Vehicle-mounted vector magnetic compensation method based on dot product invariant constraint

Abstract

The invention provides a vehicle-mounted vector magnetic compensation method based on dot product invariant constraint, and the method comprises the steps: constructing a magnetic interference model of a vehicle carrier, building an error equation of measurement interference, and achieving the compensation of the related errors of the magnetic vector measurement of the vehicle carrier; calculating a rotation matrix from the earth system to the carrier system, wherein the rotation matrix is used for constructing an auxiliary vector; constructing a constraint according to the auxiliary vector, and constructing a vector compensation parameter solving equation according to a dot product invariant constraint; according to the vector compensation parameter solving equation, carrying out vector compensation parameter solving based on an improved genetic algorithm; parameter optimization is carried out based on residual Gaussian process regression, it is assumed that a vector geomagnetic field meets a Gaussian regression model, and the geomagnetic field gradient is estimated through Gaussian regression according to a Maxwell equation set by utilizing the physical principle that the geomagnetic field vector is the gradient of scalar magnetic potential under the non-rotation condition and considering that internal physical relation exists among three axes of the vector geomagnetic field. According to the invention, various errors including installation errors can be compensated.

Description

Technical Field

[0001] The invention relates to the field of magnetic compensation of magnetic field vector measurement, and in particular to a vehicle-mounted vector magnetic compensation method based on dot product invariant constraints. Background Art

[0002] Vector magnetic measurement plays an important role in the field of geomagnetic navigation. Geomagnetic navigation can be used as a supplement to navigation in a satellite-based environment, and has many advantages such as strong anti-interference ability and good concealment. The existing mainstream geomagnetic navigation technology is mainly based on scalar magnetic navigation. Vector magnetic data contains more dimensional information than scalar magnetic data. Vector geomagnetic navigation is more robust and is one of the important development directions of geomagnetic navigation technology in the future.

[0003] As a matching navigation technology, geomagnetic navigation requires geomagnetic maps and real-time magnetic survey information. High-precision magnetic survey technology is essential, but the construction of geomagnetic maps requires the use of carriers carrying magnetic survey equipment for collection and mapping. Since the carriers inevitably contain magnetic materials and the sensor processing technology is limited, the magnetic survey data will be affected by errors. If magnetic compensation is not performed, the magnetic survey accuracy is not high, and the positioning accuracy of geomagnetic navigation will also decrease. Magnetic compensation can effectively improve the magnetic survey accuracy, thereby improving the positioning accuracy of geomagnetic navigation.

[0004] Vector geomagnetic maps can be roughly divided into aeromagnetic maps, ground magnetic maps and ocean magnetic maps according to the different application carriers. The collection of these magnetic maps requires corresponding carrier compensation technology. Although there are some magnetic compensation technologies for aircraft carriers and ship carriers, there are few vector compensation technologies for vehicle carriers. At the same time, the richness of ground magnetic features is far greater than that of high altitude and ocean. At the same time, the attitude of the vehicle is limited, and it is difficult to maneuver independently in pitch and roll, making vector magnetic compensation for vehicle carriers more difficult.

[0005] The classic method for magnetic vector compensation is the ellipsoid fitting method, which uses the unchanged modulus of the geomagnetic field to solve the compensation parameters. Specifically, the ellipsoid fitting method requires the carrier to maneuver in various postures, considers the error model of the collected data to construct the ellipsoid equation, and calculates the compensation parameters by solving the ellipsoid equation.

[0006] The ellipsoid fitting method is limited by the ellipsoid equation and cannot solve the installation error. The compensation effect of the soft magnetic error is limited by the actual situation of the carrier, and the compensation effect is unstable. In addition, the ellipsoid fitting method has high requirements on the angle of carrier traversal. The vehicle carrier can only travel on the road, the carrier posture is limited, and the parameter solution is seriously pathological. The ellipsoid fitting method assumes that the background magnetic field is constant. Considering that the vehicle carrier needs a certain amount of maneuvering space for posture traversal, the assumption of a constant background magnetic field may not hold. Summary of the invention

[0007] In order to solve the problems of installation error angle compensation lack, parameter solution pathology and background magnetic field instability in the process of measuring the vector magnetic field of the vehicle carrier, the present invention provides a vehicle-mounted vector magnetic compensation method based on dot product invariance constraint, which takes the geomagnetic field vector measured by the three-axis magnetic sensor collected at the calibration site according to certain requirements and the vehicle attitude angle calculated by high-precision inertial navigation measurement as input, requires good time synchronization between the geomagnetic field vector and the attitude angle, uses a comprehensive magnetic interference model to establish interference compensation parameter constraints, uses the MOIRCGA algorithm to solve the compensation parameters, and further realizes parameter optimization through residual Gaussian process regression. The magnetic compensation algorithm proposed by this method can improve the accuracy of the geomagnetic vector after compensation and can meet the requirements of general geomagnetic navigation.

[0008] In order to achieve the above object, the present invention adopts the following technical solution: A vehicle-mounted vector magnetic compensation method based on dot product invariance constraint comprises the following steps: Step 1: construct a magnetic interference model of the vehicle carrier, establish an error equation for measuring interference through the magnetic interference model and constraints, and realize compensation for errors related to magnetic vector measurement of the vehicle carrier; Step 2: Calculate the rotation matrix from the earth system to the load system. The rotation matrix is ​​used to construct the auxiliary vector. Step 3, construct constraints based on the auxiliary vector of step 2, and construct vector compensation parameter solution equations based on the dot product invariant constraints; Step 4, solving the equation according to the vector compensation parameter, and solving the vector compensation parameter based on the improved genetic algorithm; Step 5: Parameter optimization is performed based on residual Gaussian process regression. It is assumed that the vector geomagnetic field satisfies the Gaussian regression model. Considering the intrinsic physical connection between the three axes of the vector geomagnetic field, according to Maxwell's equations, the geomagnetic field gradient is estimated using Gaussian regression based on the physical principle that the geomagnetic field vector is the gradient of the scalar magnetic potential in the irrotational case.

[0009] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the above-mentioned vehicle-mounted vector magnetic compensation method based on dot product invariant constraint are implemented.

[0010] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-mentioned vehicle-mounted vector magnetic compensation method based on dot product invariant constraints.

[0011] Beneficial effects: 1. The ellipsoid equation has only 9 parameters, which causes the compensation parameters solved by the ellipsoid fitting method to differ from the actual compensation parameters by a rotation matrix. This results in the method having no ability to compensate for installation errors. The present invention applies the dot product invariant constraint to the field of vehicle carrier vector magnetic compensation. There are 12 parameters in the dot product invariant constraint, which can cover various interference forms such as soft magnetism and hard magnetism, and therefore can compensate for various errors including installation errors.

[0012] 2. The ellipsoid fitting has high requirements on posture. The present invention introduces additional information by using the dot product invariant constraint, improves the robustness of compensation, and reduces the requirements on posture. On the other hand, insufficient posture leads to pathological parameter solution. The present invention introduces an improved genetic algorithm to solve parameters, and uses fitting rather than iteration to overcome the influence of parameter pathology.

[0013] 3. After considering the gradient of the background magnetic field, the present invention performs secondary optimization on the compensation parameters, which can make the compensation results closer to the actual situation. Many current methods do not consider the influence of the background magnetic field gradient change, which is real and will affect the accuracy of parameter solution. Therefore, considering the background magnetic field gradient can improve the magnetic compensation effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a flow chart of a vehicle-mounted vector magnetic compensation method based on dot product invariant constraint of the present invention; Figure 2 A comparison diagram of the compensation results of the vehicle-mounted vector magnetic compensation method of the present invention and the ellipsoid fitting method (x-axis); Figure 3 A comparison diagram of the compensation results of the vehicle-mounted vector magnetic compensation method of the present invention and the ellipsoid fitting method (y-axis); Figure 4 Shown is a comparison diagram of the compensation results of the vehicle-mounted vector magnetic compensation method of the present invention and the ellipsoid fitting method (z axis). DETAILED DESCRIPTION

[0015] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0016] In the description of the present invention, it should be understood that the terms "upper", "lower", "top", "bottom", "inner", "outer", etc. used in the present invention to indicate the orientation or position relationship is based on the orientation or position relationship shown in the drawings, which is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0017] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0018] like Figure 1 As shown, a vehicle-mounted vector magnetic compensation method based on dot product invariant constraint of the present invention comprises the following steps: Step 1: Construct a magnetic interference model of the vehicle carrier. Through the model and related constraints, the error equation of the measurement interference can be established to compensate for the related errors of the magnetic vector measurement of the vehicle carrier, including: The magnetic interference model of the vehicle carrier is: ; Therefore, the calibration model is ,in, , ; and That is the parameter required for compensation, by calculating , The specific value of the measured magnetic vector Compensation for true magnetic vector .

[0019] in, is the projection vector value of the geomagnetic field in the carrier coordinate system, and the three-axis values ​​are ; is the measured magnetic vector value in the carrier coordinate system, and the three-axis values ​​are ; is the misalignment error matrix between the measurement coordinate system of the magnetic sensor and the carrier coordinate system; is the magnetic sensor sensitivity error matrix; is the non-orthogonal error matrix of the magnetic sensor; is the soft magnetic error matrix of the vehicle carrier; is the hard magnetic error vector of the vehicle carrier; is the zero bias error vector of the magnetic sensor.

[0020] Unless otherwise specified, the geodetic coordinate system of the present invention is The Northeast Sky coordinate system, the carrier coordinate system It is the upper right front coordinate system.

[0021] Step 2: Calculate the rotation matrix from the earth system to the load system. The rotation matrix is ​​used to construct the auxiliary vector, including: Auxiliary vectors are constant vectors of the artificially selected geodetic coordinate system. , the vehicle attitude angle is measured as the yaw angle , pitch angle , roll angle , then the rotation matrix from the earth system to the load system is: Rotation matrix for rotation around yaw angle for: ; Rotation matrix for rotation around pitch angle for: ; Rotation matrix for rotation around the roll angle for: ; Total rotation matrix for:

[0022] Then the auxiliary vector in the carrier coordinate system is .

[0023] Step 3: Based on the auxiliary vector of step 2, construct a vector compensation parameter solution equation according to the dot product invariant constraint, including: Note that the auxiliary vector and the geomagnetic field vector are both constant vectors in the geodetic coordinate system, so the dot product of the two should be a fixed value, and this fixed value is valid in any rectangular coordinate system, that is: ; The superscript T represents the transpose of the matrix. Represents a fixed constant.

[0024] Expand it according to the carrier coordinate system components, let , , have , where i, j represent the subscripts of the elements in each variable, i, j = 1, 2, 3. If enough and sufficient data are collected, the calculation can be done. and The specific value of .

[0025] Specifically, step 3 includes the following steps: Step 3.1 collects calibration data, wherein the calibration data includes magnetic vector data and posture data; Step 3.2 constructs an auxiliary vector based on the pose data; Step 3.3 calculates the dot product result of the auxiliary vector and the magnetic vector data; Step 3.4 Use the constraint that the dot product result remains unchanged to construct the equation to solve the compensation parameters , ; Step 3.5 According to the formula You can get the real magnetic data after rough compensation .

[0026] Step 4: Calculate the vector compensation parameters based on the improved genetic algorithm to solve the compensation parameter equation, including: Taking into account the limited mobility of vehicle carriers, it is difficult to collect sufficient data and the parameter solution is ill-conditioned. The MOIRCGA (multi-offspring improved real coded genetic) algorithm is selected to solve the compensation parameters. This algorithm is an improvement on the genetic algorithm, which enhances the convergence efficiency and search ability of the algorithm and can overcome the problem of ill-conditioned solution to a certain extent.

[0027] The input of the algorithm is the population size , mutation probability , crossover probability , variable upper and lower bounds , number of elites , crossover probability Generally set to 1.

[0028] The initial population individuals are set as and , a total of 12 compensation parameters, solve the individual that minimizes the objective function, the objective function f is set to Note that this assumes is 1, so the solution is and Just relative values, add constraints To fix the absolute value of the parameter.

[0029] Specifically, step 4 includes the following steps: Step 4.1 Initialize the population; Step 4.2 calculates the fitness function of the individuals and sorts them; Step 4.3 determines whether the cut-off condition is met, if not, go to step 4.4, if met, go to step 4.14; Step 4.4 Perform sorted group selection (SGS) and select parental pairs; Step 4.5 uses the heuristic normal distribution and direction-based crossover operator (HNDDBX). Specifically, the results after sorting and grouping are and ,generate offspring individuals; Step 4.6 performs a replacement operation to randomly generate new individuals to replace duplicate individuals in the offspring population; Step 4.7 calculates the fitness function of the individuals and sorts them; Step 4.8 From the parent and The offspring individuals are retained The elite individuals with the highest fitness; Step 4.9 executes a combined mutation algorithm of the comprehensive Cauchy mutation operator, Gaussian mutation operator, and Levy mutation operator; specifically, the algorithm executes the corresponding mutation operator on the mutant individuals according to the result of the iteration round modulo 3. The Cauchy mutation operator is , the Gaussian mutation operator is , the Levy mutation operator is ,in For the individual after mutation, For the individual to be mutated, is the standard Cauchy distribution, is a Gaussian distribution, is the mean of the Gaussian distribution, is the variance of the Gaussian distribution, For Levi's flight path, , These are all artificially set parameters; Step 4.10 calculates the fitness function of the individuals and sorts them; Step 4.11 makes An elite individual, mutant individuals and Non-mutant individuals make up the current population; Step 4.12 Select The best individuals form the central group, retaining the new population an elite individual; Step 4.13 Go to step 4.3; Step 4.14 outputs the average running time, average number of iterations and optimal individual value.

[0030] Step 5: Parameter optimization based on residual Gaussian process regression, including: In the assumption of the dot product invariance method, the geomagnetic field vector is used to construct an equation to solve the compensation parameters. However, considering the limitation of the vehicle carrier, the spatial range of data acquisition is large, and the geomagnetic field vector has a spatial gradient. It is a very constant vector in the geodetic coordinate system. This constraint condition is not satisfied. Therefore, it is necessary to correct the geomagnetic field spatial gradient. The present invention assumes that the vector geomagnetic field satisfies the Gaussian regression model. Considering the inherent physical connection between the three axes of the vector geomagnetic field, according to Maxwell's equations, the geomagnetic field vector is the gradient of the scalar magnetic potential under the irrotation condition. This method uses Gaussian regression to estimate the geomagnetic field gradient, including: Step 5.1. Build a Gaussian regression model: ; in, is the scalar magnetic potential, represents a Gaussian random process, is the kernel function of the Gaussian process, and the specific form is selected according to the actual situation. is the spatial position coordinate, is the geomagnetic field vector value in the geodetic coordinate system. If there is no special explanation, the geomagnetic field involved in step 5 is the geomagnetic field value in the geodetic coordinate system, and the subscript i represents a specific location, that is, is the coordinate of position i, is the magnetic field at position i; represents the gradient, For noise and , N represents Gaussian distribution, is a noise-related hyperparameter, represents a three-dimensional unit matrix; Specifically, the input data selects the vector magnetic data and the corresponding spatial coordinates after the rough compensation of the calibration data, and the output data is the vector magnetic data of the predicted position that conforms to the Gaussian regression model; Step 5.2. Perform maximum likelihood estimation on the hyperparameters involved in the Gaussian regression model to obtain the optimal hyperparameters. The maximum likelihood function for: ; in, is the hyperparameter vector, The subscript represents the hyperparameter value of the kernel function. ,Right now , is the number of data, represent The row straightens the vector, represents the 3n-dimensional unit matrix, is the matrix composed of magnetic vectors of training data; Step 5.3. Get the vector geomagnetic field of the predicted location: ; in, To predict the location, is the predicted magnetic field at the predicted position. In this step, the predicted position is the same as the training position. is an n-dimensional vector, where the i-th element is , is an n-dimensional matrix, where the element in the i-th row and j-th column is , That is, the coordinates of the i-th and j-th data in the training data; , Respectively represent the mean and variance of the predicted magnetic field; Step 5.4. Use the LM algorithm to minimize the residual between the prediction result and the input data: ; in, It is the rotation matrix from the carrier coordinate system to the earth coordinate system.

[0031] like Figure 2 , Figure 3 , Figure 4 The figure shows the comparison of the results after compensation by the vehicle-mounted vector magnetic compensation method of the present invention and the ellipsoid fitting method. Table 1 is a comparison table of the results.

[0032] Table 1

[0033]

[0034] In this experiment, dual fluxgates were mounted at different positions of the carrier to collect magnetic data. The distance between the two was 60 cm, which can be ignored in the spatial scale of the geomagnetic field, that is, the magnetic data collected by the two should be roughly consistent after compensation; at the same time, the two were mounted at different positions of the carrier, and the interference of the carrier and the sensor itself could not be ignored, so the data before compensation was quite different. From the results, it can be seen that the present invention can effectively compensate for the magnetic interference error, while the ellipsoid fitting is limited by the data pathology and cannot solve the effective compensation coefficient.

[0035] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the above-mentioned vehicle-mounted vector magnetic compensation method based on dot product invariant constraint are implemented.

[0036] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-mentioned vehicle-mounted vector magnetic compensation method based on dot product invariant constraints.

[0037] It should be understood by those skilled in the art that the embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The solutions in the embodiments of the present invention may be implemented in various computer languages, for example, object-oriented programming language Java and interpreted scripting language JavaScript, etc.

[0038] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0039] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0040] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0041] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0042] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A vehicle-mounted vector magnetic compensation method based on dot product invariance constraint, characterized in that: The steps include: Step 1: construct a magnetic interference model of the vehicle carrier, establish an error equation for measuring interference through the magnetic interference model and constraints, and realize compensation for errors related to magnetic vector measurement of the vehicle carrier; Step 2: Calculate the rotation matrix from the earth system to the load system. The rotation matrix is ​​used to construct the auxiliary vector. Step 3, construct constraints based on the auxiliary vector of step 2, and construct vector compensation parameter solution equations based on the dot product invariant constraints; Step 4, solving the equation according to the vector compensation parameter, and solving the vector compensation parameter based on the improved genetic algorithm; Step 5: Parameter optimization is performed based on residual Gaussian process regression. It is assumed that the vector geomagnetic field satisfies the Gaussian regression model. Considering the intrinsic physical connection between the three axes of the vector geomagnetic field, according to Maxwell's equations, the geomagnetic field gradient is estimated using Gaussian regression based on the physical principle that the geomagnetic field vector is the gradient of the scalar magnetic potential in the irrotational case.

2. The vehicle-mounted vector magnetic compensation method based on dot product invariant constraint according to claim 1 is characterized in that: The step 1 comprises: The magnetic interference model of the vehicle carrier is: ; Intermediate parameters , intermediate parameters ; Therefore, the calibration model is ,in, , ; and That is the parameter required for compensation, by calculating , The specific value of will measure the magnetic vector Compensation for true magnetic vector ; in, is the projection vector value of the geomagnetic field in the carrier coordinate system, and the three-axis values ​​are ; is the measured magnetic vector value in the carrier coordinate system, and the three-axis values ​​are ; is the misalignment error matrix between the measurement coordinate system of the magnetic sensor and the carrier coordinate system; is the magnetic sensor sensitivity error matrix; is the non-orthogonal error matrix of the magnetic sensor; is the soft magnetic error matrix of the vehicle carrier; is the hard magnetic error vector of the vehicle carrier; is the zero bias error vector of the magnetic sensor; Geodetic coordinate system The Northeast Sky coordinate system, the carrier coordinate system It is the upper right front coordinate system.

3. The vehicle-mounted vector magnetic compensation method based on dot product invariant constraint according to claim 1 is characterized in that: The step 2 comprises: The auxiliary vector is a constant vector of the artificially selected geodetic coordinate system. , the vehicle attitude angle is measured as the yaw angle , pitch angle , roll angle , then the rotation matrix from the earth system to the load system is: Rotation matrix for rotation around yaw angle for: ; Rotation matrix for rotation around pitch angle for: ; Rotation matrix for rotation around the roll angle for: ; Total rotation matrix for: ； Then the auxiliary vector in the carrier coordinate system is .

4. The vehicle-mounted vector magnetic compensation method based on dot product invariant constraint according to claim 1 is characterized in that: The step 3 comprises: The auxiliary vector and the geomagnetic field vector are both constant vectors in the geodetic coordinate system, so the dot product of the two is a fixed value, and this fixed value is valid in any rectangular coordinate system, that is: ； The superscript T represents the transpose of the matrix. represents a fixed constant; Expand it according to the carrier coordinate system components, let , , have , i, j represent the subscripts of the elements in each variable, i, j = 1, 2, 3; collect enough and sufficient data to calculate and The specific value of .

5. The vehicle-mounted vector magnetic compensation method based on dot product invariance constraint according to claim 4 is characterized in that: The step 3 comprises the following steps: Step 3.1 calculates the dot product result of the auxiliary vector and the magnetic vector data; Step 3.2 Use the constraint that the dot product result remains unchanged to construct the equation to solve the compensation parameters , ; Step 3.3 According to the formula Get the real magnetic data after rough compensation .

6. The vehicle-mounted vector magnetic compensation method based on dot product invariance constraint according to claim 1 is characterized in that: The step 4 comprises: Select multi-generation improved real-coded genetic algorithm to solve compensation parameters, whose input is the population size , mutation probability , crossover probability , variable upper and lower bounds , number of elites , crossover probability Generally set to 1; The initial population individuals are set as and , a total of 12 compensation parameters, solve the individual that minimizes the objective function, the objective function f is set to , assuming is 1, so the solution is and Is a relative value, add constraints To fix the absolute value of the parameter.

7. The vehicle-mounted vector magnetic compensation method based on dot product invariance constraint according to claim 6 is characterized in that: The step 4 comprises the following steps: Step 4.1 Initialize the population; Step 4.2 calculates the fitness function of the individuals and sorts them; Step 4.3 determines whether the cut-off condition is met, if not, go to step 4.4, if met, go to step 4.14; Step 4.4 Perform sorting and grouping selection, select parental pairs; Step 4.5 uses the heuristic normal distribution and the direction-based crossover operator to generate offspring individuals; Step 4.6 performs a replacement operation to randomly generate new individuals to replace duplicate individuals in the offspring population; Step 4.7 calculates the fitness function of the individuals and sorts them; Step 4.8 From the parent and The offspring individuals are retained The elite individuals with the highest fitness; Step 4.9 executes a combined mutation algorithm of the integrated Cauchy mutation operator, Gaussian mutation operator, and Levy mutation operator; Step 4.10 calculates the fitness function of the individuals and sorts them; Step 4.11 makes Elite individuals, mutant individuals and Non-mutant individuals make up the current population; Step 4.12 Select The best individuals form the central group, retaining the new population an elite individual; Step 4.13 Go to step 4.3; Step 4.14 outputs the average running time, average number of iterations and optimal individual value.

8. The vehicle-mounted vector magnetic compensation method based on dot product invariance constraint according to claim 1 is characterized in that: The step 5 comprises the following steps: Step 5.

1. Build a Gaussian regression model: ； in, is the scalar magnetic potential, represents a Gaussian random process, is the kernel function of the Gaussian process, and the specific form is selected according to the actual situation. , is the spatial position coordinate, is the geomagnetic field vector value in the geodetic coordinate system, and its subscript i represents a specific location. is the coordinate of position i, is the magnetic field at position i; represents the gradient, For noise and , N represents Gaussian distribution, is a noise-related hyperparameter, represents a three-dimensional unit matrix; The input data is the vector magnetic data and the corresponding spatial coordinates after the rough compensation of the calibration data, and the output data is the vector magnetic data of the predicted position that conforms to the Gaussian regression model; Step 5.

2. Perform maximum likelihood estimation on the hyperparameters involved in the Gaussian regression model to obtain the optimal hyperparameters. The maximum likelihood function for: ； in, is the hyperparameter vector, The subscript represents the hyperparameter value of the kernel function. ,Right now , is the number of data, represent The row straightens the vector, represents the 3n-dimensional unit matrix, is the matrix composed of magnetic vectors of training data; Step 5.

3. Get the vector geomagnetic field of the predicted location: ； in, To predict the location, is the predicted magnetic field at the predicted position, the predicted position is the same as the training position, is an n-dimensional vector, where the i-th element is , is an n-dimensional matrix, where the element in the i-th row and j-th column is , That is, the coordinates of the i-th and j-th data in the training data; , Respectively represent the mean and variance of the predicted magnetic field; Step 5.

4. Use the LM algorithm to minimize the residual between the prediction result and the input data: ； in, It is the rotation matrix from the carrier coordinate system to the earth coordinate system.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of a vehicle-mounted vector magnetic compensation method based on dot product invariant constraints as described in any one of claims 1-8 are implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of a vehicle-mounted vector magnetic compensation method based on dot product invariant constraints as described in any one of claims 1 to 8 are implemented.

Images