Precise Positioning Method and System for the Virtual-Reality Fusion of Intelligent Internet of Things Devices and the Metaverse

Abstract

The present invention provides a precise positioning method and system for the virtual-real integration of intelligent Internet of Things devices and the metaverse, which relates to the technical field of the metaverse. It includes obtaining the position and spatial relationship information of intelligent Internet of Things devices, and using an asynchronous iterative optimization algorithm to construct a global positioning knowledge graph to achieve high-precision collaborative positioning. A corresponding virtual scene and virtual anchor points are established in the virtual space of the metaverse, and a virtual-real mapping relationship verification mechanism is established. The virtual objects are dynamically adjusted according to the positioning results and the verification mechanism, and two-way interaction between the physical space and the virtual space is realized through a dynamic adjustment function. The present invention improves the accuracy and stability of virtual-real integration, and enhances the user's immersion and interaction experience.

Description

Technical Field

[0001] The present invention relates to the metaverse technology, and particularly to a precise positioning method and system for the virtual-real integration of intelligent Internet of Things devices and the metaverse. Background Art

[0002] With the rapid development of the Internet of Things and virtual reality technologies, the integration of intelligent Internet of Things devices and the virtual space of the metaverse has become an important research direction. This integration aims to achieve seamless connection between the physical world and the virtual world, providing users with a more immersive and interactive experience. In this context, precise positioning technology has become a key link connecting the physical space and the virtual space.

[0003] Traditional positioning methods mainly rely on GPS or other sensor data of a single device, and it is difficult to meet the positioning requirements in high-precision and complex environments. At the same time, the corresponding relationship between the virtual space and the physical space lacks an effective verification and adjustment mechanism, resulting in the problem of spatio-temporal drift in the virtual-real integration process. In addition, the existing virtual-real interaction methods are often one-way, and it is difficult to achieve real-time two-way interaction between the physical space and the virtual space. Summary of the Invention

[0004] Embodiments of the present invention provide a precise positioning method and system for the virtual-real integration of intelligent Internet of Things devices and the metaverse, which can solve the problems in the prior art.

[0005] In the first aspect of the embodiments of the present invention, a precise positioning method for the virtual-real integration of intelligent Internet of Things devices and the metaverse is provided, including:

[0006] Obtaining the position information of multiple intelligent Internet of Things devices and the spatial relationship information of the intelligent Internet of Things devices, where the position information includes geographical coordinates and attitude data; based on the position information and the spatial relationship information, constructing a global positioning knowledge graph through an asynchronous iterative optimization algorithm to achieve collaborative high-precision positioning of multiple intelligent Internet of Things devices, and obtaining a positioning result;

[0007] Establishing a virtual scene corresponding to the physical space in the metaverse virtual space; determining virtual anchor points in the virtual scene, where the virtual anchor points correspond to physical reference objects in the physical space;

[0008] Establishing a virtual-real mapping relationship verification mechanism to verify and update the mapping relationship between the position information of the intelligent Internet of Things devices and the virtual anchor points, eliminating spatio-temporal drift in the virtual-real integration;

[0009] Dynamically adjusting the positions and states of virtual objects in the virtual scene according to the positioning result and the virtual-real mapping relationship verification mechanism;

[0010] Feedback the interaction information in the virtual scene to the intelligent IoT device, and establish a dynamic adjustment function including a position constraint term and a state constraint term to realize the two-way interaction between the physical space and the metaverse virtual space.

[0011] Based on the position information and the spatial relationship information, collaboratively construct a global positioning knowledge graph through an asynchronous iterative optimization algorithm to realize the collaborative high-precision positioning of multiple intelligent IoT devices. The obtained positioning results include:

[0012] Construct a global knowledge graph based on the position information and the spatial relationship information. The global knowledge graph includes a node set, an edge set, and a weight matrix, where the node set represents the position nodes of multiple intelligent IoT devices, the edge set represents the spatial relationships between the intelligent IoT devices, and the weight matrix represents the spatial constraint strength between the intelligent IoT devices;

[0013] Establish a local consistency constraint model according to the node set and the edge set of the global knowledge graph. The local consistency constraint model includes an Euclidean distance term and a regularization term of the device position, and dynamically balance the Euclidean distance term and the regularization term through the weight coefficients in the weight matrix to obtain a local constraint optimization objective;

[0014] Use the alternating direction multiplier method to decompose the local constraint optimization objective into multiple local sub-problems, and obtain an initial collaborative optimization position result by iteratively optimizing the local variables, global variables, and dual variables of each local sub-problem;

[0015] Input the initial collaborative optimization position result into a graph optimization framework, construct an optimization function including a position error term and an information matrix, and optimize the initial collaborative optimization position result based on the optimization function to obtain the collaborative positioning results of multiple intelligent IoT devices.

[0016] Use the alternating direction multiplier method to decompose the local constraint optimization objective into multiple local sub-problems, and obtain an initial collaborative optimization position result by iteratively optimizing the local variables, global variables, and dual variables of each local sub-problem. The obtained initial collaborative optimization position result includes:

[0017] Each local sub-problem corresponds to the position optimization of an intelligent IoT device, and each local sub-problem includes: a local variable for characterizing the position of a single intelligent IoT device, a global variable for coordinating the position consistency of multiple intelligent IoT devices, and a dual variable for ensuring convergence;

[0018] Based on a preset step size parameter, a three-step iterative strategy is adopted to solve the local sub-problem: First, optimize the local variables by fixing the global variables and dual variables, then optimize the global variables by fixing the local variables and dual variables, and finally update the dual variables. Repeat the iteration until the position difference between two adjacent iterations is less than the preset deviation threshold to obtain the initial co-optimization position result.

[0019] Input the initial co-optimization position result into the graph optimization framework, construct an optimization function including a position error term and an information matrix, and optimize the initial co-optimization position result based on the optimization function to obtain the collaborative positioning results of multiple intelligent IoT devices, including:

[0020] Decompose the optimization function into a position optimization sub-problem and a weight optimization sub-problem. The position optimization sub-problem iteratively optimizes the positions of the intelligent IoT devices by the least squares method, and the weight optimization sub-problem calculates the weight coefficients of the information matrix based on the Mahalanobis distance.

[0021] Use the gradient descent algorithm with an adaptive step size to alternately solve the position optimization sub-problem and the weight optimization sub-problem. Dynamically adjust the optimization step size according to the convergence degree of the position error term in each iteration, and use the information matrix to suppress the influence of abnormal observations on the optimization result until the preset number of iterations is satisfied to obtain the collaborative positioning results of multiple intelligent IoT devices.

[0022] Establish a virtual-real mapping relationship verification mechanism to verify and update the mapping relationship between the position information of the intelligent IoT devices and the virtual anchors, and eliminate the spatio-temporal drift of virtual-real fusion, including:

[0023] Based on the position information of the intelligent IoT devices and the position information of the corresponding anchors in the virtual space, construct a space transformation matrix including a rotation matrix and a translation vector. The space transformation matrix is used to establish the initial mapping relationship from the physical space to the virtual space.

[0024] Use the space transformation matrix to transform the position information of the intelligent IoT devices, calculate the Euclidean distance between the transformed position and the position of the virtual space anchor, and obtain the virtual-real mapping error. The virtual-real mapping error is used to characterize the accuracy of the current mapping relationship.

[0025] Calculate the cumulative drift amount of the position information of the intelligent IoT devices and the position information of the corresponding anchors in the virtual space at adjacent times, and obtain the drift change rate by taking the time derivative of the cumulative drift amount. Weightedly combine the drift change rate with the gradient of the virtual-real mapping error to construct a compensation vector with an adaptive characteristic, where the weighting coefficient is dynamically adjusted according to the change of the virtual-real mapping error.

[0026] Update the spatial transformation matrix based on the compensation vector to obtain an optimized virtual-real mapping relationship.

[0027] Feedback the interaction information in the virtual scene to the intelligent IoT device, and establish a dynamic adjustment function including a position constraint term and a state constraint term to achieve two-way interaction between the physical space and the metaverse virtual space, including:

[0028] Establish a dynamic adjustment function including a position constraint term and a state constraint term. The position constraint term ensures that the position of the virtual object is synchronized with the mapped position of the intelligent IoT device, and the state constraint term ensures that the motion state of the virtual object matches the actual state of the intelligent IoT device;

[0029] Use an iterative optimization algorithm with adaptive weights to solve the dynamic adjustment function, where the weight coefficient is dynamically updated according to the time series of the position error;

[0030] Apply the optimized position and state parameters to the virtual objects in the virtual scene to achieve synchronous interaction between the virtual objects and the physical devices.

[0031] In the second aspect of the embodiments of the present invention, a precise positioning system for the virtual-real fusion of intelligent IoT devices and the metaverse is provided, including:

[0032] The first unit is used to obtain the position information of multiple intelligent IoT devices and the spatial relationship information of the intelligent IoT devices. The position information includes geographical coordinates and attitude data; based on the position information and the spatial relationship information, a global positioning knowledge graph is collaboratively constructed through an asynchronous iterative optimization algorithm to achieve collaborative high-precision positioning of multiple intelligent IoT devices and obtain a positioning result;

[0033] The second unit is used to establish a virtual scene corresponding to the physical space in the metaverse virtual space; determine virtual anchors in the virtual scene, and the virtual anchors correspond to physical reference objects in the physical space;

[0034] The third unit is used to establish a virtual-real mapping relationship verification mechanism to verify and update the mapping relationship between the position information of the intelligent IoT device and the virtual anchor, and eliminate the spatio-temporal drift of virtual-real fusion;

[0035] The fourth unit is used to dynamically adjust the position and state of the virtual objects in the virtual scene according to the positioning result and the virtual-real mapping relationship verification mechanism;

[0036] The fifth unit is used to feedback the interaction information in the virtual scene to the intelligent IoT device, and establish a dynamic adjustment function including a position constraint term and a state constraint term to achieve two-way interaction between the physical space and the metaverse virtual space.

[0037] The third aspect of the embodiments of the present invention

[0038] A kind of electronic device is provided, including:

[0039] A processor;

[0040] A memory for storing instructions executable by the processor;

[0041] Wherein, the processor is configured to call the instructions stored in the memory to execute the method described above.

[0042] The fourth aspect of the embodiments of the present invention,

[0043] A computer-readable storage medium is provided, on which computer program instructions are stored, and when the computer program instructions are executed by a processor, the method described above is implemented.

[0044] The beneficial effects of the present application are as follows:

[0045] By using the asynchronous iterative optimization algorithm to collaboratively construct the global positioning knowledge graph, the collaborative high-precision positioning of multiple intelligent Internet of Things devices is realized, and the accuracy and reliability of positioning are improved. This method makes full use of the spatial relationship information between devices and overcomes the limitations of single-device positioning.

[0046] A virtual-real mapping relationship verification mechanism is established to verify and update the mapping relationship between the position information of intelligent Internet of Things devices and virtual anchor points, effectively eliminating the spatio-temporal drift problem in the virtual-real fusion process. This ensures the accurate correspondence between the virtual scene and the physical space, enhancing the user's immersion and interaction experience.

[0047] By establishing a dynamic adjustment function including position constraint terms and state constraint terms, the two-way interaction between the physical space and the metaverse virtual space is realized. This method can not only adjust the virtual scene in real time according to the changes in the physical world, but also feedback the interaction information in the virtual world to the physical device, thus realizing the seamless fusion and interaction between the virtual and the real, and greatly improving the practicality and experience of metaverse applications. Description of the Drawings

[0048] Figure 1 It is a schematic flowchart of the accurate positioning method for the virtual-real fusion of intelligent Internet of Things devices and the metaverse in the embodiments of the present invention;

[0049] Figure 2 It is a logical schematic diagram of the collaborative positioning of multiple intelligent Internet of Things devices;

[0050] Figure 3 It is a schematic framework diagram of the accurate positioning system for the virtual-real fusion of intelligent Internet of Things devices and the metaverse in the embodiments of the present invention. Detailed Embodiments

[0051] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to 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 fall within the scope of protection of the present invention.

[0052] The technical solutions of the present invention will be described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments.

[0053] Figure 1 As shown in the flowchart of the precise positioning method for the virtual-real fusion of intelligent Internet of Things devices and the metaverse in the embodiments of the present invention, Figure 1 as shown, the method includes:

[0054] Obtain the position information of multiple intelligent Internet of Things devices and the spatial relationship information of the intelligent Internet of Things devices. The position information includes geographical coordinates and attitude data; based on the position information and the spatial relationship information, collaboratively construct a global positioning knowledge graph through an asynchronous iterative optimization algorithm to achieve the collaborative high-precision positioning of multiple intelligent Internet of Things devices and obtain a positioning result;

[0055] Establish a virtual scene corresponding to the physical space in the metaverse virtual space; determine virtual anchor points in the virtual scene, and the virtual anchor points correspond to physical reference objects in the physical space;

[0056] Establish a virtual-real mapping relationship verification mechanism to verify and update the mapping relationship between the position information of the intelligent Internet of Things devices and the virtual anchor points, and eliminate the spatio-temporal drift of virtual-real fusion;

[0057] According to the positioning result and the virtual-real mapping relationship verification mechanism, dynamically adjust the positions and states of virtual objects in the virtual scene;

[0058] Feedback the interaction information in the virtual scene to the intelligent Internet of Things devices, and establish a dynamic adjustment function including position constraint terms and state constraint terms to achieve two-way interaction between the physical space and the metaverse virtual space.

[0059] In an optional implementation manner, based on the position information and the spatial relationship information, collaboratively constructing a global positioning knowledge graph through an asynchronous iterative optimization algorithm to achieve the collaborative high-precision positioning of multiple intelligent Internet of Things devices and obtain a positioning result includes:

[0060] Construct a global knowledge graph based on the position information and the spatial relationship information. The global knowledge graph includes a node set, an edge set, and a weight matrix. The node set represents the position nodes of multiple intelligent IoT devices, the edge set represents the spatial relationships between the intelligent IoT devices, and the weight matrix represents the spatial constraint strength between the intelligent IoT devices;

[0061] Establish a local consistency constraint model according to the node set and the edge set of the global knowledge graph. The local consistency constraint model includes an Euclidean distance term and a regularization term of the device position. The Euclidean distance term and the regularization term are dynamically balanced by the weight coefficients in the weight matrix to obtain a local constraint optimization objective;

[0062] Use the alternating direction method of multipliers to decompose the local constraint optimization objective into multiple local sub-problems. By iteratively optimizing the local variables, global variables, and dual variables of each local sub-problem, an initial cooperative optimization position result is obtained;

[0063] Input the initial cooperative optimization position result into a graph optimization framework, construct an optimization function including a position error term and an information matrix, and optimize the initial cooperative optimization position result based on the optimization function to obtain the cooperative positioning results of multiple intelligent IoT devices.

[0064] The present invention provides a method for realizing cooperative high-precision positioning of multiple intelligent IoT devices based on an asynchronous iterative optimization algorithm. This method realizes the cooperative positioning of multiple devices by constructing a global positioning knowledge graph, using a local consistency constraint model and a graph optimization framework. Figure 2 It is a logical schematic diagram for the cooperative positioning of multiple intelligent IoT devices. The specific implementation method is as follows:

[0065] The position information can be the initial estimated position coordinates of the device, such as (x, y, z). The spatial relationship information includes the relative distance, angle, etc. between the devices. For example, the distance between device A and device B is 5 meters, and the included angle between device B and device C is 30 degrees.

[0066] Construct a global positioning knowledge graph based on the collected information. This graph contains three key elements: a node set, an edge set, and a weight matrix. The node set represents the position nodes of each intelligent IoT device, and each node corresponds to the three-dimensional coordinates of a device. The edge set represents the spatial relationships between the devices, such as distance edges, angle edges, etc. The weight matrix represents the strength of the spatial constraints between the devices, and the weight values can be determined according to factors such as measurement accuracy. For example, the weight of high-precision distance measurement can be set to 0.9, and the weight of low-precision angle measurement can be set to 0.6.

[0067] A local consistency constraint model is established based on the global knowledge graph. This model consists of two parts: an Euclidean distance term and a regularization term. The Euclidean distance term is used to constrain the relative position relationship between devices, and the regularization term is used to smooth the optimization result. The two terms are dynamically balanced through the coefficients in the weight matrix to obtain the local constraint optimization objective. For example, the weight of the Euclidean distance term can be set to 0.7, and the weight of the regularization term can be set to 0.3.

[0068] The alternating direction method of multipliers is used to decompose the local constraint optimization objective into multiple local subproblems. Each subproblem corresponds to the position optimization of a device. By iteratively optimizing the local variables, global variables, and dual variables of each subproblem, an initial collaborative optimization position result is obtained. Specifically, the maximum number of iterations can be set to 100, and the convergence threshold can be set to 0.001 meters.

[0069] In each iteration, the local variables are updated first. For each device, the optimal position of the device is calculated based on the position information of other devices and the spatial relationship constraints. Then the global variables are updated by taking the weighted average of the local variables of all devices. Finally, the dual variables are updated to adjust the difference between the local variables and the global variables. This process is repeated until the maximum number of iterations is reached or convergence occurs.

[0070] After obtaining the initial collaborative optimization position result, it is input into the graph optimization framework for further optimization. An optimization function including a position error term and an information matrix is constructed. The position error term represents the difference between the optimized position and the initial position, and the information matrix represents the uncertainty of position measurement. Based on this optimization function, the initial result is optimized to obtain the final collaborative positioning result.

[0071] In the graph optimization process, the Levenberg-Marquardt algorithm can be used for solution. The maximum number of iterations is set to 50, and the convergence threshold is set to 0.0001 meters. In each iteration, the error and Jacobian matrix of the current position are first calculated. Then the position increment is obtained by solving the linear equations, and the current position is updated.

[0072] In an alternative embodiment, using the alternating direction method of multipliers to decompose the local constraint optimization objective into multiple local subproblems, and obtaining the initial collaborative optimization position result by iteratively optimizing the local variables, global variables, and dual variables of each local subproblem includes:

[0073] Each local subproblem corresponds to the position optimization of an intelligent IoT device, where each local subproblem includes: local variables for characterizing the position of a single intelligent IoT device, global variables for coordinating the position consistency of multiple intelligent IoT devices, and dual variables for ensuring convergence;

[0074] Based on the preset step size parameter, a three-step iterative strategy is adopted to solve the local sub-problem: First, fix the global variables and dual variables to optimize the local variables, then fix the local variables and dual variables to optimize the global variables, and finally update the dual variables. Repeat the iteration until the position difference between two adjacent iterations is less than the preset deviation threshold to obtain the initial co-optimization position result.

[0075] Each local sub-problem contains three types of variables: Local variables are used to represent the positions of individual devices, global variables are used to coordinate the consistency of the positions of multiple devices, and dual variables are used to ensure the convergence of the algorithm.

[0076] Taking a system with 5 intelligent IoT devices as an example, the position of each device is represented by three-dimensional coordinates (x, y, z). Then each local sub-problem contains:

[0077] - Local variables: (x_i, y_i, z_i), i = 1,2,3,4,5;

[0078] - Global variables: (X, Y, Z);

[0079] - Dual variables: (λ_x_i, λ_y_i, λ_z_i), i = 1,2,3,4,5;

[0080] Next, a three-step iterative strategy is adopted to solve these local sub-problems:

[0081] The first step: Fix the global variables and dual variables and optimize the local variables.

[0082] For each device i, solve the following optimization problem:

[0083] Minimize f_i(x_i, y_i, z_i) + (λ_x_i)(x_i - X) + (λ_y_i)(y_i - Y) + (λ_z_i)(z_i - Z) + (ρ / 2)[(x_i - X)^2 + (y_i - Y)^2 + (z_i - Z)^2];

[0084] Where f_i is the local objective function of device i, and ρ is the preset penalty parameter.

[0085] The second step: Fix the local variables and dual variables and optimize the global variables.

[0086] Solve the following optimization problem:

[0087] Minimize Σ_i=1^5 [(λ_x_i)(x_i - X) + (λ_y_i)(y_i - Y) + (λ_z_i)(z_i -Z) + (ρ / 2)[(x_i - X)^2 + (y_i - Y)^2 + (z_i - Z)^2]];

[0088] Step 3: Update the dual variables.

[0089] For each device i, update the dual variables:

[0090] λ_x_i = λ_x_i + ρ(x_i - X);

[0091] λ_y_i = λ_y_i + ρ(y_i - Y);

[0092] λ_z_i = λ_z_i + ρ(z_i - Z);

[0093] Repeat the above three steps until the convergence condition is met. Here, the position difference between two adjacent iterations is used as the convergence criterion.

[0094] To better illustrate the implementation process of this method, a specific numerical example is given below:

[0095] Suppose there are 5 intelligent IoT devices with the initial positions as follows:

[0096] Device 1: (0, 0, 0); Device 2: (1, 1, 1); Device 3: (2, 2, 2); Device 4: (3, 3, 3); Device 5: (4, 4, 4);

[0097] Set the algorithm parameters: penalty parameter ρ = 1.0, step size parameter α = 0.1, maximum number of iterations is 100, and convergence threshold is 0.01 meters.

[0098] Iteration 1:

[0099] 1. Optimize the local variables:

[0100] Device 1: (0.1, 0.1, 0.1); Device 2: (1.1, 1.1, 1.1); Device 3: (2.0, 2.0, 2.0); Device 4: (2.9, 2.9, 2.9); Device 5: (3.9, 3.9, 3.9);

[0101] 2. Optimize the global variables:

[0102] (X, Y, Z) = (2.0, 2.0, 2.0);

[0103] 3. Update dual variables:

[0104] Device 1: (-1.9, -1.9, -1.9); Device 2: (-0.9, -0.9, -0.9); Device 3: (0, 0, 0); Device 4: (0.9, 0.9, 0.9); Device 5: (1.9, 1.9, 1.9);

[0105] The 2nd iteration:

[0106] 1. Optimize local variables:

[0107] Device 1: (0.29, 0.29, 0.29); Device 2: (1.19, 1.19, 1.19); Device 3: (2.0, 2.0, 2.0); Device 4: (2.81, 2.81, 2.81); Device 5: (3.71, 3.71, 3.71);

[0108] 2. Optimize global variables:

[0109] (X, Y, Z) = (2.0, 2.0, 2.0);

[0110] 3. Update dual variables:

[0111] Device 1: (-1.71, -1.71, -1.71); Device 2: (-0.81, -0.81, -0.81); Device 3: (0, 0, 0); Device 4: (0.81, 0.81, 0.81); Device 5: (1.71, 1.71, 1.71).

[0112] Continue the iteration until the convergence condition is met or the maximum number of iterations is reached. In this example, after about 50 iterations, the algorithm converges to the following results:

[0113] Device 1: (1.8, 1.8, 1.8); Device 2: (1.9, 1.9, 1.9); Device 3: (2.0, 2.0, 2.0); Device 4: (2.1, 2.1, 2.1); Device 5: (2.2, 2.2, 2.2); Global variables: (X, Y, Z) = (2.0, 2.0, 2.0).

[0114] It can be seen that the final co - optimization position results make the positions of all devices closer to the central position represented by the global variables, while still maintaining the relative position relationship between the devices.

[0115] The advantages of this method are as follows:

[0116] 1. Decompose complex global optimization problems into multiple simple local sub-problems, reducing the computational complexity. 2. Achieve the coordination of multiple device positions by introducing global variables and dual variables. 3. Adopt an iterative optimization strategy to ensure the convergence and stability of the algorithm. 4. Can flexibly handle different numbers and types of intelligent IoT devices, with good scalability.

[0117] In practical applications, the penalty parameter, step size parameter, and convergence threshold can be adjusted according to specific scenarios to balance the optimization effect and computational efficiency. At the same time, other constraint conditions (such as device communication range, obstacle avoidance, etc.) can be combined to further improve the accuracy and practicality of the collaborative positioning of intelligent IoT devices.

[0118] In an alternative embodiment, the initial collaborative optimization position result is input into a graph optimization framework to construct an optimization function including a position error term and an information matrix, and the initial collaborative optimization position result is optimized based on the optimization function to obtain the collaborative positioning results of multiple intelligent IoT devices, including:

[0119] Decompose the optimization function into a position optimization sub-problem and a weight optimization sub-problem, where the position optimization sub-problem iteratively optimizes the positions of intelligent IoT devices by the least squares method, and the weight optimization sub-problem calculates the weight coefficient of the information matrix based on the Mahalanobis distance;

[0120] Use the gradient descent algorithm with an adaptive step size to alternately solve the position optimization sub-problem and the weight optimization sub-problem, dynamically adjust the optimization step size according to the convergence degree of the position error term in each iteration, and use the information matrix to suppress the influence of abnormal observations on the optimization result until the preset number of iterations is satisfied, obtaining the collaborative positioning results of multiple intelligent IoT devices.

[0121] Exemplarily, decompose the optimization function into a position optimization sub-problem and a weight optimization sub-problem, and the position optimization sub-problem iteratively optimizes the positions of intelligent IoT devices by the least squares method. Specifically, for each intelligent IoT device, an observation equation is established according to its relative position relationship with surrounding devices. Then, all observation equations are combined into an overdetermined system of equations, and the least squares method is used to solve this system of equations to obtain the correction amount of the device position. Through multiple iterations, the device position is continuously corrected to gradually approach the true position.

[0122] The weight optimization sub-problem calculates the weight coefficients of the information matrix based on the Mahalanobis distance. Specifically, for the observations between each pair of devices, the Mahalanobis distance is calculated. The larger the Mahalanobis distance, the more likely the observation is an outlier, and its weight should be smaller. According to the calculated Mahalanobis distance, an exponential decay function is used to convert it into a weight coefficient between 0 and 1. These weight coefficients form the diagonal elements of the information matrix, which are used to suppress the influence of outlier observations during the optimization process.

[0123] The gradient descent algorithm with an adaptive step size is used to alternately solve the position optimization sub-problem and the weight optimization sub-problem. In each iteration, first fix the weights and solve the position optimization sub-problem; then fix the positions and solve the weight optimization sub-problem. The specific steps are as follows:

[0124] 1. Initialize the device positions and weights. The initial co-optimization position result can be used as the initial value of the device positions, and all weights are initialized to 1.

[0125] 2. Fix the weights and solve the position optimization sub-problem. Use the gradient descent method to update the device positions, and the update formula is:

[0126] New position = Old position - Step size * Gradient;

[0127] where the gradient is obtained by differentiating the error function.

[0128] 3. According to the updated positions, recalculate the Mahalanobis distance of the observations and update the weights.

[0129] 4. Fix the positions and solve the weight optimization sub-problem. Also use the gradient descent method to update the weights.

[0130] 5. Calculate the position error term and determine whether the convergence condition is satisfied. If satisfied, end the iteration; otherwise, continue the next iteration.

[0131] During the iteration process, the optimization step size is dynamically adjusted according to the convergence degree of the position error term. Specifically, the following strategy can be adopted:

[0132] - If the position error term decreases continuously for several consecutive iterations, increase the step size to accelerate the convergence speed.

[0133] - If the position error term oscillates or increases, decrease the step size to improve the optimization stability.

[0134] - Set the upper and lower limits of the step size to prevent the step size from being too large or too small, resulting in optimization failure.

[0135] For example, the following formula can be used to dynamically adjust the step size:

[0136] New step size = Old step size * (1 + α * sign(Old error - New error));

[0137] Where α is a positive constant less than 1, controlling the amplitude of step size adjustment; the sign function returns -1, 0, or 1, indicating the direction of error change.

[0138] To further improve the optimization effect, the information matrix is used to suppress the influence of abnormal observations on the optimization result in each iteration. Specifically, when calculating the error function and gradient, the observations are multiplied by the corresponding weight coefficients. In this way, due to the small weights, the influence of abnormal observations will be greatly reduced.

[0139] Repeat the above iterative process until the preset number of iterations or convergence conditions are met, and finally obtain the collaborative positioning results of multiple intelligent IoT devices.

[0140] The following gives a specific data case to illustrate the implementation process and effect of this method:

[0141] Suppose there are 5 intelligent IoT devices, and their initial position coordinates (unit: meter) are respectively:

[0142] Device 1: (0, 0); Device 2: (10, 0); Device 3: (5, 8.66); Device 4: (2, 3); Device 5: (8, 4);

[0143] The actual distances (unit: meter) between the devices are:

[0144] 1-2: 10.0, 1-3: 10.0, 1-4: 3.61, 1-5: 8.94; 2-3: 10.0, 2-4: 8.54, 2-5: 4.47; 3-4: 6.71, 3-5: 5.83; 4-5: 6.32.

[0145] Due to measurement errors, the observed distances (unit: meter) are:

[0146] 1-2: 10.2, 1-3: 9.8, 1-4: 3.5, 1-5: 9.1;

[0147] 2-3: 10.1, 2-4: 8.7, 2-5: 4.3;

[0148] 3-4: 6.8, 3-5: 5.7;

[0149] 4-5: 6.5;

[0150] Apply this method for collaborative positioning optimization, set the maximum number of iterations to 100, the initial step size to 0.1, the step size adjustment parameter α to 0.1, and the convergence threshold to 0.01 meters. After optimization, the obtained device position coordinates (unit: meters) are:

[0151] Device 1: (0.02, -0.03); Device 2: (10.05, 0.04); Device 3: (4.98, 8.69); Device 4: (1.97, 2.98); Device 5: (8.03, 3.97).

[0152] It can be seen that the optimized position coordinates are very close to the actual positions, and the maximum error does not exceed 0.07 meters. At the same time, by analyzing the weight coefficients, it can be found that the weights of the observations between 1 - 5 are relatively low (about 0.6), indicating that the algorithm has successfully identified and suppressed the influence of this abnormal observation value.

[0153] Through this case, it can be verified that this method can effectively achieve the collaborative positioning of multiple intelligent IoT devices and has good robustness to abnormal observation values. In practical applications, the parameter settings can be adjusted according to specific scenarios and requirements to obtain the best positioning effect.

[0154] In an alternative implementation, a virtual - physical mapping relationship verification mechanism is established to verify and update the mapping relationship between the position information of the intelligent IoT device and the virtual anchor point, and eliminate the spatio - temporal drift of virtual - physical fusion, including:

[0155] Based on the position information of the intelligent IoT device and the position information of the corresponding anchor point in the virtual space, construct a space transformation matrix including a rotation matrix and a translation vector, and the space transformation matrix is used to establish an initial mapping relationship from the physical space to the virtual space;

[0156] Use the space transformation matrix to transform the position information of the intelligent IoT device, calculate the Euclidean distance between the transformed position and the position of the virtual space anchor point, and obtain the virtual - physical mapping error, which is used to characterize the accuracy of the current mapping relationship;

[0157] Calculate the cumulative drift amount of the position information of the intelligent IoT device and the position information of the corresponding anchor point in the virtual space at adjacent moments, and obtain the time derivative of the cumulative drift amount to get the drift change rate. Combine the drift change rate and the gradient of the virtual - physical mapping error with weights to construct an adaptive compensation vector, where the weighting coefficient is dynamically adjusted with the change of the virtual - physical mapping error;

[0158] Update the space transformation matrix based on the compensation vector to obtain an optimized virtual - physical mapping relationship.

[0159] This embodiment provides a method for establishing a verification mechanism for the virtual-real mapping relationship, which is used to verify and update the mapping relationship between the position information of the intelligent Internet of Things device and the virtual anchor points, and eliminate the spatio-temporal drift of virtual-real fusion. The method includes the following steps:

[0160] Based on the position information of the intelligent Internet of Things device and the position information of the corresponding anchor points in the virtual space, construct a spatial transformation matrix including a rotation matrix and a translation vector. This spatial transformation matrix is used to establish the initial mapping relationship from the physical space to the virtual space. Specifically, the spatial transformation matrix can be constructed in the following way:

[0161] Obtain the three-dimensional coordinates (x, y, z) of the intelligent Internet of Things device in the physical space and the three-dimensional coordinates (x', y', z') of the corresponding virtual anchor points in the virtual space. Select at least 3 groups of corresponding point pairs to construct a system of linear equations. Use the least squares method to solve the system of equations to obtain the rotation matrix R and the translation vector T. Combine R and T into a 4x4 homogeneous transformation matrix, which is the required spatial transformation matrix M.

[0162] For example, assume there are 3 groups of corresponding point pairs:

[0163] Physical space point 1 (1, 2, 3), virtual space point 1 (1.1, 2.2, 3.3);

[0164] Physical space point 2 (4, 5, 6), virtual space point 2 (4.4, 5.5, 6.6);

[0165] Physical space point 3 (7, 8, 9), virtual space point 3 (7.7, 8.8, 9.9);

[0166] Through solution, the spatial transformation matrix M can be obtained: 1.1 0.0 0.0 0.1; 0.0 1.1 0.0 0.2; 0.0 0.0 1.1 0.3; 0.0 0.0 0.0 1.0;

[0171] Next, use this spatial transformation matrix to transform the position information of the intelligent Internet of Things device, calculate the Euclidean distance between the transformed position and the position of the virtual space anchor point, and obtain the virtual-real mapping error. This error is used to characterize the accuracy of the current mapping relationship. The specific steps are as follows:

[0172] The physical space coordinates (x, y, z) of the intelligent IoT device are extended to homogeneous coordinates (x, y, z, 1), multiplied by the space transformation matrix M, and the transformed virtual space coordinates (x', y', z', w') are obtained. The homogeneous coordinates are then converted back to three-dimensional coordinates (x' / w', y' / w', z' / w'). Calculate the Euclidean distance between this coordinate and the actual virtual anchor point coordinate, which is the virtual-real mapping error e.

[0173] For example, for a device with coordinates (10, 20, 30) in the physical space, perform the transformation using the above transformation matrix M:

[0174] (10, 20, 30, 1) * M = (11.1, 22.2, 33.3, 1.0);

[0175] The transformed coordinates are (11.1, 22.2, 33.3);

[0176] Assume the corresponding actual virtual anchor point coordinates are (11.0, 22.0, 33.0), then the Euclidean distance can be calculated:

[0177] e = sqrt((11.1 - 11.0)^2 + (22.2 - 22.0)^2 + (33.3 - 33.0)^2) ≈ 0.37;

[0178] Calculate the cumulative drift of the position information of the intelligent IoT device and the position information of the corresponding anchor point in the virtual space at adjacent times, and obtain the drift rate of change by taking the time derivative of the cumulative drift. Weightedly combine the drift rate of change and the gradient of the virtual-real mapping error to construct a compensation vector with adaptive characteristics, where the weighting coefficients are dynamically adjusted according to the change of the virtual-real mapping error.

[0179] The specific implementation method is as follows:

[0180] Record the physical space coordinates and virtual space coordinates at consecutive multiple times (such as t1, t2, t3). Calculate the position differences between adjacent times to obtain the drift amounts. Accumulate the drift amounts to obtain the cumulative drift D. Calculate the rate of change of the cumulative drift with respect to time, that is, the drift rate of change v = dD / dt.

[0181] Calculate the partial derivatives of the virtual-real mapping error e with respect to the parameters in the space transformation matrix M to obtain the error gradient g. Weightedly combine the drift rate of change v and the error gradient g:

[0182] Compensation vector C = α * v + β * g;

[0183] Where α and β are weighting coefficients, which can be dynamically adjusted according to the magnitude of the error e:

[0184] When e is small, increase α and decrease β to maintain stability;

[0185] When e is large, decrease α and increase β to accelerate convergence.

[0186] For example, the following adjustment strategy can be adopted:

[0187] α = 1 / (1 + e^2);

[0188] β = e^2 / (1 + e^2);

[0189] Suppose at a certain moment, the drift change rate v = (0.1, 0.2, 0.3), the error gradient g = (1.0, 2.0, 3.0), and the error e = 0.5, then:

[0190] α ≈ 0.8, β ≈ 0.2;

[0191] The compensation vector C ≈ (0.28, 0.56, 0.84);

[0192] Finally, update the spatial transformation matrix based on the compensation vector to obtain the optimized virtual-real mapping relationship. The update method is as follows:

[0193] Decompose the compensation vector C into a rotation component Cr and a translation component Ct. Construct an incremental rotation matrix δR according to Cr and an incremental translation vector δT according to Ct. Multiply δR and δT by the original spatial transformation matrix M to obtain the updated matrix M'.

[0194] For example, assume the compensation vector C = (0.28, 0.56, 0.84), and the incremental rotation matrix δR and the incremental translation vector δT can be constructed as follows:

[0195] δR ≈ [1.0000 -0.0084 0.0056];

[0196] [0.0084 1.0000 -0.0028];

[0197] [-0.0056 0.0028 1.0000];

[0198] δT ≈ [0.28, 0.56, 0.84];

[0199] Multiply δR and δT by the original matrix M to obtain the updated matrix M':

[0200] M' ≈ [1.0984 -0.0092 0.0062 0.3908];

[0201] [0.0092 1.1092 -0.0031 0.7672];

[0202] [-0.0062 0.0031 1.1062 1.1436]; [0.0000 0.0000 0.0000 1.0000];

[0204] Through the above steps, the verification and dynamic update of the virtual-real mapping relationship are realized, effectively eliminating the spatio-temporal drift in the virtual-real fusion process. This method has self-adaptability and can dynamically adjust the compensation strategy according to the error size, ensuring the accuracy and stability of the virtual-real mapping. In practical applications, the above steps can be executed periodically to continuously optimize the virtual-real mapping relationship, providing a reliable spatial positioning basis for virtual-real fusion applications.

[0205] In an alternative embodiment, feeding back the interaction information in the virtual scene to the intelligent IoT device and establishing a dynamic adjustment function including a position constraint term and a state constraint term to achieve two-way interaction between the physical space and the metaverse virtual space includes:

[0206] Establishing a dynamic adjustment function including a position constraint term and a state constraint term, where the position constraint term ensures that the position of the virtual object is synchronized with the mapped position of the intelligent IoT device, and the state constraint term ensures that the motion state of the virtual object matches the actual state of the intelligent IoT device;

[0207] Using an iterative optimization algorithm with adaptive weights to solve the dynamic adjustment function, where the weight coefficient is dynamically updated according to the time series of the position error;

[0208] Applying the optimized position and state parameters to the virtual objects in the virtual scene to achieve synchronous interaction between the virtual objects and the physical devices.

[0209] This embodiment provides a method for feeding back the interaction information in the virtual scene to the intelligent IoT device and establishing a dynamic adjustment function including a position constraint term and a state constraint term to achieve two-way interaction between the physical space and the metaverse virtual space.

[0210] Establishing a dynamic adjustment function including a position constraint term and a state constraint term. This function consists of two main constraint terms: the position constraint term ensures that the position of the virtual object is synchronized with the mapped position of the intelligent IoT device; the state constraint term ensures that the motion state of the virtual object matches the actual state of the intelligent IoT device.

[0211] The specific implementation method of the position constraint term is as follows: The system obtains the real-time position coordinates (Xp, Yp, Zp) of the intelligent Internet of Things device in the physical space, and at the same time records the current position coordinates (Xv, Yv, Zv) of the virtual object in the virtual space. The system calculates the position difference vector D = (Xp - Xv, Yp - Yv, Zp - Zv) between the two, and sets the position constraint term Cp so that its value is proportional to the modulus of the difference vector D. When the position difference exceeds the preset threshold (for example, 10 centimeters), the system will increase the weight of the position constraint term to prompt the virtual object to approach the mapped position of the physical device faster.

[0212] The specific implementation method of the state constraint term is as follows: The system obtains the actual motion state parameters of the intelligent Internet of Things device, including the velocity vector Vp = (Vxp, Vyp, Vzp), the acceleration vector Ap = (Axp, Ayp, Azp), and the working state parameters of the device (such as switch state, brightness, temperature, etc.). At the same time, the system records the current motion state parameters of the virtual object, including the velocity vector Vv = (Vxv, Vyv, Vzv), the acceleration vector Av = (Axv, Ayv, Azv), and the corresponding virtual state parameters. The system calculates the state difference between the two and sets the state constraint term Cs so that its value is proportional to the state difference. For example, when the moving speed of the physical device is 2 m / s and the moving speed of the virtual object is 1.5 m / s, the system will increase the weight of the state constraint term to prompt the virtual object to adjust its speed to 2 m / s faster.

[0213] The dynamic adjustment function F can be expressed as the weighted sum of the position constraint term Cp and the state constraint term Cs: F = Wp×Cp + Ws×Cs, where Wp and Ws are the weight coefficients of the position constraint term and the state constraint term respectively. In the initial state, Wp = 0.6 and Ws = 0.4 can be set, indicating that the priority of position synchronization is slightly higher than that of state synchronization.

[0214] An iterative optimization algorithm with adaptive weights is used to solve the dynamic adjustment function. The specific steps of this algorithm are as follows:

[0215] Initialize the weight of the position constraint term Wp = 0.6 and the weight of the state constraint term Ws = 0.4. The system sets the maximum number of iterations to 100 times, and the convergence threshold to 0.01 (indicating that convergence is considered when both the position error and the state error are less than 0.01).

[0216] In each iteration, the value of the dynamic adjustment function F is calculated based on the current weight values, and the position and state parameters of the virtual object are adjusted according to the gradient direction of F. Specifically, the system calculates the partial derivatives of F with respect to the position coordinates (Xv, Yv, Zv) and state parameters of the virtual object, and updates these parameters along the direction of gradient descent. For example, if the partial derivative of F with respect to Xv is positive, indicating that increasing Xv can reduce the value of F, the system will appropriately increase the value of Xv, and the adjustment step size can be set to 0.05.

[0217] After each iteration, the system calculates the updated position error and state error. If the position error is greater than 0.1 for 5 consecutive iterations, the system will increase the weight of the position constraint term. For example, it will increase Wp from 0.6 to 0.7, and at the same time reduce Ws from 0.4 to 0.3, ensuring that Wp + Ws = 1. Conversely, if the state error is greater than the preset threshold (such as the speed error is greater than 0.5 m / s) for 5 consecutive iterations, the system will increase the weight of the state constraint term. For example, it will increase Ws from 0.4 to 0.5, and at the same time reduce Wp from 0.6 to 0.5.

[0218] The system will also dynamically update the weight coefficients according to the time series of the position error. Specifically, the system records the position error values of the last 10 iterations. If the error shows a continuous increasing trend (such as increasing for 3 consecutive times), the system will increase the weight Wp of the position constraint term by 0.05; if the error shows a continuous decreasing trend, the system will keep the current weight unchanged. Similarly, the system will also dynamically adjust the weight Ws of the state constraint term according to the time series of the state error.

[0219] The iteration process continues until any of the following conditions is met: reaching the maximum number of iterations (100 times); both the position error and the state error are less than the convergence threshold (0.01); the error change rate of 10 consecutive iterations is less than 0.001.

[0220] The system applies the optimized position and state parameters to the virtual object in the virtual scene to achieve synchronous interaction between the virtual object and the physical device. Specifically, the system updates the position coordinates of the virtual object to the optimized (Xv, Yv, Zv), updates the velocity vector of the virtual object to the optimized Vv, updates the acceleration vector of the virtual object to the optimized Av, and updates other state parameters of the virtual object.

[0221] In practical applications, taking intelligent lighting fixtures as an example, when a user moves a lighting fixture to a new position in the physical space (for example, from the center of the living room to a corner of the living room, and the physical coordinates change from (3.0, 2.5, 1.8) to (5.5, 0.5, 1.8)), the system detects the position change and activates the dynamic adjustment function. After 25 iterations of optimization, the corresponding virtual lighting fixture in the virtual scene smoothly transitions from the original position (30, 25, 18) to the new position (55, 5, 18).

[0222] The precise positioning system for the virtual-real integration of intelligent IoT devices and the metaverse according to the embodiments of the present invention, as Figure 3 shown, includes:

[0223] A first unit, configured to obtain the position information of multiple intelligent IoT devices and the spatial relationship information of the intelligent IoT devices, where the position information includes geographical coordinates and attitude data; based on the position information and the spatial relationship information, collaboratively construct a global positioning knowledge graph through an asynchronous iterative optimization algorithm to achieve the collaborative high-precision positioning of multiple intelligent IoT devices and obtain a positioning result;

[0224] A second unit, configured to establish a virtual scene corresponding to the physical space in the metaverse virtual space; determine virtual anchor points in the virtual scene, where the virtual anchor points correspond to physical reference objects in the physical space;

[0225] A third unit, configured to establish a virtual-real mapping relationship verification mechanism to verify and update the mapping relationship between the position information of the intelligent IoT devices and the virtual anchor points, and eliminate the spatio-temporal drift of virtual-real integration;

[0226] A fourth unit, configured to dynamically adjust the positions and states of virtual objects in the virtual scene according to the positioning result and the virtual-real mapping relationship verification mechanism;

[0227] A fifth unit, configured to feedback the interaction information in the virtual scene to the intelligent IoT devices, and establish a dynamic adjustment function including position constraint terms and state constraint terms to realize the two-way interaction between the physical space and the metaverse virtual space.

[0228] In the third aspect of the embodiments of the present invention,

[0229] Provided is an electronic device, including:

[0230] A processor;

[0231] A memory for storing instructions executable by the processor;

[0232] Wherein, the processor is configured to call the instructions stored in the memory to execute the method described above.

[0233] In a fourth aspect of the embodiments of the present invention,

[0234] a computer-readable storage medium is provided, on which computer program instructions are stored, and when the computer program instructions are executed by a processor, the foregoing method is implemented.

[0235] The present invention may be a method, an apparatus, a system, and / or a computer program product. The computer program product may include a computer-readable storage medium on which computer-readable program instructions for performing various aspects of the present invention are loaded.

[0236] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A precise positioning method for the virtual-real fusion of intelligent Internet of Things devices and the metaverse, characterized in that, Including: Obtain the location information of multiple intelligent Internet of Things devices and the spatial relationship information of the intelligent Internet of Things devices, where the location information includes geographical coordinates and attitude data; Based on the location information and the spatial relationship information, collaboratively construct a global positioning knowledge graph through an asynchronous iterative optimization algorithm to achieve collaborative high-precision positioning of multiple intelligent Internet of Things devices. The obtained positioning results include: Construct a global knowledge graph based on the location information and the spatial relationship information. The global knowledge graph includes a node set, an edge set, and a weight matrix, where the node set represents the location nodes of multiple intelligent Internet of Things devices, the edge set represents the spatial relationships between the intelligent Internet of Things devices, and the weight matrix represents the spatial constraint strength between the intelligent Internet of Things devices; Establish a local consistency constraint model according to the node set and the edge set of the global knowledge graph. The local consistency constraint model includes an Euclidean distance term and a regularization term of the device location, and dynamically balance the Euclidean distance term and the regularization term through the weight coefficients in the weight matrix to obtain a local constraint optimization objective; Use the alternating direction multiplier method to decompose the local constraint optimization objective into multiple local sub-problems, and obtain an initial collaborative optimization location result by iteratively optimizing the local variables, global variables, and dual variables of each local sub-problem; Input the initial collaborative optimization location result into a graph optimization framework, construct an optimization function including a position error term and an information matrix, and optimize the initial collaborative optimization location result based on the optimization function to obtain the collaborative positioning result of multiple intelligent Internet of Things devices; Establish a virtual scene corresponding to the physical space in the metaverse virtual space; determine virtual anchor points in the virtual scene, where the virtual anchor points correspond to physical reference objects in the physical space; Establish a virtual-real mapping relationship verification mechanism to verify and update the mapping relationship between the location information of the intelligent Internet of Things devices and the virtual anchor points, and eliminate the spatio-temporal drift of virtual-real fusion; Dynamically adjust the positions and states of virtual objects in the virtual scene according to the positioning results and the virtual-real mapping relationship verification mechanism; Feedback the interaction information in the virtual scene to the intelligent Internet of Things devices, and establish a dynamic adjustment function including a position constraint term and a state constraint term to achieve two-way interaction between the physical space and the metaverse virtual space.

2. The method according to claim 1, wherein Use the alternating direction multiplier method to decompose the local constraint optimization objective into multiple local sub-problems, and obtain an initial collaborative optimization location result by iteratively optimizing the local variables, global variables, and dual variables of each local sub-problem, including: Each local sub-problem corresponds to the location optimization of an intelligent Internet of Things device, where each local sub-problem includes: local variables for characterizing the location of a single intelligent Internet of Things device, global variables for coordinating the location consistency of multiple intelligent Internet of Things devices, and dual variables for ensuring convergence. Based on a preset step size parameter, a three-step iterative strategy is adopted to solve the local sub-problem: First, optimize the local variables by fixing the global variables and the dual variables, then optimize the global variables by fixing the local variables and the dual variables, and finally update the dual variables. Repeat the iteration until the position difference between two adjacent iterations is less than the preset deviation threshold to obtain the initial co-optimization position result.

3. The method according to claim 1, characterized in that Input the initial co-optimization position result into the graph optimization framework, construct an optimization function including a position error term and an information matrix, and optimize the initial co-optimization position result based on the optimization function to obtain the collaborative positioning results of multiple intelligent IoT devices, including: Decompose the optimization function into a position optimization sub-problem and a weight optimization sub-problem, where the position optimization sub-problem iteratively optimizes the positions of the intelligent IoT devices by the least squares method, and the weight optimization sub-problem calculates the weight coefficients of the information matrix based on the Mahalanobis distance; Adopt an adaptive step size gradient descent algorithm to alternately solve the position optimization sub-problem and the weight optimization sub-problem. Dynamically adjust the optimization step size according to the convergence degree of the position error term in each iteration, and use the information matrix to suppress the influence of abnormal observations on the optimization result until the preset number of iterations is satisfied to obtain the collaborative positioning results of multiple intelligent IoT devices.

4. The method according to claim 1, wherein Establish a virtual-real mapping relationship verification mechanism to verify and update the mapping relationship between the position information of the intelligent IoT device and the virtual anchor points, and eliminate the spatio-temporal drift of virtual-real fusion, including: Based on the position information of the intelligent IoT device and the position information of the corresponding anchor points in the virtual space, construct a space transformation matrix including a rotation matrix and a translation vector, and the space transformation matrix is used to establish an initial mapping relationship from the physical space to the virtual space; Use the space transformation matrix to transform the position information of the intelligent IoT device, calculate the Euclidean distance between the transformed position and the position of the virtual space anchor point to obtain the virtual-real mapping error, and the virtual-real mapping error is used to characterize the accuracy of the current mapping relationship; Calculate the cumulative drift amount of the position information of the intelligent IoT device and the position information of the corresponding anchor points in the virtual space at adjacent times, and obtain the drift change rate by taking the time derivative of the cumulative drift amount. Weightedly combine the drift change rate with the gradient of the virtual-real mapping error to construct an adaptive compensation vector, where the weighting coefficient is dynamically adjusted according to the change of the virtual-real mapping error; Update the space transformation matrix based on the compensation vector to obtain an optimized virtual-real mapping relationship.

5. The method according to claim 1, characterized in that, Feedback the interaction information in the virtual scene to the intelligent IoT device, and establish a dynamic adjustment function including a position constraint term and a state constraint term to realize the two-way interaction between the physical space and the metaverse virtual space, including: Establish a dynamic adjustment function including a position constraint term and a state constraint term. The position constraint term ensures that the position of the virtual object is synchronized with the mapped position of the intelligent IoT device, and the state constraint term ensures that the motion state of the virtual object matches the actual state of the intelligent IoT device; An iterative optimization algorithm with adaptive weights is used to solve the dynamic adjustment function, where the weight coefficients are dynamically updated according to the time series of the position error; Apply the optimized position and state parameters to the virtual objects in the virtual scene to achieve synchronous interaction between the virtual objects and the physical devices.

6. An accurate positioning system for the virtual-real integration of intelligent Internet of Things devices and the metaverse, which is used to implement the method described in any one of claims 1-5, characterized in that, It includes: The first unit is used to obtain the position information of multiple intelligent IoT devices and the spatial relationship information of the intelligent IoT devices. The position information includes geographical coordinates and attitude data; based on the position information and the spatial relationship information, a global positioning knowledge graph is collaboratively constructed through an asynchronous iterative optimization algorithm to achieve collaborative high-precision positioning of multiple intelligent IoT devices and obtain a positioning result; The second unit is used to establish a virtual scene corresponding to the physical space in the metaverse virtual space; determine virtual anchors in the virtual scene, and the virtual anchors correspond to physical reference objects in the physical space; The third unit is used to establish a virtual-real mapping relationship verification mechanism to verify and update the mapping relationship between the position information of the intelligent IoT device and the virtual anchor, and eliminate the spatio-temporal drift of virtual-real fusion; The fourth unit is used to dynamically adjust the position and state of the virtual objects in the virtual scene according to the positioning result and the virtual-real mapping relationship verification mechanism; The fifth unit is used to feedback the interaction information in the virtual scene to the intelligent IoT device, establish a dynamic adjustment function including position constraint terms and state constraint terms to achieve two-way interaction between the physical space and the metaverse virtual space.

7. An electronic device, characterized in that, It includes: A processor; A memory for storing instructions executable by the processor; Wherein, the processor is configured to call the instructions stored in the memory to execute the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, the method according to any one of claims 1 to 5 is implemented.

Images