Non-magnetic stripe magnetic navigation method and system using matrix tag radio frequency signal strength

Abstract

The application belongs to the technical field of navigation, and specifically discloses a magnetic stripe-free magnetic navigation method and system using matrix tag radio frequency signal strength, which comprises the following steps: receiving radio frequency signals broadcast by a radio frequency tag matrix arranged in a navigation area; collecting the received signal strength RSSI, angle of arrival AOA and phase difference PDOA parameters of at least three radio frequency tag nodes in real time through a multi-antenna array, and calculating the real-time position information of a mobile device based on the parameters; generating virtual magnetic stripe data according to a preset path and through a signal field strength-space coordinate-magnetic navigation protocol three-layer mapping mechanism; converting the virtual magnetic stripe data into analog magnetic field signal output conforming to a standard magnetic navigation sensor protocol; and performing path tracking and dynamic deviation correction control on the mobile device according to the analog magnetic field signal based on a model predictive control algorithm. The application does not need to rely on a physical magnetic stripe, can dynamically adjust a path, and is easy to deploy and maintain.

Description

Technical Field

[0001] This application belongs to the field of navigation technology, and more specifically, relates to a magnetic navigation method and system without magnetic stripes that utilizes the radio frequency signal strength of matrix tags. Background Technology

[0002] Currently, magnetic navigation technology is widely used in industrial automation, smart warehousing, and other scenarios. Traditional magnetic navigation devices require laying physical magnetic strips on the ground to provide navigation paths for mobile devices. However, laying magnetic strips on the ground has many drawbacks: on the one hand, the construction process is cumbersome, requiring ground pretreatment and consuming a lot of manpower and time; on the other hand, once the magnetic strips are laid, the path is fixed, and if the navigation route needs to be adjusted, the original magnetic strips must be removed and re-laid, resulting in poor flexibility and difficulty in adapting to dynamically changing application scenarios; in addition, after long-term use, the magnetic strips are susceptible to wear and tear, and interference from external magnetic fields, leading to a decrease in navigation accuracy and increasing maintenance costs.

[0003] Therefore, how to provide a navigation solution that does not rely on physical magnetic strips, can dynamically adjust the path, and is easy to deploy and maintain is an urgent problem to be solved. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the purpose of this application is to provide a magnetic navigation method and system that utilizes the radio frequency signal strength of matrix tags without relying on physical magnetic stripes, can dynamically adjust the path, and is easy to deploy and maintain.

[0005] To achieve the above objectives, in a first aspect, this application provides a magnetic navigation method without magnetic stripes that utilizes the radio frequency signal strength of matrix tags, comprising the following steps: S10, receive radio frequency signals broadcast by the radio frequency tag matrix deployed in the navigation area, each radio frequency signal containing a unique spatial coordinate code of its corresponding radio frequency tag node; S20, the received signal strength RSSI, angle of arrival AOA and phase difference PDOA parameters of at least three radio frequency tag nodes are collected in real time through the multi-antenna array on the mobile device, and the real-time location information of the mobile device is calculated based on the parameters; S30, virtual magnetic stripe data is generated according to the preset path and through the three-layer mapping mechanism of signal field strength-spatial coordinates-magnetic navigation protocol. The virtual magnetic stripe data is used to convert the preset path into gradient direction control instructions for the radio frequency signal strength field. S40, convert the virtual magnetic stripe data into an analog magnetic field signal output that conforms to the standard magnetic navigation sensor protocol; S50, based on the model predictive control algorithm, performs path tracking and dynamic correction control on the mobile device according to the simulated magnetic field signal.

[0006] The magnetic navigation method without magnetic stripes, utilizing the radio frequency signal strength of matrix tags provided in this application, has the following advantages: First, by deploying a matrix of radio frequency tags to completely replace physical magnetic strips on the ground, the cumbersome ground laying and modification projects can be completely eliminated, resulting in a revolutionary reduction in deployment cycle and a significant decrease in overall cost, while fundamentally solving the maintenance problem of magnetic strip wear. Second, relying on the three-layer mapping mechanism of signal field strength-spatial coordinates-magnetic navigation protocol, changes in the path can be added by redefining the tag activation area and virtual magnetic strip data through software, without any physical modification. This supports online dynamic topology reconstruction and rapid path switching, meeting the needs of highly flexible production. Finally, by generating simulated magnetic field signals that are fully compatible with standard magnetic navigation sensor protocols, existing traditional magnetic navigation equipment can be directly connected to this system without hardware modifications, effectively protecting existing investments and achieving a smooth upgrade transition from traditional navigation to flexible navigation.

[0007] As a further preferred embodiment, in step S10, the matrix adopts a honeycomb topology structure with equilateral triangles or regular hexagons closely teeming together, the spacing between adjacent RFID tag nodes is d=0.5m~2m, and each cellular unit contains 3 anchor tags to form a positioning reference surface; the rate of change of the signal intensity gradient of the RFID signal in space is greater than or equal to 15dB / m, and the resolution error is less than 3 cm.

[0008] As a further preferred embodiment, in step S20, when calculating the real-time location information of the mobile device based on the parameters, a signal strength-spatial relationship model is established and applied: S=K·P t ·d -n ·e^(-αd) In the formula, S is the received signal strength, K is a constant, and P t Where is the transmit power, d is the distance, n is the path loss exponent, and α is the environmental attenuation coefficient. n and α can be adaptively calibrated.

[0009] As a further preferred option, step S20 also includes: The acquired received signal strength RSSI parameters are filtered using a sliding window filtering and Kalman filtering fusion algorithm to select the effective signal range, which is [S min +σ, S max -σ], where S min S is the minimum received signal strength. max The maximum received signal strength is σ, and the standard deviation of the ambient noise is σ. Multipath interference is suppressed by using the STAP algorithm with spatiotemporal adaptive processing.

[0010] As a further preferred embodiment, in step S30, the virtual magnetic stripe data B(x) is obtained through magnetic field feature modeling, and the modeling formula is: B(x) = B0·exp(-x² / 2w²) In the formula, B0 is the magnetic field center strength, w is the effective width of the equivalent magnetic strip, and x is the lateral distance from the center line.

[0011] As a further preferred option, in step S30, an improved method is adopted. The algorithm performs dynamic path planning to generate the virtual magnetic stripe data, the improvement The cost function of the algorithm is f=g+h+λ·S, where g is the path cost, h is the heuristic function, S is the signal strength confidence weight, and λ is adjustable from 0.1 to 0.5. It supports global path replanning within 30 minutes and switching virtual branches within 5 seconds.

[0012] As a further preferred embodiment, step S40 specifically involves: using an integrated Hall effect sensor array, outputting a 0-5V analog voltage signal via a digital-to-analog converter (DAC) with a signal-to-noise ratio greater than or equal to 60dB, and employing a PID feedforward compensation algorithm with a response time of less than 10ms; and outputting a correction pulse signal when a lateral deviation Δx greater than 2cm is detected.

[0013] As a further preferred option, step S50 specifically involves: adopting a dual closed-loop control architecture, with an outer loop path tracking loop bandwidth of 5Hz and an inner loop wheel speed servo loop bandwidth of 50Hz, predicting the motion trajectory 200ms in advance based on the model predictive control (MPC) algorithm, with a maximum tracking error of less than 1cm, and supporting local path replanning when the signal strength drops by more than 20dB, the cumulative positioning error is greater than 5cm, or a new target point is issued by the host computer, with replanning taking less than 100ms.

[0014] As a further preferred embodiment, in step S10, each of the radio frequency tag nodes is configured with an environmental disturbance detection unit. When metal obstruction or electromagnetic interference is detected, the unit automatically switches between a transmit power of 10-100mW and a dual-band frequency hopping mode of 433MHz / 915MHz to make the signal field strength stability CV value less than 5%.

[0015] Secondly, this application provides a magnetic navigation system without magnetic stripes, comprising the steps of implementing the method described in any one of the above, including: The signal receiving module is used to receive radio frequency signals broadcast by the radio frequency tag matrix deployed in the navigation area. Each radio frequency signal contains a unique spatial coordinate code of its corresponding radio frequency tag node. The signal detection and processing module is used to collect the received signal strength RSSI, angle of arrival (AOA), and phase difference (PDOA) parameters of at least three radio frequency tag nodes in real time through a multi-antenna array on the mobile device, and calculate the real-time location information of the mobile device based on the parameters. The virtual route generation module is used to generate virtual magnetic stripe data based on a preset path and through a three-layer mapping mechanism of signal field strength-spatial coordinates-magnetic navigation protocol. The virtual magnetic stripe data is used to convert the preset path into gradient direction control commands for the radio frequency signal strength field. The magnetic navigation simulation module is used to convert the virtual magnetic stripe data into a simulated magnetic field signal output that conforms to the standard magnetic navigation sensor protocol; The motion control module is used to perform path tracking and dynamic correction control of the mobile device based on the model predictive control algorithm and the simulated magnetic field signal.

[0016] It is understandable that the beneficial effects of the second aspect mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here. Attached Figure Description

[0017] Figure 1 This is a flowchart of the magnetic navigation method without magnetic stripes that utilizes the radio frequency signal strength of matrix tags provided in this application; Figure 2 This is an architecture diagram of the magnetic navigation method without magnetic stripes provided in this application; Figure 3 This is a comparison curve of RSSI signal stability under different filtering methods provided in the embodiments of this application. Figure 4 This is a comparison curve of the multipath interference effect before and after enabling the STAP algorithm provided in this application embodiment; Figure 5 This is a bar chart comparing the path operation efficiency of the traditional solution and the solution of this application in the embodiments of this application; Figure 6 This is a line graph comparing the AGV operating efficiency of the traditional solution provided in this application and the solution in this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0019] like Figure 1 As shown, this application provides a magnetic navigation method without magnetic stripes that utilizes the radio frequency signal strength of matrix tags, including steps S10 to S50, which are detailed below: Step S10: Receive radio frequency signals broadcast by the radio frequency tag matrix deployed within the navigation area. Each radio frequency signal contains a unique spatial coordinate code for its corresponding radio frequency tag node. This step constructs the physical foundation layer of the system by receiving signals broadcast by several RFID tag nodes deployed in a cellular topology. The unique spatial coordinate code built into each tag forms a programmable, dynamic electronic map, providing a spatial reference for subsequent precise positioning and virtual path generation, thus eliminating the need to lay any physical magnetic strips on the ground.

[0020] Step S20: The received signal strength RSSI, angle of arrival (AOA), and phase difference (PDOA) parameters of at least three RFID tag nodes are collected in real time through the multi-antenna array on the mobile device, and the real-time location information of the mobile device is calculated based on the parameters.

[0021] This step utilizes a multi-antenna array to simultaneously acquire multi-dimensional radio frequency signal parameters. Specifically, it obtains real-time centimeter-level location information of the mobile device by establishing and solving a correlation model between signal strength and space. This method can achieve a positioning accuracy of ±1cm, providing core data support for generating high-precision virtual navigation commands.

[0022] Step S30: Based on the preset path and through the three-layer mapping mechanism of signal field strength-spatial coordinates-magnetic navigation protocol, virtual magnetic stripe data is generated. The virtual magnetic stripe data is used to convert the preset path into gradient direction control commands for the radio frequency signal strength field.

[0023] This step is the core of achieving "navigation protocols compatible with magnetic stripes but without physical magnetic stripes". Through a three-layer mapping mechanism, the preset path defined in the software can be mapped to a radio frequency signal intensity field gradient change model overlaid on real spatial coordinates, thereby generating virtual magnetic stripe data for controlling the mobile device to travel in a preset direction, simulating the magnetic field distribution navigation characteristics of traditional physical magnetic stripes.

[0024] Step S40: Convert the virtual magnetic stripe data into an analog magnetic field signal output that conforms to the standard magnetic navigation sensor protocol.

[0025] This step converts the virtual magnetic stripe data generated in step S30 into an analog signal that conforms to the standards of traditional magnetic navigation sensors in terms of electrical characteristics and communication protocols through protocol conversion. This method allows sensors and control systems of traditional magnetic navigation devices (such as AGVs) that rely on physical magnetic stripes to directly recognize and process this analog signal without any hardware modifications, achieving smooth compatibility.

[0026] Step S50: Based on the model predictive control algorithm, the mobile device is subjected to path tracking and dynamic correction control according to the simulated magnetic field signal.

[0027] This step utilizes model predictive control algorithms to process simulated magnetic field signals, enabling high-precision closed-loop control of mobile device motion. This method can predict the trajectory in advance and perform dynamic correction, ensuring that the mobile device travels stably and accurately along the virtual path, and supports rapid dynamic response when encountering interference or needing to change routes.

[0028] The magnetic navigation method without magnetic stripes, utilizing the radio frequency signal strength of matrix tags provided in this application, has the following advantages: First, by deploying a matrix of radio frequency tags to completely replace physical magnetic strips on the ground, the cumbersome ground laying and modification projects can be completely eliminated, resulting in a revolutionary reduction in deployment cycle and a significant decrease in overall cost, while fundamentally solving the maintenance problem of magnetic strip wear. Second, relying on the three-layer mapping mechanism of signal field strength-spatial coordinates-magnetic navigation protocol, changes in the path can be added by redefining the tag activation area and virtual magnetic strip data through software, without any physical modification. This supports online dynamic topology reconstruction and rapid path switching, meeting the needs of highly flexible production. Finally, by generating simulated magnetic field signals that are fully compatible with standard magnetic navigation sensor protocols, existing traditional magnetic navigation equipment can be directly connected to this system without hardware modifications, effectively protecting existing investments and achieving a smooth upgrade transition from traditional navigation to flexible navigation.

[0029] Based on the same inventive concept, this application also provides a magnetic navigation system without a magnetic stripe, comprising: The signal receiving module is used to receive radio frequency signals broadcast by the radio frequency tag matrix deployed in the navigation area. Each radio frequency signal contains a unique spatial coordinate code of its corresponding radio frequency tag node. The signal detection and processing module is used to collect the received signal strength RSSI, angle of arrival (AOA), and phase difference (PDOA) parameters of at least three RFID tag nodes in real time through the multi-antenna array on the mobile device, and calculate the real-time location information of the mobile device based on these parameters. The virtual route generation module is used to generate virtual magnetic stripe data based on a preset path and through a three-layer mapping mechanism of signal field strength-spatial coordinates-magnetic navigation protocol. The virtual magnetic stripe data is used to convert the preset path into gradient direction control commands for the radio frequency signal strength field. The magnetic navigation simulation module is used to convert virtual magnetic stripe data into simulated magnetic field signal output that conforms to the standard magnetic navigation sensor protocol; The motion control module is used to perform path tracking and dynamic correction control of the mobile device based on the model predictive control algorithm and the simulated magnetic field signal.

[0030] It should be noted that the functions of each module mentioned above can be found in the detailed description of the methods provided above, and will not be repeated here.

[0031] In one embodiment, the technical solution to achieve the above objective can be as follows: This embodiment provides a magnetic navigation system without magnetic stripes that utilizes the radio frequency signal strength of matrix tags. A virtual magnetic stripe navigation environment is constructed through spatial mapping of the radio frequency signal strength field, solving the problem of rigid constraints that traditional magnetic navigation relies on physical magnetic stripes.

[0032] The core innovation of this embodiment lies in establishing a three-layer mapping mechanism of signal field strength-spatial coordinates-magnetic navigation protocol, realizing a breakthrough technical solution of "no physical magnetic strip but compatible with magnetic strip navigation protocol".

[0033] like Figure 2 As shown, the magnetic stripe-free magnetic navigation system provided in this embodiment includes a signal receiving module 100, a signal detection and processing module 200, a virtual route generation module 300, a magnetic navigation simulation module 400, and a motion control module 500.

[0034] The signal receiving module 100 provided in this embodiment is used to receive radio frequency signals broadcast by an RFID matrix deployed within the navigation area. The RFID matrix constitutes the physical foundation layer of the system, employing a close-packed topology of equilateral triangles or regular hexagons. The spacing d between adjacent tags is preferably 0.5-2m, forming a cellular signal coverage unit. Three anchor tags are deployed within each cellular unit to form a positioning reference plane, ensuring a signal strength gradient change rate ≥15dB / m and a spatial coordinate calculation error <3cm.

[0035] In this embodiment, each tag node integrates an environmental disturbance detection unit, which identifies metal obstruction or electromagnetic interference by monitoring the background noise spectrum characteristics. When the interference intensity exceeds a preset threshold, the tag automatically switches the transmission power (adjustable range 10-100mW) and enables a 433MHz / 915MHz dual-band frequency hopping mode to maintain the stability of the signal field strength with a CV value of <5%.

[0036] Each tag has a unique spatial coordinate code (including XYZ three-dimensional coordinate values ​​and regional attribute code). The radio frequency data packets are broadcast synchronously to send coordinate information, creating a dynamic electronic map that can be edited online. The tag nodes can be added, deleted, or their coordinate parameters can be modified through the console software.

[0037] The signal detection and processing module 200 provided in this embodiment is configured with three orthogonal dipole antennas to form a three-dimensional receiving array, and synchronously collects three-dimensional parameters of signal strength RSSI, angle of arrival AOA, and phase difference PDOA, and establishes the following signal strength-space relationship model: S=K·P t ·d -n ·e^(-αd) In the formula, S is the received signal strength (dBm), K is the system constant, and P tLet be the transmit power, d be the distance between the tag and the receiving antenna, n be the path loss exponent, and α be the environmental attenuation coefficient. The parameters n and α are adaptively calibrated online using the least squares method to cope with dynamic environmental changes.

[0038] In this embodiment, the dynamic interval filtering algorithm uses a sliding window filter (window length N=10) to remove instantaneous outliers, and combines it with a Kalman filter to predict the signal intensity change trend, thereby filtering the effective interval [S] in real time. min +σ, S max -σ], where σ is the standard deviation of environmental noise. The system continuously outputs the optimal tag triplet (i.e., the tag with the strongest signal and its two next nearest neighbor tags) for subsequent localization calculation.

[0039] In complex metallic environments, the module employs the Space-Time Adaptive Processing (STAP) algorithm, which constructs a two-dimensional spatiotemporal filter to distinguish between direct waves and multipath reflection signals, ensuring that the vernier positioning accuracy remains within ±2cm.

[0040] The virtual route generation module 300 provided in this embodiment is used to implement the core algorithm layer functions. First, it models the magnetic stripe path features, abstracting the magnetic field distribution of the traditional physical magnetic stripe into a mathematical model: B(x) = B0·exp(-x² / 2w²) Where B0 is the magnetic field center strength, w=5cm is the effective width of the equivalent magnetic strip, and x is the lateral distance from the center line. The virtual route is defined as a continuous sequence of spatial coordinates, ensuring that the direction of the synthetic signal intensity gradient detected by the mobile device is always consistent with the tangent direction of the preset path.

[0041] The dynamic path planning engine adopts an improved The algorithm's cost function is designed as f = g + h + λ·S, where g is the path length cost, h is the heuristic estimation cost, and S is the signal strength confidence weight (λ = 0.1-0.5 adjustable). This algorithm supports global path replanning for sites of tens of thousands of square meters within 30 minutes. Notably, the system supports a "virtual branching" function, allowing users to switch between preset path branches within 5 seconds by activating different area label combinations via software, without any physical modifications.

[0042] The magnetic navigation simulation interface converts virtual route data into virtual magnetic field data frames that conform to standard magnetic navigation sensor protocols (such as CANopenDS406). These data frames contain magnetic field strength B and gradient. Parameters such as B and curvature κ allow the upper-level magnetic navigation simulation module to be directly compatible without modifying the existing AGV control logic.

[0043] The magnetic navigation simulation module 400 provided in this embodiment is responsible for protocol conversion and signal output. It integrates a Hall effect sensor array (3×3 layout, 3cm spacing) to collect the magnetic field parameters generated by the virtual route in real time, and converts them into 0-5V analog voltage signals through a high-speed digital-to-analog converter (DAC), with a signal-to-noise ratio ≥60dB, completely replicating the electromagnetic induction characteristics of the physical magnetic strip.

[0044] The feedback closed-loop control adopts a PID feedforward compensation algorithm, with a system response time of <10ms, accurately simulating the closed-loop dynamic characteristics of "deviation-detection-correction" in magnetic stripe navigation. When a lateral deviation Δx > 2cm is detected, the module outputs a correction pulse signal. The pulse width is linearly related to the deviation, and the proportional coefficient can be adjusted according to the vehicle's dynamic characteristics.

[0045] The module has a built-in Modbus / PROFIBUS / CAN bus adapter, which can be directly connected to various PLCs / motion controllers to achieve plug-and-play replacement of traditional magnetic navigation systems without modifying the upper-level control program.

[0046] The motion control module 500 provided in this embodiment adopts a dual closed-loop control architecture, with an outer loop being a path tracking loop (bandwidth 5Hz) and an inner loop being a wheel speed servo loop (bandwidth 50Hz). The core algorithm is based on model predictive control (MPC), which uses the system dynamics model to predict the motion trajectory 200ms in advance, achieving high-precision control with a maximum tracking error of <1cm.

[0047] The dynamic replanning trigger conditions include: ① a sudden drop in signal strength > 20dB (determined as obstacle occlusion); ② cumulative positioning error > 5cm; ③ the host computer sends a new target point. Once triggered, the system completes local path optimization calculation within 100ms to ensure that the continuity of motion is not affected.

[0048] The module provides a digital twin interface, which can upload device pose, signal strength map and path data to the cloud platform in real time via 4G / 5G or Wi-Fi. It supports offline simulation and parameter optimization based on digital twin technology, and the measured path efficiency can be improved by 15%-25%.

[0049] The magnetic stripe-free magnetic navigation system provided in this embodiment has the following advantages: 1. Revolutionary Deployment: No ground modifications required; deployment cycle reduced from weeks to hours; labels can be used simply by pasting them on; overall costs reduced by over 60%. Label lifespan > 5 years; maintenance-free design completely solves the problem of magnetic strip wear.

[0050] 2. Ultimate Flexibility: Path changes are achieved by redefining the tag activation area through software, supporting online dynamic topology reconstruction and completing path switching within 30 seconds, meeting the flexible production requirements of Industry 4.0. It can support 100+ virtual paths concurrently, with a density 10 times higher than magnetic strips.

[0051] 3. Breakthrough in compatibility: The unique magnetic navigation protocol simulation layer allows traditional magnetic navigation AGVs to be connected to this system without any hardware modifications, protecting existing investments and achieving a smooth upgrade transition.

[0052] 4. Intelligent Enhancement: Based on signal strength fingerprint positioning, its anti-interference capability is 3 times better than that of a single magnetic strip. Multi-sensor fusion enables positioning accuracy to reach ±1cm, a significant improvement over the ±5cm accuracy of traditional magnetic strips. Dynamic obstacle avoidance and path optimization functions expand the application boundaries of magnetic navigation.

[0053] 5. Zero maintenance cost: The system has self-diagnostic function and can remotely monitor the tag battery status (low power warning) and signal quality, enabling predictive maintenance and reducing operation and maintenance costs by 90%.

[0054] The following are specific implementation cases of this application, verifying the technical advantages of the solution through actual test data. The data, technical principles, and actual effects form a complete logical closed loop: Example 1: Application Scenarios of Intelligent Warehousing AGV This system was deployed in a 5000㎡ automated warehouse, with a total of 1200 RFID tags arranged in a regular hexagonal matrix (1m spacing between adjacent tags). The AGV (Automated Guided Vehicle) is equipped with a signal detection and processing module and a magnetic navigation simulation module. The system first establishes a full-field signal strength fingerprint map through tag broadcast signals, and then defines 10 material transport paths (including 3 virtual branch paths) through a virtual route generation module to achieve multi-area material flow coverage.

[0055] During operation, to address the issues of RF signals being susceptible to environmental noise interference and large fluctuations in RSSI parameters, this solution employs a fusion algorithm of "sliding window filtering + Kalman filtering" for signal optimization. The measured data is shown in Table 1 below. As can be seen from Table 1, the fusion filtering algorithm, through the dual effects of eliminating instantaneous outliers via the sliding window and predicting signal change trends via Kalman filtering, narrowed the signal fluctuation range from -85 to -40 dBm (without filtering) to -75 to -48 dBm, reduced the standard deviation to 1.8 dBm, increased the effective signal ratio to 98.2%, and ultimately stabilized the average positioning error at 0.8 cm, providing a high-precision positional foundation for subsequent virtual magnetic stripe data generation. Table 1

[0056] Meanwhile, the warehouse contains obstructions such as metal shelving frames and stacked goods, which can easily generate multipath interference. This solution employs the Space-Time Adaptive Processing (STAP) algorithm, which constructs a two-dimensional spatiotemporal filter to distinguish between direct and reflected waves. The measured data is shown in Table 2 below. The data indicates that after enabling the STAP algorithm, the proportion of multipath signals decreased from 38.6% to 8.2%, the average positioning error improved from 2.5cm to 1.1cm, and the number of errors greater than 2cm accounted for only 1.8%, effectively solving the positioning interference problem in complex environments.

[0057] Table 2

[0058] In dynamic obstacle avoidance scenarios, when temporary stacking of goods in a certain area causes signal attenuation >20dB, the system triggers a local path replanning mechanism, based on improvements. The algorithm completes path optimization within 50ms, and the AGV automatically avoids obstacles. The positioning accuracy is maintained within ±1.2cm throughout the process, which is 22% more efficient than the original magnetic strip navigation solution (which is prone to failure due to magnetic strip obstruction).

[0059] Example 2: Flexible Production Line Transformation Scenario A car assembly workshop needs to upgrade its existing magnetic navigation AGV system. The core requirements are to shorten the path adjustment cycle and reduce the cost of modification and maintenance. This solution retains the original AGV body and PLC controller, only removing the physical magnetic strips on the ground and replacing them with the wireless tag matrix and magnetic navigation simulation module of this invention. The virtual path deployment can be completed within 2 hours through software configuration. The AGV can directly recognize the virtual magnetic strip signals without any hardware modification.

[0060] Traditional physical magnetic stripe solutions rely on ground installation, requiring extensive physical operations for path deployment, modification, and switchover, resulting in extremely low efficiency. This solution, through a three-layer mapping mechanism of "signal field strength - spatial coordinates - magnetic navigation protocol," allows for rapid path adjustment simply by redefining the tag activation area or virtual path parameters in software. The efficiency comparison is shown in the table below. Table 3 data shows a 2160% improvement in path deployment efficiency over a 500㎡ area, a 9600% improvement in efficiency for modifying a single 100m path, and an 8640% improvement in efficiency for switching three virtual switches, significantly meeting the dynamic adjustment needs of flexible production lines.

[0061] Table 3

[0062] In long-term operation testing, efficiency statistics were compiled for 100 material handling tasks (each handling distance 50m), and the results are shown in Table 4 below. Due to its high positioning accuracy and absence of magnetic strip wear failure, this solution reduced the average handling time from 2.8min to 2.2min, increased the overall equipment efficiency (OEE) from 62.5% to 80.5%, and eliminated downtime caused by path failures throughout the entire process, significantly improving stability compared to traditional solutions. Table 4

[0063] Figure 3 is a comparison curve of RSSI signal stability under different filtering methods provided in this embodiment. Figure 3 It can be seen that: the unfiltered curve fluctuates violently, oscillating greatly in the range of -85 to -40 dBm, reflecting strong interference from environmental noise; the sliding window filtered curve shows reduced fluctuation amplitude, concentrated in the range of -78 to -45 dBm, but still exhibits small continuous fluctuations; the fused filtered curve tends to be stable, remaining within the range of -75 to -48 dBm, with no obvious abnormal fluctuations, intuitively reflecting the synergistic optimization effect of dual filtering.

[0064] Figure 4 is a comparison curve of the multipath interference impact before and after enabling the STAP algorithm provided in this embodiment. Figure 4 It can be seen that: the STAP curve without STAP enabled has a large error dispersion, with multiple samples having errors exceeding 2cm, reaching a maximum of 4.2cm, reflecting the superposition interference of multipath signals; the STAP curve with STAP enabled has errors concentrated in the range of 0.5~1.5cm, with small dispersion and no obvious outliers, intuitively demonstrating the suppression effect of the STAP algorithm on multipath interference.

[0065] Figure 5 is a bar chart comparing the path operation efficiency of the traditional solution and the solution of this application provided in this embodiment. Figure 5 It can be seen that: the traditional columnar solution takes significantly longer, with the switching of 3 branch lines taking 43,200 seconds (12 hours), reflecting the inefficiency of physical operation; the columnar solution of this application takes extremely short time, with the longest operation (500㎡ deployment) taking only 7,200 seconds (2 hours), which directly reflects the efficiency advantage of software-based operation.

[0066] Figure 6 is a line graph comparing the AGV operating efficiency of the traditional solution and the solution of this application provided in this embodiment. Figure 6 It can be seen that: the traditional solution curve shows large fluctuations in time consumption, with some handling processes taking more than 3 minutes due to magnetic strip interference, and the overall trend is upward (due to increased magnetic strip wear); the solution curve of this application shows that the time consumption is stable in the range of 2.1~2.3 minutes, with no obvious fluctuations, and the overall process remains stable, which intuitively reflects the long-term stability and efficiency of the solution.

[0067] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A magnetic navigation method without magnetic stripes utilizing the radio frequency signal strength of matrix tags, characterized in that, Includes the following steps: S10, receive radio frequency signals broadcast by the radio frequency tag matrix deployed in the navigation area, each radio frequency signal containing a unique spatial coordinate code of its corresponding radio frequency tag node; S20, the received signal strength RSSI, angle of arrival AOA and phase difference PDOA parameters of at least three radio frequency tag nodes are collected in real time through the multi-antenna array on the mobile device, and the real-time location information of the mobile device is calculated based on the parameters; S30, virtual magnetic stripe data is generated according to the preset path and through the three-layer mapping mechanism of signal field strength-spatial coordinates-magnetic navigation protocol. The virtual magnetic stripe data is used to convert the preset path into gradient direction control commands for the radio frequency signal strength field. S40, convert the virtual magnetic stripe data into an analog magnetic field signal output that conforms to the standard magnetic navigation sensor protocol; S50, based on the model predictive control algorithm, performs path tracking and dynamic correction control on the mobile device according to the simulated magnetic field signal; In step S30, the virtual magnetic stripe data B(x) is obtained through magnetic field feature modeling, and the modeling formula is: B(x) = B0·exp(-x² / 2w²) In the formula, B0 is the magnetic field center strength, w is the effective width of the equivalent magnetic strip, and x is the lateral distance from the center line.

2. The magnetic navigation method without magnetic stripes using the radio frequency signal strength of matrix tags as described in claim 1, characterized in that, In step S10, the matrix adopts a honeycomb topology structure with equilateral triangles or regular hexagons closely teeming together. The spacing between adjacent RFID tag nodes is d = 0.5m to 2m. Each cellular unit contains 3 anchor tags to form a positioning reference plane. The rate of change of the signal intensity gradient of the RFID signal in space is greater than or equal to 15dB / m, and the resolution error is less than 3 cm.

3. The magnetic navigation method without magnetic stripes using the radio frequency signal strength of matrix tags as described in claim 1, characterized in that, In step S20, when calculating the real-time location information of the mobile device based on the parameters, a signal strength-spatial relationship model is established and applied: S=K·P t ·d -n ·e^(-αd) In the formula, S is the received signal strength, K is a constant, and P t Where is the transmit power, d is the distance, n is the path loss exponent, and α is the environmental attenuation coefficient. n and α can be adaptively calibrated.

4. The magnetic navigation method without magnetic stripes using the radio frequency signal strength of matrix tags as described in claim 1, characterized in that, Step S20 also includes: The acquired received signal strength RSSI parameters are filtered using a sliding window filtering and Kalman filtering fusion algorithm to select the effective signal range, which is [S min +σ, S max -σ], where S min S is the minimum received signal strength. max The maximum received signal strength is σ, and the standard deviation of the ambient noise is σ. Multipath interference is suppressed by using the STAP algorithm with spatiotemporal adaptive processing.

5. The magnetic navigation method without magnetic stripes using the radio frequency signal strength of matrix tags as described in claim 1, characterized in that, In step S30, an improved method is adopted. The algorithm performs dynamic path planning to generate the virtual magnetic stripe data, the improvement The cost function of the algorithm is f=g+h+λ·S, where g is the path cost, h is the heuristic function, S is the signal strength confidence weight, and λ is adjustable from 0.1 to 0.

5. It supports global path replanning within 30 minutes and switching virtual branches within 5 seconds.

6. The magnetic navigation method without magnetic stripes using the radio frequency signal strength of matrix tags as described in claim 1, characterized in that, Step S40 specifically involves: using an integrated Hall effect sensor array, outputting a 0-5V analog voltage signal via a digital-to-analog converter (DAC) with a signal-to-noise ratio greater than or equal to 60dB, and employing a PID feedforward compensation algorithm with a response time of less than 10ms. When a lateral deviation Δx greater than 2cm is detected, a correction pulse signal is output.

7. The magnetic navigation method without magnetic stripes using the radio frequency signal strength of matrix tags as described in claim 1, characterized in that, Step S50 specifically involves: adopting a dual closed-loop control architecture, with an outer loop path tracking loop bandwidth of 5Hz and an inner loop wheel speed servo loop bandwidth of 50Hz. Based on the Model Predictive Control (MPC) algorithm, the motion trajectory is predicted 200ms in advance, with a maximum tracking error of less than 1cm. It also supports triggering local path replanning when the signal strength drops by more than 20dB, the cumulative positioning error exceeds 5cm, or a new target point is sent by the host computer. The replanning time is less than 100ms.

8. The magnetic navigation method without magnetic stripes using the radio frequency signal strength of matrix tags as described in claim 1, characterized in that, In step S10, each of the radio frequency tag nodes is configured with an environmental disturbance detection unit. When metal obstruction or electromagnetic interference is detected, the unit automatically switches between a transmit power of 10-100mW and a dual-band frequency hopping mode of 433MHz / 915MHz to make the signal field strength stability CV value less than 5%.

9. A magnetic navigation system without magnetic stripes, used to implement the steps of the method according to any one of claims 1 to 8, characterized in that, include: The signal receiving module is used to receive radio frequency signals broadcast by the radio frequency tag matrix deployed in the navigation area. Each radio frequency signal contains a unique spatial coordinate code of its corresponding radio frequency tag node. The signal detection and processing module is used to collect the received signal strength RSSI, angle of arrival (AOA), and phase difference (PDOA) parameters of at least three radio frequency tag nodes in real time through a multi-antenna array on the mobile device, and calculate the real-time location information of the mobile device based on the parameters. The virtual route generation module is used to generate virtual magnetic stripe data based on a preset path and through a three-layer mapping mechanism of signal field strength-spatial coordinates-magnetic navigation protocol. The virtual magnetic stripe data is used to convert the preset path into gradient direction control commands for the radio frequency signal strength field. The magnetic navigation simulation module is used to convert the virtual magnetic stripe data into a simulated magnetic field signal output that conforms to the standard magnetic navigation sensor protocol; The motion control module is used to perform path tracking and dynamic correction control of the mobile device based on the model predictive control algorithm and the simulated magnetic field signal.

Images