coil. This represents the change in capacitance of the probe coil when the electromagnetic field frequency is high. f H hour, This represents the change in resistance of the probe coil. The change in capacitance of the probe coil.

[0077] Based on this, the response of a mixture including a specific product and a metal can be obtained at an intermediate frequency after calibration at low and high frequencies. To achieve an engineering-acceptable accuracy prediction target, some simplification is required. Considering that the capacitive reactance effect of non-metals is amplified at low frequencies, while the inductive reactance effect of metals is weakened; and that the eddy current effect of metals increases rapidly with increasing frequency, drowning out the non-metal signal—that is, non-metals dominate at low frequencies and metals dominate at high frequencies—this invention simplifies the calibration equation to: This leads to the determination of the product's dielectric loss factor A, capacitive coupling factor B, and metal eddy current resistance coefficient. k 1 and the metal eddy current resistivity k Solution to 2.

Claims

1. A method of implementing precision prediction of a metal detector, characterized by, include: After setting the electromagnetic field frequency between the probe coils of the metal detector to a preset low frequency, the first resistance change and the first reactance change of the probe coil when the product sample passes through the electromagnetic field are obtained; the dielectric loss factor and capacitive coupling factor of the product sample are calculated, where the dielectric loss factor is proportional to the first resistance change and the capacitive coupling factor is proportional to the first reactance change. Calculate the predicted resistance and reactance changes of multiple metal spheres, where any two metal spheres have different diameters; The calculation of the predicted resistance and reactance changes of any metal ball includes: determining the frequency of the electromagnetic field to be a preset high-frequency frequency, obtaining the second resistance change and second reactance change of the probe coil when the metal ball passes through the electromagnetic field, calculating the metal eddy current resistance coefficient and the metal inductive reactance attenuation coefficient, wherein the metal eddy current resistance coefficient is proportional to the second resistance change and the metal inductive reactance attenuation coefficient is proportional to the second reactance change; obtaining the preset operating frequency, and calculating the predicted resistance and reactance changes of the metal ball based on the operating frequency, dielectric loss factor, capacitive coupling factor, eddy current resistance coefficient, and inductive reactance attenuation coefficient; Obtain the principal feature vector of the product sample, and obtain the resistance change threshold and reactance change threshold based on the principal feature vector; determine the minimum diameter among all metal spheres whose predicted resistance change value is greater than the resistance change threshold or whose predicted reactance change value is greater than the reactance change threshold as the accuracy of the metal detector.

2. The method for predicting the accuracy of a metal detector according to claim 1, characterized in that, The formula for calculating the metal eddy current resistance coefficient is: wherein k 1 is the metal eddy current resistance coefficient, is the second resistance change amount, f H is the high frequency frequency.

3. The method for predicting the accuracy of a metal detector according to claim 2, characterized in that, The formula for calculating the metal inductance attenuation coefficient is: wherein k 2 is the metal inductance attenuation coefficient, is the second reactance variation.

4. The method for predicting the accuracy of a metal detector according to claim 3, characterized in that, The formula for calculating the product dielectric loss factor is: wherein A is the product dielectric loss factor, is the first resistance change amount, f L is the low frequency.

5. The method for predicting the accuracy of a metal detector according to claim 4, characterized in that, The formula for calculating the product susceptance coupling factor is: wherein B is the product susceptance coupling factor, is the first amount of reactance change.

6. The method for realizing precision prediction of a metal detector according to claim 5, characterized in that, The calculation of the predicted resistance and reactance changes of the metal sphere includes: Calculate the predicted resistance change of the metal sphere, using the following formula: ； for the resistance change prediction value, f M for the operating frequency; Calculate the predicted change in reactance of the metal sphere, using the following formula: ； is the predicted value of the change in the electrical reactance; determining a predicted vector of the metal ball as .

7. The method of claim 1, wherein, The main feature vectors of the product sample include: After setting the electromagnetic field frequency between the probe coils of the metal detector to the operating frequency, the target feature vectors at multiple consecutive moments during the process of the product sample passing through the electromagnetic field are obtained, where the first... j The target feature vector at each time step is During the process of the product sample passing through the electromagnetic field, For the probe coil at the first j The change in resistance at each moment. For the probe coil at the first j The change in probe coil reactance at each moment. j Represents an index; The target feature vector with the largest modulus among all target feature vectors at all times is determined as the principal feature vector of the product sample.

8. The method of claim 7, wherein, The resistance change threshold and reactance change threshold obtained from the principal eigenvector include: The resistance change threshold is calculated by a preset first margin coefficient, and the calculation formula is: , is the resistance change amount corresponding to the main feature vector, k x is the first margin coefficient, and k x >1; The reactance change threshold is calculated by a preset second margin coefficient, and a calculation formula is: is a reactance change amount corresponding to the feature vector, k y is the second margin coefficient, and k y >1.

9. The method for predicting the accuracy of a metal detector according to claim 1, characterized in that, The changes in the first resistance and the first reactance of the probe coil when the product sample passes through an electromagnetic field include: After setting the electromagnetic field frequency between the probe coils of the metal detector to a preset low-frequency frequency, the first characteristic vector at multiple consecutive moments during the process of the product sample passing through the electromagnetic field is obtained, where the first characteristic vector is... s The first feature vector at time is During the process of the metal sphere passing through the electromagnetic field, For the probe coil at the first s The change in resistance at each moment. For the probe coil at the first s The change in probe coil reactance at each moment. s Represents an index; The resistance change corresponding to the first eigenvector with the largest modulus is determined as the first resistance change; the reactance change corresponding to the first eigenvector with the largest modulus is determined as the first reactance change.

10. A method for predicting the accuracy of a metal detector according to claim 9, characterized in that, The changes in the second resistance and the second reactance of the probe coil when the metal ball passes through the electromagnetic field include: Obtain the second eigenvector at multiple consecutive moments during the process of the metal sphere passing through the electromagnetic field, where the... s The second feature vector at each time step is During the process of the metal sphere passing through the electromagnetic field, For the probe coil at the first s The change in resistance at each moment. For the probe coil at the first s The change in probe coil reactance at each moment; The resistance change corresponding to the second eigenvector with the largest modulus is determined as the second resistance change; the reactance change corresponding to the second eigenvector with the largest modulus is determined as the second reactance change.

Images