Ship magnetic field simulation method based on magnetic dipole

Abstract

The invention discloses a ship magnetic field simulation method based on magnetic dipoles, and the method comprises the steps: building a three-dimensional distribution physical model of the magnetic dipoles in a space magnetic field, and building a magnetic dipole simulation model; building a multi-magnetic dipole simulation model to simulate a magnetic field generated by the magnetic simulation source through coordinate transformation and magnetic field superposition of the magnetic dipoles; variable control modification is carried out by adopting a variable classification control and single variable principle; the maximum error or the root-mean-square error of the fitting magnetic field is used as a fitness function of a genetic algorithm, and parameters of each magnetic dipole are optimized through the genetic algorithm; different simulation analysis experiment principles are formulated for a pull-type magnetoelectric simulation source and a novel magnetic simulation source; calculating a magnetic field and a target magnetic field, and finally obtaining a better solution considering the number of magnets and a magnetic field fitting effect; the ship magnetic field can be accurately simulated; and a foundation is laid for magnetoelectric simulation and optimization of a novel magnetic simulation source of a subsequent sweeping ship or an unmanned ship.

Description

Technical Field

[0001] The present invention relates to the field of magnetic field simulation, and particularly to a method for simulating the magnetic field of a ship based on magnetic dipoles. Background Art

[0002] Magnetic fuze mines are an important threat to the navigation of modern steel ships. The magnetic simulation source is one of the important equipments in the mine-sweeping operation of mine-sweeping ships. The traditional towed target-setting magnetic simulation source consists of a powered cable, a stressed cable, a solenoid coil and its internal iron core, a ship power supply system, etc. The ship power supply system energizes the coil, and the solenoid coil and its internal iron core generate a strong magnetic field. Using several solenoid coils as magnetic simulation bodies can simulate the magnetic field of a ship in space. During the actual mine-sweeping process, several solenoid coils are linearly arranged at a certain distance behind the ship in a towed manner. The equipment has a depth-setting function and can maintain a stable relative position during the towing process, thus ensuring the accuracy and stability of the simulation.

[0003] The magnetic signal of the ship is simulated to detonate the mine, but the magnetic field of the degaussed ship is complex. The existing magnetic simulation source has insufficient accuracy in simulating the magnetic field of the ship. In recent years, various countries have successively used new unmanned equipment to participate in mine-sweeping, and the parameters of the electromagnetic simulation source have changed again. Summary of the Invention

[0004] The problem to be solved by the present invention is that the existing magnetic simulation source has insufficient accuracy in simulating the magnetic field of a ship.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is:

[0006] A method for simulating the magnetic field of a ship based on magnetic dipoles, comprising the following steps:

[0007] Step 1: Regard each magnet of the electromagnetic simulation source as a magnetic dipole, that is, a circular current, establish a three-dimensional distribution physical model of the magnetic dipole in the space magnetic field, and build a magnetic dipole simulation model on this basis;

[0008] Step 2: Build a multi-magnetic dipole simulation model to simulate the magnetic field generated by the magnetic simulation source through the coordinate transformation and magnetic field superposition of the magnetic dipole;

[0009] Step 3: Adopt the principle of variable classification control and single variable to control and modify variables; among them, variable classification control means dividing variables into magnetic moment magnitude variables and position variables; the single variable principle means observing the change of simulation accuracy when only one or a class of variables is changed;

[0010] Step 4: On the basis of the principle established in Step 3, use the maximum error or root mean square error of the fitted magnetic field as the fitness function of the genetic algorithm, and optimize the parameters of each magnetic dipole through the genetic algorithm.

[0011] Step 5: Formulate different simulation analysis experiment principles for the drag-type magnetoelectric simulation source and the new magnetic simulation source; combine the application of the future new electromagnetic simulation source to limit and optimize the analysis of the electromagnetic simulation source parameters, compare the calculated magnetic field with the target magnetic field, i.e., the measured magnetic field, and finally obtain a better solution considering the number of magnets and the magnetic field fitting effect.

[0012] Further, the method for establishing the three-dimensional distribution physical model of the magnetic dipole in the space magnetic field in Step 1 is as follows:

[0013] The following is the derivation of the analytical solution of the magnetic field distribution of the magnetic dipole in space:

[0014] Based on the basic knowledge of engineering electromagnetic fields, define the magnetic vector potential:

[0015]

[0016] According to the magnetic flux continuity theorem,

[0017]

[0018] According to the Weyl gauge,

[0019]

[0020] From (2-1) and (2-2),

[0021]

[0022] From the vector analysis of (2-4),

[0023]

[0024] From (2-3) and (2-5),

[0025]

[0026] That is,

[0027]

[0028] (0-28)

[0029] From the vector analysis of (2-7),

[0030]

[0031] Furthermore, the magnetic vector potential of the magnetic dipole is obtained:

[0032]

[0033] From (2-1) and (2-12),

[0034]

[0035] Further decomposition of (2-13) gives:

[0036]

[0037] As can be seen from formula (2-14), its analytical solution contains the components of the three-axis magnetic moment of the magnetic dipole in three directions, thus forming a 3×3 nine-variable matrix. Assuming the use of a rectangular coordinate system, a simple magnetic substance with a regular shape is decomposed into three three-axis magnetic dipoles. The magnetic fields of the three three-axis magnetic dipoles are expressed using the analytical solution and decomposed into X, Y, and Z components to obtain the analytical solution of the magnetic field of the most basic magnetic dipole. The explanations of the above parameter symbols are as follows:

[0038] B magnetic induction intensity, representing the force exerted by the magnetic field on a moving charge;

[0039] A vector potential, an auxiliary field quantity used to describe the spatial distribution of the magnetic field;

[0040] Gradient operator, used to calculate the local change rate of a scalar field or the rotation of a vector field;

[0041] J: Current density, describing the current intensity per unit volume or area;

[0042] μ 0 : Magnetic permeability, the magnetic permeability in vacuum, is a physical constant;

[0043] Magnetic dipole moment, describing the magnetic magnitude and direction of the magnetic dipole;

[0044] Position vector, representing the position from the magnetic dipole to the integration point;

[0045] r: Magnitude of the position vector, i.e., the distance from the magnetic dipole to the integration point;

[0046] ∮: Integration symbol for a closed path, indicating a line integral over a closed path;

[0047] ∫: Integration symbol, indicating an integration operation on a certain function;

[0048] x, y, z represent the coordinate positions of the magnetic dipole in three-dimensional space relative to the intersection of the midlines of the length and width of the ship's top view and the intersection of the water plane;

[0049] m x , m y , m z respectively represent the magnetic dipole moment vectors Components in the x, y, and z directions.

[0050] Furthermore, the magnetic fields of the three triaxial magnetic dipoles are represented by analytical solutions and decomposed into X, Y, and Z components. The specific method for obtaining the analytical solutions of the magnetic fields of the most basic magnetic dipoles is as follows:

[0051] 3.1 When the magnetic dipole moment is along the x direction, the three components of the magnetic field generated at the point (x, y, z) are respectively

[0052]

[0053] 3.2 When the magnetic dipole moment is along the y direction, the three components of the magnetic field generated at the point (x, y, z) are respectively

[0054]

[0055] 3.3 When the magnetic dipole moment is along the z direction, the three components of the magnetic field generated at the point (x, y, z) are respectively

[0056]

[0057] Furthermore, the magnetic moment magnitude variables in step three are divided into the magnetic moment magnitude variables of the drag-type electromagnetic simulation source and the new-type electromagnetic simulation source components.

[0058] Furthermore, the parameters for optimizing each magnetic dipole in step four include the number of magnetic dipoles, their positions, as well as the magnitude and direction of the magnetic moment.

[0059] Furthermore, in the selection of chromosomes for the genetic algorithm, the best individual preservation method is used to prevent the failure to preserve chromosomes with high fitness due to the large random selection error of the roulette wheel.

[0060] Furthermore, in step five, the method for formulating different simulation analysis experimental principles for the drag-type magnetoelectric simulation source and the new-type magnetic simulation source is as follows:

[0061] For the drag-type magnetic simulation source:

[0062] 1. Optimize from the number of magnets, decreasing from 10 to 3 in sequence to analyze the accuracy of the simulated magnetic field;

[0063] 2. Optimize from the direction of the magnets, from having only magnets in the X direction to having only magnets in the X and Z directions, and then to having magnets in the X, Y, and Z directions simultaneously;

[0064] 3. Combine the X and Z direction magnets into one magnet and conduct an optimization analysis by comprehensively considering the number of magnets and the direction of the magnets;

[0065] For the new-type electromagnetic simulation source:

[0066] Optimize the number of magnets from three to eight and analyze the optimization results by comparison.

[0067] The beneficial effects and features of the present invention are as follows:

[0068] (1) The ship magnetic field simulation method based on magnetic dipoles of the present invention sets a more accurate standard for a certain component of the magnetic simulation source, that is, the root mean square error curve drops to 80% from the best fitting effect of a certain group of experiments, on the basis of comprehensively considering simplicity and accuracy, and can simulate the ship magnetic field more accurately; it lays a foundation for the magnetic-electric simulation and optimization of a new magnetic simulation source for subsequent minesweepers or unmanned ships. Description of the Drawings

[0069] Figure 1 It is the curve of the Z-component of the simulated magnetic field when the number of magnets in the X direction of the preferred embodiment of the present invention ranges from 3 to 10;

[0070] Figure 2 It is the curve of the rms change of each component when the number of magnets in the X direction of the preferred embodiment of the present invention ranges from 3 to 10;

[0071] Figure 3 It is the curve of the X-component of the simulated magnetic field when the number of magnets in the X direction of the preferred embodiment of the present invention ranges from 3 to 10;

[0072] Figure 4 It is the comparison of the Z-component of the simulated magnetic field and the target magnetic field in the preferred embodiment of the present invention;

[0073] Figure 5 It is the comparison of the similarity of the magnetic field curves of the Z and Y components under the chord in the preferred embodiment of the present invention;

[0074] Figure 6 It is the curve of the rms change of each component when the number of magnets in the X direction ranges from 3 to 10 after adding two magnets in the Z direction in the preferred embodiment of the present invention;

[0075] Figure 7 It is the comparison of the fitting effect data before and after adding two Z-component magnets in the preferred embodiment of the present invention;

[0076] Figure 8 It is the curve of the rms change of each component when the number of magnets in the X direction ranges from 3 to 6 after adding two magnets in the Z and Y directions in the preferred embodiment of the present invention;

[0077] Figure 9 It is the comparison diagram of the curves of the Z and Y components under the chord of the actual ship and the curves of the Z and Y components under the chord of the simulated magnetic field in the preferred embodiment of the present invention;

[0078] Figure 10 It is the comparison of the magnetic field of the actual ship and the curve of the Y-component magnetic field under the keel of the simulated magnetic field in the preferred embodiment of the present invention;

[0079] Figure 11 It is a bar chart of fitting data when using four and five X - direction magnets before and after the integration of X and Z - direction magnets in the preferred embodiment of the present invention (each cluster of bar charts in the figure represents from left to right: 4X after integration, 5X after integration, 4X before integration, 5X before integration);

[0080] Figure 12 It is the change curve of each component rms when the number of magnets used in the new equipment of the preferred embodiment of the present invention changes from 3 to 8;

[0081] Figure 13 It is the comparison of the fitting performance among the new equipment, the improved current equipment and the current equipment in the preferred embodiment of the present invention (each cluster of bar charts in the figure represents from left to right: new equipment, improved existing equipment, existing equipment); Detailed implementation manners

[0082] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0083] Please refer to Figures 1 to 5 , the present invention relates to a method for simulating the magnetic field of a ship based on magnetic dipoles, including the following steps:

[0084] Step 1: Regard each magnet of the electromagnetic simulation source as a magnetic dipole, that is, a circular current, establish a three - dimensional distribution physical model of the magnetic dipole in the space magnetic field, and build a magnetic dipole simulation model based on this;

[0085] Step 2: Build a multi - magnetic - dipole simulation model to simulate the magnetic field generated by the magnetic simulation source through the coordinate transformation and magnetic field superposition of the magnetic dipole;

[0086] Step 3: Use the variable classification control and the single - variable principle for variable control modification; where variable classification control means dividing variables into magnetic moment magnitude variables and position variables; the single - variable principle means observing the change of simulation accuracy when only changing one or a class of variables. In step 3, the magnetic moment magnitude variables are divided into three component magnetic moment magnitude variables in the X, Y, and Z directions. For example, to experiment with the influence of the number of magnetic dipoles changing from three to ten on the fitting effect, then each experiment should only change the number of magnetic dipoles, and other variables should remain the same; if only studying the influence of the number of X - direction magnetic moments on the simulation effect, then each experiment should only change the number of X - direction magnetic dipoles, and the states of the Y - and Z - direction magnetic dipoles remain unchanged.

[0087] According to the research status at home and abroad, the number of magnets in existing towed magnetic simulation sources does not exceed ten, and there is no Y-component magnetic moment. There is only a small part of the Z-component magnetic moment, and most of them are X-component magnetic moments. Therefore, subsequent research on towed magnetic simulation sources has been improved based on the X-direction magnetic moment, adding an appropriate number of Z- and Y-direction magnetic moments; the number of new electromagnetic simulation sources is unknown, but considering the USV size and safety distance, the number will not exceed seven, and should be around five. The magnets can generate three-direction magnetic moments, and the new electromagnetic simulation sources should be analyzed with different numbers of magnets;

[0088] Step 4: On the basis of the principle established in Step 3, use the maximum error or root mean square error of the fitted magnetic field as the fitness function of the genetic algorithm, and optimize the parameters of each magnetic dipole through the genetic algorithm; specifically, optimizing the parameters of each magnetic dipole includes the number of magnetic dipoles, position, and the magnitude and direction of the magnetic moment.

[0089] Step 5: Develop different simulation analysis experimental principles for towed magnetoelectric simulation sources and new magnetic simulation sources; combine the application of future new electromagnetic simulation sources to limit and optimize the analysis of electromagnetic simulation source parameters, compare the calculated magnetic field with the target magnetic field, that is, the measured magnetic field, and finally obtain a better solution considering the number of magnets and the magnetic field fitting effect.

[0090] In the said Step 5, the method of developing different simulation analysis experimental principles for towed magnetoelectric simulation sources and new magnetic simulation sources is specifically as follows:

[0091] For towed magnetic simulation sources:

[0092] 1. Optimize from the number of magnets, decreasing from 10 to 3 in turn to analyze the accuracy of the simulated magnetic field;

[0093] 2. Optimize from the direction of the magnets, from only X-direction magnets to only X- and Z-direction magnets, and then to magnets with X-, Y-, and Z-directions simultaneously;

[0094] 3. Combine the X- and Z-direction magnets into one magnet, and comprehensively consider the number of magnets and the direction of the magnets for optimization analysis;

[0095] For new electromagnetic simulation sources:

[0096] Optimize from the number of magnets, with the number ranging from three to eight, and analyze by comparing the optimization results.

[0097] As a preferred solution, the method of establishing the three-dimensional distribution physical model of magnetic dipoles in the space magnetic field in Step 1 is as follows:

[0098] The following is the derivation of the analytical solution of the magnetic field distribution of magnetic dipoles in space:

[0099] From the basic knowledge of engineering electromagnetic fields, define the magnetic vector potential:

[0100]

[0101] According to the magnetic flux continuity theorem,

[0102]

[0103] According to the Weyl gauge,

[0104]

[0105] From (2-1) and (2-2), we get

[0106]

[0107] From the vector analysis of (2-4),

[0108]

[0109] From (2-3) and (2-5), we get

[0110]

[0111] That is,

[0112]

[0113] From the vector analysis of (2-7),

[0114]

[0115]

[0116] Furthermore, the magnetic vector potential of the magnetic dipole is obtained:

[0117]

[0118] From (2-1) and (2-12), we get

[0119]

[0120] Further decomposition from (2-13) gives:

[0121]

[0122] According to formula (2-14), its analytical solution includes the components of the three-axis magnetic moment of the magnetic dipole in three directions, thus forming a 3×3 nine-variable matrix. Assuming the use of a rectangular coordinate system, a simple magnetic substance with a regular shape is decomposed into three three-axis magnetic dipoles. The magnetic fields of the three three-axis magnetic dipoles are expressed using the analytical solution and decomposed into X, Y, and Z components to obtain the analytical solution of the magnetic field of the most basic magnetic dipole.

[0123] Among them, the magnetic fields of three three-axis magnetic dipoles are represented by analytical solutions and decomposed into X, Y, and Z components. The specific method for obtaining the analytical solutions of the magnetic fields of the most basic magnetic dipoles is as follows:

[0124] 3.1 When the magnetic dipole moment is along the x-direction, the three components of the magnetic field generated at the point (x, y, z) are respectively

[0125]

[0126]

[0127] 3.2 When the magnetic dipole moment is along the y-direction, the three components of the magnetic field generated at the point (x, y, z) are respectively

[0128]

[0129] 3.3 When the magnetic dipole moment is along the z-direction, the three components of the magnetic field generated at the point (x, y, z) are respectively

[0130]

[0131] In the selection of chromosomes in the genetic algorithm, the best individual preservation method is used to prevent the inability to preserve chromosomes with high fitness due to the large random selection error of the roulette wheel. In the early stage, a cyclic traversal magnetoelectric simulation analysis algorithm was tried, but it had the disadvantages of large computational complexity and slow operation. The algorithm logic of the cyclic traversal method was generated in the magnetic measurement practice. When measuring the magnetism of a ship model, the large plane measurement method was adopted. Since a magnetic sensor array was used to measure the spatial three-component magnetic field distribution, the distribution of measurement points was discrete and finite, and the obtained magnetic field data was also discrete and finite. When fitting the magnetic field, only these measurement points were used as the inspection points. In contrast, in the built magnetic dipole simulation model, only the three-component magnetism of each measurement point needed to be calculated, so the positions of the calculation points were determined, and the relatively small computational complexity was also conducive to the rapid implementation of the algorithm.

[0132] There are many variables in the magnetic dipole simulation model, including the number of magnetic dipoles, positions, and the magnitude and direction of the magnetic moments. Due to the limited computing power of the cyclic traversal algorithm, only the magnetic field generated by the magnetic moment along the X-axis was considered. First, the number of magnetic dipoles was determined. On this basis, the positions and magnitudes of the magnetic moments of each magnetic dipole were optimized. Here, the meaning of "optimization" is to loop through all possible values of the variable and calculate once to achieve a rough global search, and increase the constraint conditions of each variable according to the previous calculation results. It is a manual optimization. In theory, as long as the number of calculations is large enough, it is possible to achieve a transition from global search to local search and gradually approach the optimal solution.

[0133] In this algorithm, the matrix multiplication for solving overdetermined equations is not used to fit using the least squares method. There are multiple metrics for measuring the degree of magnetic field fitting, which is a multi-objective problem. In the case of single-objective optimization, there is only one objective, and any two solutions can be compared in terms of a single objective to obtain an uncontroversial optimal solution. However, multi-objective optimization is the opposite of traditional single-objective optimization. When multiple objectives need to be achieved in a certain problem, due to the easy existence of internal conflicts between objectives, the optimization of one objective comes at the expense of the degradation of other objectives. Therefore, it is very difficult to have a unique optimal solution. Instead, coordination and compromise are made among them to make the overall objective reach the optimal as much as possible. Since minimizing the root mean square error using the least squares method is not necessarily the best result, an algorithm similar to the evolutionary algorithm is used, that is, the optimal individual is preserved. The actual value of each measurement point corresponding to each case is subtracted from the calculated value, and the absolute values are summed. The parameters of the set of variables with the smallest absolute value are recorded. When further looping through later, the initial variable settings approach this parameter, and on this basis, the optimal solution is continued to be searched for.

[0134] When using this algorithm, it is very difficult to estimate the variable limit range in the first calculation. Due to the lack of reference, it is difficult to set the limit range of the magnetic moment of each magnetic dipole and the limit range of the position, that is, the loop range. A too large range will lead to an excessive amount of calculation and a decrease in the refinement degree of loop segmentation, and more calculations are required to gradually narrow the range; a too small range will result in the obtained optimal solution may only be a local optimal solution within the set range, and it is impossible to jump out of the local optimum to reach the global optimum.

[0135] Based on the above two-point analysis, when starting to use the loop traversal algorithm, it is necessary to perform a loop search over a large range. Even if the number of sampled calculation points is very small, it can be compensated by increasing the number of manual adjustments. The ranges of each variable are narrowed in a better direction, gradually searching from the global to the local.

[0136] The loop traversal algorithm is a preliminary attempt of the magnetoelectric simulation algorithm. Since the loop tries all possible values of all variables and calculates, in terms of algorithm implementation, it is loop within loop, and there are as many loops as there are variables; for MATLAB, the speed of the loop is much lower than matrix calculation. Therefore, the disadvantage is that when solving multi-variable problems, especially complex problems with more than 20 variables, this loop traversal method cannot run when the number of variables increases, and the amount of calculation will increase exponentially, and finally it cannot be calculated; in addition, as the loop refinement, that is, the strengthening of the local search intensity, the amount of calculation will also increase exponentially. In the actual application process of the algorithm, without particularly fine-segmented loops, a single loop calculation even takes forty minutes. Such a cumbersome approximation operation and huge amount of calculation are obviously not conducive to rapid analysis. In this case, it is necessary to refer to some optimization algorithms for solving multi-objective overdetermined equations.

[0137] Based on the principle of biological evolution, reproduction, and continuous adaptation to environmental changes in nature, the genetic algorithm has good global search ability and can avoid problems encountered in traditional algorithms. Since its proposal in the 1980s, various improvements have been made over the past decade. The disadvantage of this algorithm is its weak local search ability. For the same input parameters, the output optimal solutions are different. Although the output solutions are different, the optimal solution results are roughly the same, and the fitness values are similar, which can be accepted.

[0138] The main body of the genetic algorithm includes the following procedures: chromosome encoding, population initialization, chromosome fitness calculation, chromosome selection, chromosome crossover, and chromosome mutation. Since the genetic algorithm is improved for magnetoelectric simulation analysis and optimization, the improved program is called the magnetoelectric simulation optimization algorithm based on the improved genetic algorithm.

[0139] The magnetoelectric simulation optimization algorithm based on the improved genetic algorithm is similar to the traditional genetic algorithm in terms of chromosome encoding, population initialization, chromosome crossover, and mutation. The fitness calculation relies on the fitness function, which evaluates the fitness value of each chromosome. The best individual preservation will select the superior chromosomes and eliminate the inferior ones according to the fitness value. The fitness function is not fixed and changes according to the specific problems to be solved. For example, in this algorithm, the goal is to accurately fit the target magnetic field, so the fitness function is set as the maximum error or root mean square error of the fitted magnetic field, etc., which changes with the change of the evaluation index. In the selection of chromosomes, instead of using the traditional roulette wheel method, the best individual preservation method is used to prevent the inability to preserve chromosomes with high fitness due to the large random selection error of the roulette wheel.

[0140] The specific calculation examples are as follows:

[0141] 1. Starting from the structure and composition of the conventional towed electromagnetic simulation source, it is found that the limitation of the parameters of a single magnet lies in the direction of the magnetic moment of the magnet. The direction of the magnetic moment generated by a single magnet should be the X direction. After improvement, it can generate magnetic moments in both the X and Z directions simultaneously, but it is impossible to generate a stable magnetic moment in the Y direction. Considering the limitations of the above variables, through comprehensive analysis of the number of magnets and the direction of the magnets, and through horizontal and vertical comparisons, a better solution is obtained.

[0142] In actual application of the magnetic simulation source, the magnetic moments of the magnets are all in the X direction. Therefore, subsequent research is improved based on the magnetic moment in the X direction.

[0143] Analysis of the number of magnets, that is, the comparison and analysis of the ship magnetic simulation effects from 3 magnet simulation sources to 10 magnetic simulation sources. Taking the magnetic field curve obtained by visualizing the magnetic field data and the root mean square errors of the fitted magnetic field Z component (rmsZ), the fitted magnetic field X component (rmsX), and the fitted magnetic field Y component (rmsY) obtained by mathematical calculation analysis as the basis for comparison, the data is shown in Table 1:

[0144] Table 1 Fitting effect data of each component when the number of magnets in the X direction ranges from 3 to 10

[0145] Number of dipoles 3 4 5 6 7 8 9 10 absZ 216.21 146.85 123.58 114.84 98.13 97.04 93.12 95.22 relZ 41.19 29.07 25.25 24.14 23.56 23.54 23.5 23.49 rmsZ 87.33 61.62 53.53 51.17 49.95 49.9 49.82 49.81 absX 141.53 86.69 57.58 63.22 51.61 52.09 51.63 51.17 relX 44.91 26.84 19.62 18.5 17.32 16.92 16.97 16.92 rmsX 60.23 36 26.32 24.81 23.23 22.69 22.76 22.69 absY 142.31 129.75 123.23 125.1 120.21 117.22 117.3 117.2 relY 71.29 67.8 67.32 67.06 67.01 66.98 66.99 67.02 rmsY 51.67 49.14 48.79 48.6 48.57 48.55 48.56 48.58

[0146] First of all, and most importantly, is the fitting of the Z component of the ship's magnetic field. As Figure 1 , it can be visually observed after data visualization that as the number of magnetic moments in the X direction increases, the Z-component curve of the magnetic field simulated by the magnets becomes smoother and smoother. Starting from 5X, there is no obvious distortion in the curve, and it is impossible to use conventional methods to observe and distinguish whether it is the magnetic field of the electromagnetic simulation source.

[0147] As Figure 2 , after data comparison and analysis, it is found that when the number of X-direction magnets ranges from 3 to 10, rmsZ ranges from 87.33 nT to 49.81 nT. Plotting the rmsZ - number of X-component magnets curve and observing the rmsZ - number of X-component magnets curve, according to the previous relatively accurate definition, the relatively accurate standard for the rmsZ of this set of data of the magnetic simulation source is to drop to 57.31 nT. It can be found that the magnitude of rmsZ just reaches the relatively accurate standard when the number of X-direction magnets increases to 5. Thus, it can be seen that when only using X-direction magnets, after the number reaches 5, increasing the number of X-direction magnets has little effect on the fitting effect of the Z-component magnetic field. This is also the reason why most electromagnetic simulation sources in various countries currently use 5 magnets.

[0148] Secondly, analyze the fitting of the X component of the ship's magnetic field. As Figure 3 , observing the X-component curve of the magnetic field simulated by the magnets, it is found that the distortion of the X-component magnetic field curve under the keel is very obvious. Below 6X, it can be clearly seen that the curve is not smooth and oscillates. Compared with the Z-component curve, its fluctuation is greater, and it can be considered that its requirement for the number of magnets is higher than that for the fitting of the Z-component magnetic field to the number of magnets. And observing the influence of the increase in the number of magnets on rmsX, according to the relatively accurate definition, the relatively accurate standard for the rmsX of this set of data of the magnetic simulation source is to drop to 30.20 nT. It can be found that the magnitude of rmsX just reaches the relatively accurate standard when the number of X-direction magnets increases to 5. It can also be clearly found that after increasing the number of X-direction magnets to 5, further increasing the number of X-direction magnets has a very poor effect on reducing the fitting error of rmsX, further verifying the simplicity and relative accuracy of using 5 X-direction magnets in actual combat.

[0149] Finally, analyze the fitting of the Y component of the ship's magnetic field. First, observe the curve of the Y component under the keel of the magnetic field simulated by the magnet. It can be found that regardless of the number of magnets in the X direction, the magnetic magnitude is always zero, which is also consistent with the conclusion of the spatial magnetic field characteristics of the magnetic moment in the X direction. However, in fact, the magnetic field of the Y component under the ship's keel is not zero. Therefore, this is a defect of the electromagnetic simulation source magnetic field. Then, look at the change of the fitting error of the Y component of the ship's magnetic field with the increase in the number of magnets in the X direction. It is found that rmsY is basically a straight line and the value is very large, which further shows that the increase in the number of X-component magnets has no effect on the fitting of the Y-component magnetic field of the ship.

[0150] 2. Consider the optimization results of the magnetic dipole moment direction

[0151] Summarize the problems found by only using the magnetic moment in the X direction. The first problem is that the magnetic field of the Z component of the ship simulated is always smaller than the target magnetic field as a whole, with an overall bias, as Figure 5 .

[0152] Considering that only the magnets in the Z direction can generate a positive magnetic field in a large range, it is decided to add magnets in the Z direction in the subsequent experiments. The second problem is that the curve of the X component of the magnetic field simulated by the magnet is also not smooth, but it gradually improves with the increase in the number of magnets in the X direction; the third problem is that the curve of the Y component under the keel simulated by the magnet is always zero, which is a defect of the magnetic field simulated by the magnet. From the very large rmsY, it can be seen that the fitting effect of the Y component of the magnetic field simulated by the magnet is very poor. At the same time, after comparing the experimental data many times, it is found that the curves of the Z component magnetic field and the Y component magnetic field of the magnetic field simulated by the magnet at the same position are very similar, as Figure 4 , using a dual-axis graph to draw the two curves on one graph. The difference is that the range of the left Y-axis is 2.085 times that of the right Y-axis, and the two are in a multiple relationship, which does not conform to the actual ship's magnetic field. This is also a defect of the Y component of the magnetic field simulated by the magnet.

[0153] Regarding the problem that the magnetic field of the Z component of the ship simulated is always smaller than the target magnetic field as a whole, considering that it is difficult to arrange the magnets of the Z component alone and the total number of magnets cannot be too large, therefore, the total number of magnetic moments is limited to less than 10. And to meet the requirements of simplicity and economy, only two separate Z-component magnets are arranged to establish a model for discussion, and the fitting error data are as follows.

[0154] Table 2 Fitting effect data of each component when the number of X-direction magnets ranges from 3 to 8 after adding two Z-direction magnets

[0155] Number of X dipoles 3 4 5 6 7 8 absZ 138.5 90.79 71.58 70.53 66.55 63.42 relZ 28.32 18.33 14.94 14.07 12.66 12.62 rmsZ 60.04 38.86 31.68 29.84 26.83 26.75 absX 110.4 66.08 65.98 56.06 50.01 45.3 relX 33.74 20.37 16.79 15.17 16.13 15.78 rmsX 45.25 27.32 22.52 20.35 21.63 21.16 absY 125.96 111.79 100.35 101.29 100.4 98.32 relY 80.24 75.09 66.17 67.22 64.54 65.09 rmsY 58.15 54.43 47.95 48.72 46.78 47.17

[0156] The data is analyzed using the same method as in the previous section. First, a horizontal comparison is made. According to a more accurate definition, the more accurate standard for the magnetic simulation source rmsZ of this group of data is to drop to 33.41 nT, the more accurate standard for rmsX is to drop to 25.98 nT, and the more accurate standard for rmsY is to drop to 49.37 nT. As Figure 6 , it can be found that as the number of magnets in the X direction increases, rmsX, rmsY, and rmsZ also increase to reach the more accurate standard when there are 5 magnets in the X direction. At this time, it is the simplest and has a better effect. After that, adding more magnets in the X direction does not help much in improving the fitting accuracy. Then, a vertical comparison is made. As Figure 7 , after adding two Z-component magnets and comparing with the case where there were only X-direction magnets before, it is found that the fitting effect rmsZ of the Z component of the ship's magnetic field has been greatly improved. Only 6 magnets, namely 4 X-direction magnets and 2 Z-direction magnets, are used to achieve a fitting effect that cannot be achieved with 10 X-direction magnets. This also shows that when using X- and Z-direction magnets simultaneously, the effect of simulating the Z component of the ship's magnetic field is better than using only X-direction magnets. Then, analyzing the fitting of the simulated magnetic field X and Y components, it is found that the value of rmsX decreases slightly, indicating that adding two Z-component magnets is helpful for fitting the X component of the ship's magnetic field. However, it can also be found that the rmsY curve almost coincides and the value remains high, indicating that it is of no help for fitting the Y component of the ship's magnetic field, and the fitting effect of the Y component has always been poor.

[0157] Regarding the third problem, the curve of the Y component under the keel simulated by the magnet is always zero, the magnetic field fitting degree is low, and the similarity between the curve of the Z component and the Y component of the magnetic field simulated by the magnet is high and does not match the actual magnetic field of the ship. Analyzing from a theoretical perspective, this is because the magnetic moments of most magnets are in the X direction, there is no Y component in the magnetic field directly below, and the Z-component magnetic field and the Y-component magnetic field are symmetric with respect to the X axis, so they are similar, which explains the above problems. The root cause of the problem is that the directions of the magnets cannot all be in the X direction, and interference can be introduced by adding magnetic moments perpendicular to the X axis. Therefore, it is decided to add Y-direction magnets in the next experiment. Observing the Y-component magnetic field under the ship's keel, it is found that its curve is very simple, and it can be simply simulated with 3 or even 2 Y-direction magnets. Starting from the principle of simplicity that the fewer the actual number of magnets used, the better, only 2 Y-direction magnets are used for analysis in the next experiment, and the data obtained is shown in Table 3.

[0158] Table 3 Fitting effect data of each component when the number of X-direction magnets ranges from 3 to 6 after adding two Y- and Z-direction magnets

[0159]

[0160]

[0161] First, a horizontal comparison is made. After adding two Z-direction magnets and two Y-direction magnets, as the number of X-direction magnets increases from 3 to 6, according to a more accurate definition, the more accurate standard for the magnetic simulation source rmsZ of this group of data is to drop to 26.95 nT, the more accurate standard for rmsX is to drop to 27.27 nT, and the more accurate standard for rmsY is to drop to 30.00 nT. Observe Figure 8 It can be found that the fitting error of the three components of the magnetic field slows down when using 4 X-direction magnets, and fully reaches the more accurate standard after using 5 X-direction magnets; then a vertical comparison is made. Compared with only using X- and Z-direction magnets, using X-, Y-, and Z-direction magnets has an effect on the fitting of the Z component of the ship's magnetic field, and rmsZ drops by nearly 10 nT, but the two rmsX curves basically coincide, indicating that there is no effect on the fitting of the X component of the ship's magnetic field; the rmsY curve moves significantly downward, rmsY decreases by nearly 20 nT, and the relative error of the fitted magnetic field of the Y component (relY) drops by nearly 30%, indicating that adding two Y-direction magnets has an obvious effect on the fitting of the Y component of the ship's magnetic field. At the same time, as Figure 9 shown, the magnetic field of the Y component under the keel generated by the magnets is no longer a straight line, and as Figure 9 shown, the similarity between the curves of the Z component and the Y component under the chord of the magnetic field simulated by the magnets is broken. After repeated comparison with the magnetic field measured by the actual ship, it is found that the relationship between the two is closer to the actual magnetic field of the ship, which not only makes up for the defects of the magnetic field of the magnetic simulation source, but also further improves the level of simulating the magnetic field of the ship.

[0162] 3. Optimization results obtained by comprehensively considering the number of magnetic dipoles and the direction of magnetic moments

[0163] Based on the above analysis, the current optimal way of using magnets is obtained, that is, 5 X-direction magnets, 2 Z-direction magnets, and 2 Y-direction magnets. However, this requires 9 magnets, which is nearly twice as many as those used in the current magnetic simulation source. Although the magnetic field fitting effect is very good, it does not meet the requirements of practicality and simplicity. To solve the problem of too many magnets, the idea of generating multiple-direction magnetic moments with a single magnet is proposed. Considering the actual application of the current towed electromagnetic simulation source, it is possible to generate X- and Z-direction magnetic moments simultaneously if a single magnet is improved, but it is more difficult to generate Y-direction magnetic moments, and generally only separate magnets can be used to generate them. Therefore, 5 X-component magnetic moments and two Z-component magnetic moments are integrated into 5 magnets, and two Y-component magnetic moments are specially generated by two magnets. The improved experimental data are shown in Table 4.4.

[0164] Table 4 Fitting effect data of each component when the number of X-direction magnets is 4 and 5 after integrating X- and Z-direction magnets

[0165] Number of X absZ relZ rmsZ absX relX rmsX absY relY rmsY 4 96.25 18.73 39.71 93.31 25.41 34.08 69.4 35.74 25.9 5 67.37 11.09 23.5 65.82 17.07 22.9 59.31 35.23 25.53

[0166] The experimental results are longitudinally compared with the case where the X and Z magnets are not integrated, as Figure 11 From the comparison of the two groups of data, it can be seen that for the experimental group using 4 X-direction magnetic moments, the accuracy of the simulated magnetic field slightly decreases. However, for the experimental group using 5 X-direction magnetic moments, not only does it not decrease, but it even improves the accuracy. It can be seen that integrating the X and Z magnets together reduces the total number of magnets and can relatively well maintain the accuracy of the simulated magnetic field. Such a design is completely feasible.

[0167] Applicable to the experimental data analysis of the electromagnetic simulation source for new surface unmanned equipment:

[0168] Considering the current usage method of the new magnetic simulation source, it is necessary to integrate the X, Y, and Z component magnetic moments into one magnet. For example, integrating 5 X-component magnetic moments, 5 Z-component magnetic moments, and 5 Y-component magnetic moments into 5 magnets to achieve the effect that one magnet can stably generate three-component magnetic moments. Due to the limitations of the size, safety distance, and deployment range of the unmanned ship in actual operation, the total number of magnets is limited to within 8. In the specific experiment, the data fitted from 3 magnets to 8 magnets were obtained, and the experimental data are shown in Table 5.

[0169] Table 5 Fitting effect data of each component when the number of magnets used in the new equipment ranges from 3 to 8

[0170]

[0171]

[0172] First, conduct a horizontal comparison. According to a more accurate definition, the more accurate standard for the rmsZ of the magnetic simulation source in this group of data is to drop to 26.61 nT, the more accurate standard for rmsX is to drop to 27.36 nT, and the more accurate standard for rmsY is to drop to 25.07 nT.

[0173] From Figure 12 it can be seen that rmsX, rmsY, and rmsZ all decrease with the increase in the number of magnets, and they fully reach the more accurate standard after increasing to 5 magnets. The curve descent rate significantly decreases, indicating that in the case of using the new electromagnetic simulation source, 5 magnets are the optimal solution considering simplicity and accuracy, meeting the expected goal.

[0174] Compare the final optimized structure of the new electromagnetic simulation source with the final optimized structure of the traditional towed electromagnetic simulation source and the structure of the current domestic towed electromagnetic simulation source, and analyze their performance differences, as Figure 13 shown.

[0175] It can be found that the fitting effects of the final optimized structure of the new electromagnetic simulation source and the final optimized structure of the traditional towed electromagnetic simulation source on each component of the ship's magnetic field are roughly the same, and the fitting effect of the final optimized structure of the new electromagnetic simulation source is slightly better. However, compared with the current domestic towed electromagnetic simulation source, the optimized effect produced by the final optimized structure of the solution in this embodiment is twice as strong, and the simulation effect can be called a leap.

[0176] Those skilled in the art can easily understand that the above are only preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention should be included within the protection scope of the present invention.

Claims

1. A ship magnetic field simulation method based on magnetic dipole, characterized in that: The steps include: Step 1: Treat each magnet of the electromagnetic simulation source as a magnetic dipole, that is, a circular current, and establish a three-dimensional distribution physical model of the magnetic dipole in the spatial magnetic field, and build a magnetic dipole simulation model based on this; Step 2: Through the coordinate transformation of magnetic dipoles and the superposition of magnetic fields, a multi-magnetic dipole simulation model is built to simulate the magnetic field generated by the magnetic simulation source; Step 3: Modify variable control using variable classification control and single variable principle; The variable classification control refers to dividing the variables into magnetic moment size variables and position variables; the single variable principle refers to observing the change of simulation accuracy when only one or one type of variable is changed; Step 4: Based on the principle established in step 3, the maximum error or root mean square error of the fitting magnetic field is used as the fitness function of the genetic algorithm, and the parameters of each magnetic dipole are optimized by the genetic algorithm; Step 5: Formulate different simulation analysis experimental principles for towed magnetoelectric simulation sources and new magnetic simulation sources; limit and optimize the electromagnetic simulation source parameters in combination with the application of future new electromagnetic simulation sources, compare the calculated magnetic field with the target magnetic field, that is, the measured magnetic field of the ship, and finally obtain a better solution that takes into account the number of magnets and the magnetic field fitting effect.

2. The ship magnetic field simulation method based on magnetic dipole according to claim 1 is characterized in that: The method for establishing the physical model of the three-dimensional distribution of magnetic dipoles in the spatial magnetic field in the step 1 is as follows: the analytical solution of the magnetic field distribution of magnetic dipoles in space is derived as follows: Based on the basic knowledge of engineering electromagnetic fields, the magnetic vector potential is defined as: According to the flux continuity theorem, According to Weyl norm From (2-1) and (2-2) we get From the vector analysis of (2-4), we can get From (2-3) and (2-5) we get Right now From the vector analysis of (2-7) we can get Then we get the magnetic vector potential of the magnetic dipole: From (2-1) and (2-12) we get Further decomposition of (2-13) yields: It can be seen from formula (2-14) that its analytical solution includes the components of the three-axis magnetic moment of the magnetic dipole in three directions, thus forming a 3×3 nine-variable matrix. Assuming a rectangular coordinate system, the simple magnetic material with a regular shape is decomposed into three three-axial magnetic dipoles. The magnetic field of the three three-axial magnetic dipoles is represented by an analytical solution and decomposed into three components, X, Y, and Z, to obtain the most basic magnetic dipole magnetic field analytical solution, where the above parameter symbols are explained as follows: B Magnetic induction intensity, which indicates the force exerted by the magnetic field on moving charges; A vector potential, an auxiliary field quantity used to describe the spatial distribution of the magnetic field; Gradient operators, which are used to compute the local rate of change of a scalar field or the rotation of a vector field; J: current density, which describes the current intensity per unit volume or area; μ0: magnetic permeability, the magnetic permeability in a vacuum, is a physical constant; The magnetic dipole moment, which describes the magnitude and direction of the magnetic properties of a magnetic dipole; The position vector represents the position from the magnetic dipole to the integration point; r: the magnitude of the position vector, i.e. the distance from the magnetic dipole to the integration point; ∮: The integral symbol of a closed path, indicating the line integral performed on a closed path; ∫: integral symbol, indicating the integration of a function; x, y, z represent the coordinates of the magnetic dipole in three-dimensional space relative to the intersection of the length and width midline of the ship's top view and the horizontal plane; m x ,m y ,m z Denote the magnetic dipole moment vector Components in the x, y, and z directions.

3. The ship magnetic field simulation method based on magnetic dipole according to claim 2 is characterized in that the magnetic field of the three triaxial magnetic dipoles is represented by an analytical solution and decomposed into three components, X, Y, and Z, and the specific method of obtaining the most basic magnetic dipole magnetic field analytical solution is as follows: 3.1 When the magnetic dipole moment is along the x direction, the three components of the magnetic field generated at the point (x, y, z) are 3.2 When the magnetic dipole moment is along the y direction, the three components of the magnetic field generated at the point (x, y, z) are 3.3 When the magnetic dipole moment is along the z direction, the three components of the magnetic field generated at the point (x, y, z) are 4. The ship magnetic field simulation method based on magnetic dipole according to claim 1 is characterized in that: In the step three, the magnetic moment size variables are divided into drag-type electromagnetic simulation source and new electromagnetic simulation source component magnetic moment size variables.

5. The ship magnetic field simulation method based on magnetic dipole according to claim 1 is characterized in that: The parameters of each magnetic dipole optimized in step 4 include the number, position, and magnitude and direction of the magnetic moment of the magnetic dipole.

6. The ship magnetic field simulation method based on magnetic dipole according to claim 1 is characterized in that: The genetic algorithm uses the best individual preservation method in the selection of chromosomes.

7. The method for simulating a ship's magnetic field based on magnetic dipoles according to any one of claims 1 to 6, characterized in that: In step 5, the method of formulating different simulation analysis experimental principles for the towed magnetoelectric simulation source and the new magnetic simulation source is specifically as follows: For dragged magnetic simulation sources:

1. Optimize the number of magnets, and analyze the accuracy of the simulated magnetic field in descending order from 10 to 3; 2. Optimize the direction of the magnet, from only X-direction magnets to only X and Z-direction magnets, and then to X, Y, and Z-direction magnets at the same time; 3. Combine the X and Z direction magnets into one magnet, and conduct optimization analysis by comprehensively considering the number of magnets and the direction of the magnets.

8. The ship magnetic field simulation method based on magnetic dipole according to claim 7 is characterized in that: Also included are experimental principle methods for novel electromagnetic simulation sources: The number of magnets is optimized from three to eight, and the optimization results are compared and analyzed.

Images