Cross-regional precise indoor positioning method

Abstract

The present invention relates to the field of indoor positioning and communication technologies, and particularly to a cross-region precise indoor positioning method. By using 6G micro base stations, adjustable metamaterial reflectors, voxel lidar, and quantum inertial measurement units to work together, precise indoor positioning in multiple regions (such as cells, corridors, activity halls, stairs, etc.) is achieved. The core process includes incorporating multi-source data into a hyper-surface factor fusion model for continuous calculation, and combining with a digital twin map to identify the cross-region behavior of prisoners. Subsequently, the dynamic phased array mechanism of the adjustable metamaterial reflector is used to adaptively change the wireless coverage. The present invention maintains centimeter-level accuracy and millisecond-level response in corner, multipath, and base station handover scenarios, and can avoid positioning jumps or interruptions that occur in traditional technologies during area conversion, making it suitable for precise monitoring in high-security scenarios such as prisons.

Description

Technical Field

[0001] The present invention relates to the field of indoor positioning and communication technologies, and particularly to a method for accurate indoor positioning across regions. Background Art

[0002] Accurate indoor positioning across regions is of great significance in prison management. Especially when prisoners need to be tracked to centimeter-level accuracy without delay, it poses strict requirements for the rapid switching and continuity of the positioning system. The currently common method is to cooperate multiple base stations with prisoner bracelets, and the position is parsed by the base stations sending signals to the bracelets. However, when prisoners continuously move in different regions (such as cells, corridors, activity halls, stairs, etc.), the number of available base stations and signal coverage patterns within each region often vary greatly, resulting in the base station switching speed and positioning mode being unable to adapt to environmental changes in real time. If the switching is not sensitive enough or the positioning mode change is delayed by a few seconds, tracking deviation or even position loss will occur. Existing technologies often use independent base station combinations or locally add blind spot neural networks for compensation. However, due to the failure to effectively solve problems such as high floors, corner occlusion, and real-time data fusion, there are still problems such as instantaneous positioning disorders across regions, multi-path interference that cannot be corrected in time, or the inability to maintain centimeter-level accuracy during instantaneous region switching.

[0003] Some existing solutions only rely on a fixed number of base stations and use separate positioning algorithms within each region. When a prisoner walks out of the original region and enters a new region, it takes a long switching time to enable another group of base stations or a specific "blind spot compensation model", resulting in a positioning gap of several seconds or even longer; other solutions overly rely on a single wireless measurement method, and it is easy to generate error drifts when encountering corridor occlusion or multi-path of metal doors. Due to the inability to dynamically adjust channel coverage and quickly perceive region changes, existing technologies are difficult to ensure continuous and accurate positioning in complex environments such as across floors, corners, and narrow passages in prisons. Summary of the Invention

[0004] In view of the many problems existing in the above-mentioned prior art, the present invention provides a method for accurate indoor positioning across regions. The present invention enables 6G micro base stations, adjustable metamaterial reflectors, voxel lidar, and quantum inertial measurement units to work together, uses a hypersurface factor fusion model to uniformly solve multi-source data, combines a digital twin map to quickly determine the cross-region actions of prisoners, and realizes the reconstruction of wireless channels under the dynamic phased array mechanism of the adjustable metamaterial reflector. In this way, centimeter-level accuracy and millisecond-level response can be maintained all the time when switching between different floors and different functional areas within the prison, avoiding positioning interruption or significant deviation during cross-region instantaneous in traditional solutions.

[0005] A method for accurate indoor positioning across regions includes the following steps:

[0006] Deploy 6G micro base stations, adjustable metamaterial reflectors, and voxel lidars in various areas of the prison, and install a quantum inertial measurement unit in the bracelet; obtain wireless measurement data from the 6G micro base stations, point cloud data from the voxel lidars, and motion data from the bracelets, and form multi-source data for subsequent processing after time synchronization and noise filtering;

[0007] Based on the multi-source data, simultaneously incorporate the bracelet position, the phase distribution of the adjustable metamaterial reflector, and the point cloud calibration information of the voxel lidar into the hypersurface factor fusion model, and use two-way thrust iteration to nonlinearly co-solve each observation to obtain the continuous coordinates of the bracelet at different times;

[0008] Map the continuous coordinates of the bracelet to the digital twin map of the prison, identify cross-region behaviors based on the coordinate changes across region boundaries, and adjust the wireless signal coverage based on the phase distribution of the adjustable metamaterial reflector to keep the continuous coordinate update of the bracelet during cross-region;

[0009] After obtaining the continuous coordinates of the bracelet, further correct the phase distribution of the adjustable metamaterial reflector in combination with the dynamic obstacle information identified by the voxel lidar to expand and maintain the prison positioning system, so as to achieve accurate indoor positioning in each area.

[0010] Preferably, the step of deploying 6G micro base stations, adjustable metamaterial reflectors, and voxel lidars in various areas of the prison includes:

[0011] Extra adjustable metamaterial reflectors are installed in areas with corners or stairs to reduce the multipath effect caused by occlusion;

[0012] The voxel lidar is installed at a higher position in the prison to cover at least one main activity passage, facilitating subsequent acquisition of point cloud data containing human targets.

[0013] Preferably, the motion data of the quantum inertial measurement unit also includes real-time measurements of acceleration, angular velocity, and attitude changes, and is synchronously processed with the wireless measurement data obtained by the 6G micro base stations under the same time reference, so as to ensure the time alignment accuracy of the multi-source data.

[0014] Preferably, the process of incorporating the bracelet position, the phase distribution of the adjustable metamaterial reflector, and the voxel lidar point cloud calibration information into the hypersurface factor fusion model based on the multi-source data includes:

[0015] Input the motion increment provided by the quantum inertial measurement unit, the time difference of arrival and the angle of arrival obtained by the 6G micro base stations into the factor graph;

[0016] Segment and extract the human contour voxels from the point cloud data output by the voxel lidar, and use the central coordinates of the human contour voxels as the observation factors for the position of the bracelet.

[0017] Preferably, when the hyper-surface factor fusion model performs non-linear collaborative solution by bidirectional thrust iteration, it also corrects the propagation characteristics of the wireless signal path in real time according to the phase distribution of the tunable metamaterial reflector, and weights different observation residuals to improve the positioning accuracy of the bracelet in corners or locally occluded areas.

[0018] Preferably, the bidirectional thrust iteration includes:

[0019] Forward solution: Solve the factor graph for the current observation to obtain a preliminary estimate of the bracelet position;

[0020] Backward traction: Trace back the bracelet position at the previous moment within a short time window, and further correct the preliminary estimate of the bracelet position according to the human center coordinates recognized by the voxel lidar in the subsequent frames.

[0021] Preferably, the step of mapping the continuous coordinates of the bracelet to the digital twin map of the prison and identifying cross-region behaviors according to the coordinate changes across the region boundary includes:

[0022] Calibrate the geometric boundaries of cells, corridors, and activity halls in the digital twin map;

[0023] When the continuous coordinates of the bracelet enter another geometric boundary region from one geometric boundary region, it is determined that the prisoner has completed cross-region movement and a region switching instruction is generated.

[0024] Preferably, the process of adjusting the wireless signal coverage based on the phase distribution of the tunable metamaterial reflector includes:

[0025] When a cross-region behavior is detected, send an instruction to the tunable metamaterial reflector to change the phase combination of the phased array units, so as to increase the signal weight of the base station in the newly entered region;

[0026] When the bracelet is close to a corner or an area with many obstacles, further fine-tune the phase distribution of the phased array units to keep the observation error of the bracelet position within the expected range.

[0027] Preferably, the step of keeping the continuous coordinates of the bracelet updated during cross-region also includes:

[0028] Use the quantum inertial measurement unit for short-term autonomous navigation. When the voxel lidar or the tunable metamaterial reflector is in a temporary switching state resulting in insufficient observation data, preferentially use the pose increment of the quantum inertial measurement unit for positioning transition.

[0029] Preferably, the step of further calibrating the phase distribution of the tunable metamaterial reflector by combining the continuously obtained coordinates of the bracelet with the dynamic obstacle information identified by the voxel lidar to expand and maintain the prison positioning system includes:

[0030] When the voxel lidar detects a new obstacle or after the passage is transformed, update the structural information of the corresponding area in the digital twin map;

[0031] According to the updated structural information, reset the phase distribution of the tunable metamaterial reflector so that the 6G micro base station measurement data still has a low multipath error in the new environment, thereby maintaining the effective coverage of the prison positioning system.

[0032] Compared with the prior art, the advantages and beneficial effects of the present invention are as follows:

[0033] By dynamically reshaping the wireless signal path with a tunable metamaterial reflector, low-error coverage in multipath environments such as corners and metal doors is achieved;

[0034] Through the complementary observation formed by the quantum inertial measurement unit and the voxel lidar, the effect of providing continuous centimeter-level positioning can still be achieved when the base station signal is insufficient or during the handover transition;

[0035] Through the hypersurface factor fusion model, the bracelet position, reflector phase distribution, and lidar observation are incorporated together in a single iteration process, overcoming delays or drifts caused by environmental or reflector state changes;

[0036] By determining the real-time scheduling coverage strategy after the prisoner crosses the region through the digital twin map, the rapid identification of cross-region behavior and the seamless switching of the positioning mode are achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a flowchart of the method of the present invention;

[0038] Figure 2 It is a schematic diagram of the principle of the implementation logic of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0039] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present disclosure. In the following detailed description, for the sake of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present disclosure. However, obviously, one or more embodiments can also be implemented without these specific details. In addition, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessarily confusing the concepts of the present disclosure.

[0040] The terms used herein are merely for describing specific embodiments and are not intended to limit the present disclosure. The terms "including", "comprising", etc. used herein indicate the presence of the described features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0041] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0042] As Figure 1 - Figure 2 shown, a cross-regional precise indoor positioning method includes the following steps:

[0043] Deploy 6G micro base stations, adjustable metamaterial reflectors, and voxel lidars in each area of the prison, and let the bracelet be built-in with a quantum inertial measurement unit; obtain wireless measurement data from the 6G micro base stations, obtain point cloud data from the voxel lidars, and obtain motion data from the bracelets, and form multi-source data for subsequent processing after time synchronization and noise filtering;

[0044] Deploy 6G micro base stations, adjustable metamaterial reflectors, and voxel lidars in each area of the prison, and let the bracelet be built-in with a quantum inertial measurement unit; obtain wireless measurement data from the 6G micro base stations, obtain point cloud data from the voxel lidars, and obtain motion data from the bracelets, and form multi-source data for subsequent processing after time synchronization and noise filtering.

[0045] The 6G micro base station refers to a base station unit that uses the sixth-generation communication protocol or its terahertz extended frequency band. Although 6G is still in the research and development stage, "quasi-6G" experimental devices for terahertz and high millimeter wave frequency bands have emerged. "Quasi-6G" devices can be provided by communication equipment manufacturers (such as China Telecom, China Mobile, China Unicom laboratories, etc.); in the academic community, research institutions have conducted prototype verification on communication in the 300GHz-1THz band. These experimental base stations can achieve higher bandwidth and shorter coverage radius, and cooperate with multiple-input multiple-output (MIMO) antenna arrays for ranging (TOF / TDoA) and angle measurement (AOA). 6G micro base stations are generally installed in positions such as prison high walls, ceilings, or corridors, and provide centimeter-level ranging, angle measurement, and communication capabilities for indoor areas through higher bandwidth and denser deployment methods. TOF (Time of Flight) or TDoA (Time Difference of Arrival) measurement, as well as AOA (Angle of Arrival) measurement, can be realized to sense data such as the distance, direction, and signal strength between the bracelet and the base station.

[0046] An adjustable metamaterial reflector refers to arranging a large-scale phased array of microstructures on the surface of a medium or a thin film, and dynamically adjusting the reflection characteristics of electromagnetic waves by controlling the phase distribution of the microstructures. A metamaterial reflector (or an adjustable metasurface) is a type of device that uses artificial microstructures to achieve the reconstruction of the bandwidth, phase, or scattering of electromagnetic waves. There have been relevant phased reflector test platforms in university laboratories and some startups (such as Kymeta, Meta Materials Inc., etc.). In the higher frequency band (millimeter wave / terahertz), an adjustable panel with a phased array of microstructures can be fabricated to change the reflection path and direction according to an electrical control signal. In a prison environment, the reflector can be installed near corners, stairways, or blind spots, and the reflection direction, amplitude, or focal area can be adjusted through an electrical control array, thereby improving the multipath phenomenon or compensating for blind spots in the base station coverage. For example, if the signal attenuation at a certain corridor corner is relatively severe, the adjustable metamaterial reflector can automatically change the combination of phase units to form a directional reflection according to the relative position of the current base station and the bracelet, concentrating the main beam on the area where the bracelet is located and reducing the interference from complex reflection paths caused by walls, metal facilities, etc.

[0047] A voxel-based lidar performs three-dimensional environmental detection of the interior of a prison through multi-beam or phase scanning, and outputs three-dimensional point cloud data or voxel data (dividing a detailed grid in the XYZ coordinates to determine whether there are objects in each grid cell). Most existing three-dimensional lidar products (such as Velodyne, Ouster, RoboSense, Innoviz, etc.) generate three-dimensional point clouds by using multi-beam or rotational scanning; some manufacturers and research institutions are also exploring converting the output data into real-time stereo detection at the voxel (grid) level. Although "voxel-based" lidar is not yet widespread commercially, a similar effect can be approximately achieved based on high-resolution solid-state LiDAR or multi-beam LiDAR. By performing rasterization processing on the original point cloud through a backend software algorithm and dividing voxel units, the spatial positions of people or obstacles can be detected. When a humanoid target is detected, the lidar can extract the central position of the humanoid point cloud, and then use it to match and fuse with the ranging results of 6G micro base stations to ensure that the prisoners can still be located by the lidar when the base station is partially blocked or the signal attenuates.

[0048] The Quantum Inertial Measurement Unit (Quantum IMU) utilizes ultra-cold atom interferometry, atomic spin, or other quantum-level effects to significantly reduce the random drift in the measurement of acceleration and angular velocity, thereby making the cumulative error within a short period (from several seconds to dozens of seconds) much lower than that of traditional MEMS IMUs. Quantum IMUs currently mostly appear in laboratories or at the military level. Some companies (such as iXblue, Muquans, ColdQuanta, etc.) and research institutions have developed prototypes of "atomic interferometers" or "ultra-cold atom accelerometers", which can be used for high-precision requirements in environments without satellite signals, such as underwater navigation and underground positioning. In prison applications, a simplified version of the quantum gyroscope or quantum accelerometer can be used to reduce the overall size and power consumption, achieving lower random drift than traditional MEMS IMUs. When integrated into a bracelet, the Quantum IMU can maintain the prediction and calculation of the target person's position in case of temporary interruption of cross-regional or micro-base station signals, and realign and correct with the base station and radar measurements after the signal is restored.

[0049] The time alignment of the outputs of 6G micro-base stations, voxel lidar, and Quantum IMUs is achieved through a unified clock or communication protocol. Then, the multipath error, optical noise, or measurement noise of the quantum sensor itself is filtered uniformly, so that the multi-source data formed has higher confidence and consistency during subsequent processing.

[0050] In this invention, 6G micro-base stations are deployed in the main areas of prisons, such as cells, corridors, stairwells, and activity halls, to obtain wireless measurement data that can reflect the distance, angle, or channel quality between prisoners and the base stations. Through adjustable metamaterial reflectors, the directional reflection or phase regulation of electromagnetic waves in severely blocked areas (such as near corners or metal doors) is realized, reducing the multipath effect to a relatively low level. Voxel lidar is used to obtain point cloud data containing the moving positions of prisoners in three-dimensional space, and combined with the acceleration, angular velocity, and attitude information of the Quantum Inertial Measurement Unit in the bracelet to make up for the continuous positioning in case of short-term attenuation or blind spots of the base station signal. After time synchronization and noise filtering of all the collected data, it can be used as the input for subsequent fusion algorithms to maintain stable and high-precision indoor positioning in cross-regional scenarios (such as crossing floors or corners) in the prison environment.

[0051] In one embodiment, if a prisoner walks through a corridor and enters a stairwell, temporarily losing sight of some base stations, the displacement increment is estimated through the Quantum Inertial Measurement Unit. Subsequently, when the base station coverage is restored on the other side of the stairway entrance, the base station measurement values and the lidar point cloud data are fused together to quickly correct the positioning coordinates of the bracelet and achieve seamless connection.

[0052] Preferably, the steps of deploying 6G micro-base stations, adjustable metamaterial reflectors, and voxel lidar in each area of the prison include:

[0053] An adjustable metamaterial reflector is additionally provided in areas with corners or stairs to reduce the multipath effect caused by occlusion;

[0054] The voxel lidar is installed at a relatively high position in the prison so that it covers at least one main activity passage, facilitating subsequent acquisition of point cloud data containing humanoid targets.

[0055] The adjustable metamaterial reflector additionally configured at the corner or stair position can perform beam reshaping on 6G-band electromagnetic waves through the phased array microstructure inside it, achieving directional reflection or phase adjustment, thereby significantly reducing the multiple scattering paths generated by channel occlusion;

[0056] In a specific embodiment, if there are metal guardrails or concrete walls at the corner of the prison corridor, an adjustable metamaterial reflector can be mounted on the corner wall, and the phase unit combination can be dynamically changed using a small controller to focus or refract the base station radio frequency signal to the other side passage where the bracelet may appear; during this process, the reflection attenuation coefficient can also be corrected by adjusting the microstructure arrangement to increase the effective signal strength. The voxel lidar is installed at a relatively high position in the prison (such as the ceiling or high wall). By dividing fine grid voxels in the XYZ coordinate system, it outputs point cloud data including the prisoner's body shape, surrounding obstacles, etc. If large-scale human body voxel movement is recognized, the specific location of the prisoner in the corridor, activity hall or stairway can be determined by combining base station ranging, thus avoiding obvious positioning deviation when there is no direct line of sight at the corner or stair corner.

[0057] In one embodiment, if a prisoner moves quickly in a long and narrow corridor, the lidar scans at a frequency of 10 Hz or higher, and can detect the position change of the humanoid contour within a distance of 10 meters. With the help of the adjustable metamaterial reflector, the wireless signal can still maintain a certain quality at the corner, and finally achieve centimeter-level or decimeter-level dynamic positioning accuracy in the data fusion link.

[0058] Preferably, the motion data of the quantum inertial measurement unit also includes real-time measurements of acceleration, angular velocity, and attitude changes, and is synchronously processed with the wireless measurement data obtained by the 6G micro base station under the same time reference, so as to ensure the time alignment accuracy of multi-source data.

[0059] The quantum inertial measurement unit can significantly reduce random drift in a short time through methods such as ultra-cold atom interference or quantum spin sensing, and record the bracelet's triaxial acceleration , , and triaxial angular velocity , , The real-time changes are reflected in the form of Euler angles or quaternions while reflecting the attitude changes. In coordination with the programmable clock or the synchronization frame sent by the base station, the IMU measurement values and the base station measurement data can be uniformly mapped to the same timestamp.

[0060] In one embodiment, the acceleration measured by the IMU can be integrated within discrete time periods and the angular velocity to update the bracelet speed and position . If the base station TDoA measurement obtains the distance between the bracelet and the base station at time , then according to the alignment of the two time bases, and are compared for residuals and corrected in subsequent fusion steps to ensure continuous positioning of the prisoner without jumps when crossing different areas within the prison. Combining the humanoid point cloud position output by the voxel-based lidar can assist in detecting certain short-term loss-of-lock phenomena: if the error between the IMU predicted coordinates and the radar-recognized point cloud is too large, the calibration process can be triggered to keep the final positioning solution within the expected accuracy range at all times. and

[0061] Based on the multi-source data, the bracelet position, the phase distribution of the tunable metamaterial reflector, and the point cloud calibration information of the voxel-based lidar are simultaneously incorporated into the hypersurface factor fusion model. The bidirectional thrust iteration is used to nonlinearly co-solve each observation to obtain the continuous coordinates of the bracelet at different times;

[0062] Under the overall idea of accurate indoor positioning across regions, through the multi-source data jointly generated by 6G micro base stations, tunable metamaterial reflectors, voxel-based lidars, and quantum inertial measurement units, the hypersurface factor fusion model can be used to uniformly solve the bracelet position and related observation information. For the dynamically complex wireless channels and multi-source perception quantities in the prison environment, the bracelet position, the phase configuration of the tunable metamaterial reflector, and the point cloud calibration information of the voxel-based lidar are regarded as multiple mutually coupled high-dimensional factors, and iterations are carried out within the same optimization framework to avoid the limitations of traditional single algorithms that are difficult to take into account the dynamic regulation of the reflector and the calibration of the lidar. Through "bidirectional thrust iteration", both the forward calculation can be used to obtain a preliminary estimate of the real-time bracelet coordinates, and the backward traction can be used to correct the previous moment's estimate within a short time window, so that when quickly switching between different areas of the prison or when the reflector beam changes, the bracelet position can still be kept smooth and accurate.

[0063] ​In one embodiment, when a prisoner moves across floors, the quantum inertial measurement unit accumulates a low drift in a short period of time, and the tunable metamaterial reflector changes phased array units at corners for beam redirection. The lidar provides a humanoid point cloud observation of the corridor or stair space. When these observations are simultaneously incorporated into the hypersurface factor fusion model, a continuous and robust coordinate trajectory can be output through two-way thrust iterative solution, significantly reducing the jump error during the cross-region process.

[0064] In the present invention, the "hypersurface factor fusion model" is used to uniformly integrate multi-source observables and simultaneously include the phase control parameters of the tunable metamaterial reflector, so as to continuously and accurately solve the three-dimensional position of the bracelet at different times in a prison environment. Its principle can be understood as establishing a comprehensive error function in a high-dimensional space, and taking the motion increment of the quantum inertial measurement unit, the time difference of arrival and angle of arrival of the 6G micro base station, the point cloud observation of the voxel lidar, and the reflector phase parameters as "factors" and adding them to the same optimization target, and estimating the bracelet state and reflector configuration simultaneously through non-linear minimization. For ease of explanation, an exemplary calculation embodiment and the corresponding core expression are given below.

[0065] In this model, first define a certain discrete time window , let represent the state of the bracelet at time (including position, optional attitude or velocity), represent the phase distribution vector of the tunable metamaterial reflector at time (for example, the phase combination of phased array units). The point cloud calibration information of the voxel lidar can appear as the "target observation coordinate" , and the motion increment generated by the quantum inertial measurement unit can be regarded as . The time difference of arrival obtained by the 6G micro base station is denoted as , and the angle of arrival is denoted as (if there are multiple base stations, they are respectively denoted as , , etc.). To describe the influence of the reflector on the channel, a "path correction" function can be added to the model to represent how the theoretical prediction of TDoA or AOA changes due to the change of the reflection path or gain by the phased array.

[0066] Incorporate all observables into the same objective function to construct the "comprehensive residual" and minimize it. The example is as follows (omitting the small noise terms in inertial navigation):

[0067]

[0068] Among them, represents the displacement difference of the smart bracelet between adjacent moments, which is the motion increment obtained by integrating the quantum IMU within; and are respectively the "theoretical time difference of arrival" and "theoretical angle of arrival" deduced based on the position of the smart bracelet and the phase distribution of the reflector, which can be calculated through an electromagnetic propagation model or a simplified path approximation. If the reflector changes the phased array unit, these theoretical values will also change along with the update; is the humanoid center coordinate extracted by the voxel lidar, represents a function that matches the smart bracelet coordinate with the human body center (for example, there is a fixed offset when the smart bracelet is worn on the waist or wrist, and this offset relationship can be modeled to infer the voxel center of the entire human body), 、 、 、 are the weighting coefficients of different observation residuals, which are used to balance the influence of each channel in multi-source fusion.

[0069] Regarding the above comprehensive residual as a high-dimensional "hyper-surface", when different values are taken in the space, will also change accordingly; through non-linear iterative methods such as Gauss-Newton or Levenberg-Marquardt, 、 can be solved simultaneously and the smart bracelet state and phase distribution can be continuously updated until convergence, so as to obtain the overall optimal solution. Once the solution converges, the position of the smart bracelet at time can be obtained, and the most suitable reflector phase distribution can be determined, so as to ensure that the time difference of arrival and the angle of arrival are as close as possible in actual measurement and theoretical prediction in various areas of the prison (including corners, stairs, etc.), and are consistent with the humanoid position detected by the voxel lidar.

[0070] In specific applications, the dynamic adaptability of the model can also be enhanced through the two-way thrust iteration of "forward calculation + backward traction": the forward calculation quickly outputs the initial coordinate value based on the current moment's observations, and the backward traction corrects of the previous moment according to the observations of the subsequent several frames (for example, the humanoid center recognized by the lidar in the next beat), so that the entire trajectory remains smooth and accurate when encountering scenarios such as sudden beam switching or rapid personnel cross-regions. For example, when a prisoner enters a new floor and the adjustable metamaterial reflector has not yet switched to the optimal phase, resulting in a temporary increase in the TDoA measurement error, the humanoid position captured by the lidar in the subsequent frames will cause the backward traction to readjust and , while ensuring position accuracy, it also optimizes the reflector configuration to provide continuous robustness for cross - region positioning. Through this joint estimation, the system can simultaneously solve the problems of high - precision positioning and reflector channel maintenance in the complex environment of multiple floors and multiple rooms in prisons, and achieve long - term and stable cross - region tracking of prisoners.

[0071] Preferably, the process of incorporating the bracelet position, the phase distribution of the tunable metamaterial reflector, and the voxel - based lidar point cloud calibration information into the hypersurface factor fusion model based on multi - source data includes:

[0072] Input the motion increment provided by the quantum inertial measurement unit, the time difference of arrival (TDoA) and the angle of arrival (AOA) obtained from the 6G micro - base station into the factor graph together.

[0073] Segment and extract the human body contour voxels from the point cloud data output by the voxel - based lidar, and use the central coordinates of the human body contour voxels as the observation factor for the bracelet position.

[0074] In the specific implementation of the model, the motion increment of the quantum inertial measurement unit (the short - time pose change obtained by integrating acceleration and angular velocity) is combined with the time difference of arrival (TDoA) and the angle of arrival (AOA) of the 6G micro - base station in a factor graph way to form multiple residuals related to the bracelet position; while the three - dimensional point cloud output by the voxel - based lidar is first segmented, the voxel clusters that conform to the human form characteristics are detected, and their central coordinates are extracted to form another dimension of the observation factor, which is connected to the bracelet position. At this time, the phase distribution of the tunable metamaterial reflector can be set as the parameter quantity affecting the channel propagation characteristics between the "base station - bracelet". If the phased array configuration is changed in the corner or stair area, the corresponding constraints in the factor graph will also be refreshed accordingly, ensuring that the system can simultaneously correct the bracelet coordinates and the reflector state to maintain the positioning continuity during cross - region.

[0075] In one embodiment, the displacement increment , of the quantum inertial measurement unit within the time period and the human - shaped center output by the voxel - based lidar at the moment can be jointly used as residual constraints. The TDoA or AOA measured by the base station is restricted by another residual term to the bracelet position, and finally the relatively consistent optimal coordinates are obtained by solving, avoiding positioning errors caused by the failure or drift of a single observation.

[0076] In another embodiment, 6G micro base stations are first deployed in the main areas of the prison (such as cells, corridors, stairwells, and activity halls). These base stations can acquire time difference of arrival and angle of arrival data at high frequencies. At the same time, adjustable metamaterial reflectors are deployed at key positions, and their internal phased array microstructure is used to dynamically regulate the wireless signal path. In addition, voxel-based lidar is installed in the prison to scan and output three-dimensional point cloud data. Through algorithm segmentation, voxel clusters containing human outlines are extracted, and then the center coordinates of these voxel clusters are calculated as target observation values. In addition, the bracelet worn by the prisoner is equipped with a quantum inertial measurement unit, which obtains acceleration and angular velocity data through quantum interference or ultracold atom technology, and obtains short-term pose increments through integration. After all these data are synchronized in time and filtered for noise, they form the input for subsequent multi-source data fusion.

[0077] Specifically, within a discrete time window let represent the three-dimensional position state of the bracelet at time . The motion increment measured by the quantum inertial measurement unit within the time period is denoted as . The time difference of arrival obtained by the 6G micro base station is denoted as and the angle of arrival is denoted as . The humanoid center coordinates output by the voxel-based lidar are denoted as , and the phase distribution of the adjustable metamaterial reflector is denoted as the vector . To jointly solve the above multi-source data in a high-dimensional space, a comprehensive residual function is constructed as follows:

[0078]

[0079] where: represents the motion increment obtained by integrating the quantum inertial measurement unit within the time period ; represents the predicted time difference of arrival based on the bracelet position and the phase distribution of the adjustable metamaterial reflector, and its calculation can be based on the radio propagation model; represents the predicted angle of arrival based on the bracelet position and the reflector phase distribution; represents a mapping function for converting the bracelet position into a theoretical value that matches the humanoid center coordinates extracted by the lidar; , , , are weight coefficients used to balance the influence of each residual in the overall optimization.

[0080] During the implementation process, this comprehensive residual function constitutes a high-dimensional hypersurface, and its optimal solution corresponds to the optimal position of the bracelet at each moment under given observation conditions and the corresponding phase distribution of the metamaterial reflectors. When specifically solving, nonlinear least squares methods such as the Gauss-Newton method or the Levenberg-Marquardt algorithm can be used for iterative optimization. To improve the dynamic adaptability, the system introduces a two-way thrust iteration mechanism:

[0081] Forward solution: At each moment Solve the preliminary position of the bracelet according to all current observation factors (quantum IMU increment, TDoA, AOA, lidar point cloud center) ;

[0082] Backward traction: Use the real-time humanoid center coordinates provided by the lidar in subsequent moments or in more subsequent frames to correct the position at the previous moment and update the residual weights and re-iterate to solve, ensuring that the entire continuous trajectory remains smooth when switching across regions or encountering occlusions.

[0083] For example, in a specific embodiment, when a prisoner enters the stairs from the corridor, the cumulative error of the motion increment provided by the quantum IMU in a short time is small, but due to the possible temporary instability of the base station signal in the stair area, the voxel lidar timely detects the change of the humanoid point cloud in the corridor, resulting in a deviation from the actual measurement value At this time, the backward traction mechanism will adjust the previous bracelet position estimate and the corresponding reflector phase vector , making the newly calculated residual minimum, so as to obtain the corrected optimal position, ensuring continuous and accurate positioning during the cross-region process. Through this joint optimization method, this system can provide stable centimeter-level positioning accuracy in various areas within the prison (including complex environments such as corners, stairs, partitions, etc.), and at the same time, it can update the working state of the adjustable metamaterial reflector in real time, realizing dynamic adaptive control of the wireless channel, and further ensuring the robustness and long-term stability of the entire prison positioning system.

[0084] The advantage of this hypersurface factor fusion model is that it can simultaneously consider the mutual dependencies between multi-source data, avoid the overall positioning deviation caused by errors in a single measurement channel, and enable the system to achieve fast convergence and self-correction in a dynamic environment through the two-way iteration mechanism, providing strong technical support for real-time positioning in a high-security prison environment.

[0085] Preferably, when the hyper - surface factor fusion model uses two - way thrust iteration for non - linear collaborative solution, it also corrects the propagation characteristics of the wireless signal path in real - time according to the phase distribution of the tunable metamaterial reflector, and weights different observation residuals to improve the positioning accuracy of the bracelet in corner or partially blocked areas.

[0086] During the iterative calculation process, the phase distribution of the tunable metamaterial reflector will produce directional or amplitude corrections to the main propagation channel between the base station and the bracelet. For example, after the phased array is updated, a certain reflection path with a low weight may become the main signal path. The two - way thrust iteration mechanism allows temporarily relying on the reflector configuration of the previous iteration during forward derivation. When a large residual is detected (such as a sudden change in TDoA measurement at a corner), backward traction will trace back the bracelet coordinates at the previous moment and re - evaluate the observation residual weights in combination with the new phase configuration. In this way, for strong occlusion scenarios near corners or metal doors, the system can timely compensate for channel attenuation with the adjustment of the tunable metamaterial reflector and reduce the positioning error in the blind area.

[0087] Preferably, the two - way thrust iteration includes:

[0088] Forward solution: Solve the factor graph for the current - moment observables to obtain a preliminary estimate of the bracelet position;

[0089] Backward traction: Trace back the bracelet position at the previous moment within a short - time window, and further correct the preliminary estimate of the bracelet position according to the human body center coordinates identified by the voxel - type lidar in subsequent frames.

[0090] This iterative process can be divided into two steps: forward and backward. First, in the forward - solution stage, the TDoA, AOA, IMU increment, and lidar point - cloud center observations collected at the current moment are input into the factor graph to output a temporary solution of the bracelet coordinates. Subsequently, in the backward - traction stage, the observations within a short historical window are re - evaluated. If there is a large deviation between the human - form center seen by the lidar in the subsequent frame and the bracelet coordinates at the previous moment, the solution at the previous moment is updated by means of backward correction, so as to ensure that the overall trajectory remains smooth under complex spatio - temporal conditions (such as a prisoner quickly passing through a corner or a phase switch of the reflector midway).

[0091] In a specific embodiment, the size of the sliding window can be set in the range of 0.5 to 1 second. When the lidar in the subsequent frame successfully identifies the position of the prisoner's body , and the deviation from the previous bracelet - position prediction is too large, then it is incorporated into the backward - traction process, and combined with the tunable metamaterial reflector at At the moment of phase configuration, reallocate the residual weights of this segment to make the finally output trajectory closer to the actual movement path; in this way, the high-confidence estimation of the prisoner's position can still be maintained at the moment of crossing regions or going up and down stairs in the prison, improving the reliability of cross-region precise indoor positioning.

[0092] Map the continuous coordinates of the bracelet to the digital twin map of the prison, identify cross-region behaviors based on the coordinate changes across the region boundary, and adjust the wireless signal coverage based on the phase distribution of the tunable metasurface reflector, so that the bracelet can maintain continuous coordinate updates during cross-region movement.

[0093] In the application of cross-region precise indoor positioning, in order for the prison management system to real-time master the accurate position of the prisoner from one area (such as a cell) to another area (such as a corridor or an activity hall), it is necessary to map the continuous coordinates of the bracelet to the digital twin map, identify cross-region behaviors by judging when the coordinates cross the geometric boundaries pre-calibrated in the map, and trigger corresponding linkage measures on the monitoring platform. In order to cope with the changes in the wireless signal requirements of different regions, the tunable metasurface reflector will schedule the coverage priority according to its phase distribution, so that the base station signals in the newly entered area can be enhanced or redirected, ensuring that the bracelet still has reliable measurement accuracy at the critical moment of cross-region movement. In addition, if there is a short-term signal mismatch at the switching moment or in special occlusion areas, the quantum inertial measurement unit can be used to maintain short-term autonomous navigation, so that the coordinate updates do not interrupt or have significant jumps.

[0094] In one embodiment, when the prisoner leaves the cell and enters the corridor, after the digital twin map monitors that its coordinates have crossed the cell polygon boundary and entered the corridor boundary range, it immediately notifies the tunable metasurface reflector to adjust the main beam azimuth, increase the amplitude of the base station signal at the other end of the corridor, and the bracelet continuously outputs continuous coordinates during this connection process, ensuring that the entire cross-region positioning process is smooth and reliable.

[0095] In another embodiment, in the forward solution stage, the system will use all the observation data at the current moment According to the formula: Input into the factor graph, and solve the position of the bracelet and the phase parameters of the reflector at the current moment through an iterative optimization algorithm (such as the Gauss-Newton method or the Levenberg–Marquardt algorithm) And the phase parameters of the reflector . Specifically, the system first calculates the difference between the motion increment measured by the quantum IMU and the change in the bracelet position as a motion constraint. At the same time, using the TDoA and AOA data provided by the 6G micro base station, the theoretical prediction value is calculated through the propagation model and subtracted from the actual measurement value to form the corresponding residual. After adding the humanoid center coordinates extracted by lidar as a position compensation factor, all constraints are combined to form a high-dimensional non-linear optimization problem. The optimization goal is to minimize the comprehensive residual :

[0096]

[0097] In each iteration step, the system calculates the gradient and an approximation of the Hessian matrix, updates the solution until the residual converges. Through forward calculation, the system obtains a preliminary continuous coordinate trajectory of the bracelet and records the reflector phase parameters corresponding to each moment to provide a reference for subsequent steps.

[0098] In the backward traction stage, the system uses the subsequent observation data collected after forward calculation, especially the humanoid center coordinates identified by the voxel lidar in subsequent frames, to retrospectively correct the bracelet position at the previous moment. Suppose at time the humanoid center coordinates extracted by the lidar are , if there is a large deviation from the predicted position obtained by forward calculation, then this deviation is fed back to the estimation at time . The backward traction process can fine-tune the previous moment position by reconstructing the residual function within the optimization window. For example, add the following additional residual term:

[0099]

[0100] where is the weight of the backward traction. The system combines this term with the original residual function and iteratively solves again to obtain the corrected bracelet position . This iterative method combining forward and backward can effectively compensate for the instantaneous errors caused by wireless signal or reflector state adjustment during large environmental changes or cross-region processes, ensuring the smooth transition of continuous coordinates.

[0101] Example illustration: For example, at the junction of a prison corridor and a staircase, due to changes in the building structure and the appearance of obstacles (such as metal doors), the TDoA data obtained by the 6G micro base station suddenly shows a large deviation at time , and there is an obvious error between the bracelet position given by forward calculation and the humanoid center identified by the lidar. At this time, the system activates the backward traction mechanism, uses the lidar data at time as a reference to correct the estimation at time . Suppose at time , the humanoid center coordinates detected by the lidar are , and there is a large residual between the corresponding theoretical mapping value and , then through the additional residual Feed this deviation back to the optimization process. After several backward iterations, the updated position of the bracelet matches the lidar observations better, and at the same time corrects the phase parameters of the tunable metamaterial reflector to adapt to the new environment. This process ensures that even when sudden structural changes occur inside the building, the system can still adaptively adjust in a short time and provide continuous and accurate positioning data.

[0102] Combining forward calculation and backward traction, the two-way thrust iteration method realizes the efficient fusion of multi-source data, ensuring high robustness and accuracy in the continuous position calculation of the bracelet in the complex and changeable environment of the prison. The finally output continuous coordinates not only reflect the real-time position of the prisoner, but also synchronously output the optimal phase parameters of the reflector, enabling the entire wireless channel to adaptively adjust according to environmental changes and achieving continuous positioning across regions. Through the above method, this system can eliminate positioning errors caused by signal occlusion, multipath interference, and hardware switching delay in cross-region scenarios, providing strong technical support for the high-security management of prisons.

[0103] Preferably, the step of mapping the continuous coordinates of the bracelet to the digital twin map of the prison and identifying cross-region behavior based on the coordinate changes across the regional boundary includes:

[0104] Calibrate the geometric boundaries of cells, corridors, and activity halls on the digital twin map;

[0105] When the continuous coordinates of the bracelet enter another geometric boundary region from one geometric boundary region, it is determined that the prisoner has completed cross-region movement and a region switching instruction is generated.

[0106] In the digital twin map, the plane and three-dimensional structure information of the prison can be pre-loaded or manually drawn, and corresponding geometric boundaries, such as polygons, polyhedrons, or grid regions, can be set for each main functional area (including cell units, stairs, public corridors, activity halls, etc.) for coordinate collision detection in the subsequent positioning system. When the continuous coordinates of the bracelet cross from the inner region of one boundary to another boundary region, it can be determined that the prisoner has completed the region switching; at this time, the system will generate a "region switching instruction" and can synchronize this information to the tunable metamaterial reflector control module or the monitoring platform. If there is also a security access control linkage between multiple regions, subsequent operations such as access control and video monitoring will be triggered according to this instruction to strengthen the security management and traceability of the cross-region process.

[0107] In one embodiment, the activity hall is set as the coordinate boundary , and the corridor is set as , when instantly satisfies after a certain moment When this occurs, the system determines that the cross-region movement is completed and issues an alarm or status update, presenting the area change information of the prisoner to the prison guards in real time.

[0108] Preferably, the process of adjusting the wireless signal coverage based on the phase distribution of the adjustable metamaterial reflector includes:

[0109] When a cross-region behavior is detected, send an instruction to the adjustable metamaterial reflector to change the phase combination of the phased array units, so as to increase the signal weight of the base station in the newly entered area;

[0110] When the bracelet is close to an area with many corners or obstacles, further fine-tune the phase distribution of the phased array units to keep the position observation error of the bracelet within the expected range.

[0111] After determining the cross-region switch, the adjustable metamaterial reflector will automatically update the phase combination of the internal phased array units according to the priority of the area or real-time requirements, so as to reconstruct the wireless channel coverage. If a prisoner enters a relatively spacious activity hall from a stairwell, the system can increase the phase gain of the base station signal corresponding to the hall or adjust the reflector to an angle more conducive to the hall coverage, so as to ensure that the bracelet can obtain stronger ranging or angle measurement information in the hall. When a prisoner enters a corner or an area with dense obstacles (such as fixed metal partitions), the "phased array fine-tuning" method can also be used to bypass the obstacles in the signal path, improve the controllability of the observation residual, and keep the overall positioning with better accuracy in this local area.

[0112] In one embodiment, if there is a metal door at the corner at the end of the corridor, when it is detected that the prisoner is about to pass through the door and enter another area, the system will automatically send an instruction to the phased array of the reflector to change the enhancement strategy of the base station signal in the area behind the door and complete the switch within a few milliseconds to ensure continuous positioning during the process of the prisoner crossing the threshold.

[0113] Preferably, the method of keeping the bracelet continuously updated with coordinates during cross-region movement further includes:

[0114] Utilize the quantum inertial measurement unit for short-term autonomous navigation. When the voxel lidar or the adjustable metamaterial reflector is in a temporary switching state resulting in insufficient observation data, preferentially use the pose increment of the quantum inertial measurement unit for positioning transition.

[0115] To prevent the observation data from being temporarily missing due to the adjustable metamaterial reflector not completing the beam redirection instantaneously during cross-region movement, or the line of sight of the voxel lidar being temporarily blocked, the system can switch to the pose increment provided by the quantum inertial measurement unit for short-term autonomous navigation within an extremely short time. Specifically, the quantum IMU continuously calculates the bracelet coordinates through the integration of acceleration and angular velocity , it can maintain low drift accuracy within seconds even in the absence of external observations. After the voxel lidar and the reflector stabilize their outputs again, the subsequent fusion algorithm will perform differential correction on the coordinates estimated by the IMU and the new external observations to achieve smooth connection.

[0116] In one embodiment, when the prisoner is walking in the stair area across floors, if the voxel lidar loses some of the point cloud due to the floor structure and the reflector has not yet switched to the optimal beam pattern for this floor, the quantum IMU can maintain continuous coordinate updates; when the prisoner reaches the stair exit and the radar and base station signals are restored, the fusion algorithm will match the IMU trajectory with the new TDoA / lidar point cloud information to generate a consistent and accurate coordinate trajectory without the prisoner having to stop or perform any calibration operations.

[0117] After obtaining the continuous coordinates of the bracelet, combined with the dynamic obstacle information identified by the voxel lidar, further correct the phase distribution of the adjustable metamaterial reflector to expand and maintain the prison positioning system, so as to achieve accurate indoor positioning in each area.

[0118] In the overall workflow of precise indoor positioning across regions, when the continuous coordinates of the bracelet have been obtained through multi-source fusion of 6G micro base stations, adjustable metamaterial reflectors, voxel lidars, and quantum inertial measurement units, if new obstacles (such as temporarily stacked metal utensils) or changes in the venue layout (such as renovating passages, adding or removing partition walls) occur inside the prison, it is necessary to expand and maintain the positioning system itself. By using the voxel lidar to identify the dynamic changes in the prison environment in real time, the corresponding structural information can be updated in the digital twin map. At the same time, combined with the adjustable metamaterial reflector to perform secondary correction on the beam direction and phase distribution, so that even when new obstacles or new passages appear, the wireless signals emitted by the 6G micro base station can still cover the target area with low multipath error, thereby maintaining accurate monitoring of the prisoner's movement across regions.

[0119] In one embodiment, if a temporary iron fence is added in the activity hall, the voxel lidar will detect the appearance of the fence and output the modified point cloud distribution of the humanoid target. The digital twin map will simultaneously refresh the entity boundary information of the hall, and then instruct the adjustable metamaterial reflector to adjust the phased array unit to bypass the signal reflection path brought by the new fence, and finally continue to ensure high-precision positioning coverage in this area.

[0120] Preferably, the step of further correcting the phase distribution of the adjustable metamaterial reflector after obtaining the continuous coordinates of the bracelet and combining the dynamic obstacle information identified by the voxel lidar to expand and maintain the prison positioning system includes:

[0121] When the voxel-based lidar detects a new obstacle or after the passage is renovated, update the structural information of the corresponding area in the digital twin map;

[0122] According to the updated structural information, reset the phase distribution of the tunable metamaterial reflector so that the measurement data of the 6G micro base station still has a low multipath error in the new environment, thereby maintaining effective coverage of the prison positioning system.

[0123] In specific implementation, the voxel-based lidar can obtain the three-dimensional point cloud data inside the prison through regular scanning or trigger scanning, and compare the changed area with the existing map. If the shape of the newly recognized large object (such as a cuboid iron box or a newly added partition wall) does not match the original information in the database, it will mark and update the digital twin map. Subsequently, the tunable metamaterial reflector will reassign the phase combination of the phased array units by analyzing the latest map structure. For example, when the position of the newly added obstacle blocks the original main propagation path, the system can turn the beam to the auxiliary path or locally increase the attenuation compensation on the reflector surface to minimize the multipath stray interference caused by the newly added obstacle.

[0124] In one embodiment, if a baffle is temporarily added at the end of the stair passage, after the voxel-based lidar detects this hard baffle and updates the map information, the tunable metamaterial reflector immediately makes partial parameter adjustments in the phased array microstructure, enabling the signal to bypass the baffle from another side door or a better angle, ensuring the reliability of the distance measurement of the 6G micro base station and avoiding large positioning deviations caused by environmental changes. Thus, the present invention continuously expands and maintains the prison positioning system, enabling it to still provide reliable high-precision positioning services for prison guards in situations where personnel frequently move across regions or facilities are constantly changing.

[0125] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects.

[0126] The above are only the embodiments of the present application and are not used to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included within the scope of the claims of the present application.

Claims

1. A cross-region precise indoor positioning method, characterized in that The steps include: Deploy 6G micro base stations, tunable metamaterial reflectors, and voxel lidars in each area of the prison, and install a quantum inertial measurement unit in the bracelet; obtain the time difference of arrival, angle of arrival, and received signal strength indication value between the 6G micro base station and the bracelet, obtain the three-dimensional point cloud data containing the human voxel cluster of the target person and the prison environment structure information from the voxel lidar, and extract the center coordinates of the human voxel in real time, and obtain the three-axis acceleration, three-axis angular velocity, and attitude change from the bracelet and integrate them to obtain the motion increment. After time synchronization and noise filtering, multi-source data for subsequent processing is formed; Based on the multi-source data, simultaneously introduce in the hypersurface factor fusion model: the three-dimensional position coordinates of the bracelet, the phase distribution parameters of the tunable metamaterial reflector, and the center coordinates of the target human voxel extracted by the voxel lidar; use the inertial motion increment residual, time difference of arrival residual, angle of arrival residual, and point cloud matching residual as observation factors, assign weights to each observation residual respectively, and then uniformly form an objective function, and use the non-linear minimization method for joint solution; first perform forward solution in the current observation window to obtain the initial value of the bracelet position, and then use the new observations in the next frame to perform backward smoothing correction on the previous frame to form a forward-backward two-way thrust iteration, and output the continuous coordinates of the bracelet at each moment; Map the continuous coordinates of the bracelet to the digital twin map of the prison, identify cross-regional behaviors according to the coordinate changes across regional boundaries, and adjust the wireless signal coverage based on the phase distribution of the tunable metamaterial reflector, so that after the bracelet detects a cross-regional behavior during cross-regional: first send an instruction to the tunable metamaterial reflector to adjust the phase of the phased array to enhance the 6G micro base station signal in the new area; during the phase adjustment period or when the point cloud is temporarily missing, temporarily maintain short-term navigation with the pose increment of the quantum inertial measurement unit; After all observables are restored, input the above transitional solution together with the updated wireless measurement data and point cloud data into the hypersurface factor fusion model to continue the solution, so as to seamlessly output continuous coordinates; After obtaining the continuous coordinates of the bracelet, combine the dynamic obstacle information identified by the voxel lidar to further correct the phase distribution of the tunable metamaterial reflector to expand and maintain the prison positioning system, so as to achieve accurate indoor positioning in each area.

2. The cross-region precise indoor positioning method according to claim 1, wherein The step of deploying 6G micro base stations, tunable metamaterial reflectors, and voxel lidars in each area of the prison includes: Extra tunable metamaterial reflectors are installed in areas with corners or stairs to reduce the multipath effect caused by occlusion; The voxel lidar is installed at a higher position in the prison so that it covers at least one main activity passage to facilitate subsequent acquisition of point cloud data containing human-shaped targets.

3. A cross-region precise indoor positioning method according to claim 1 or 2, characterized in that The motion data of the quantum inertial measurement unit also includes real-time measurements of acceleration, angular velocity, and attitude changes, and is synchronously processed with the wireless measurement data obtained by the 6G micro base station under the same time reference, so as to ensure the time alignment accuracy of multi-source data.

4. The cross-region precise indoor positioning method according to claim 1, characterized in that The process of incorporating the bracelet position, the phase distribution of the tunable metamaterial reflector, and the calibration information of the voxel lidar point cloud in the hypersurface factor fusion model based on multi-source data includes: Input the motion increment provided by the quantum inertial measurement unit, the time difference of arrival and the angle of arrival obtained by the 6G micro base station into the factor graph together. Segment and extract the human contour voxels from the point cloud data output by the voxel lidar, and use the central coordinates of the human contour body as the observation factor for the bracelet position.

5. A cross-region precise indoor positioning method according to claim 4, wherein When the hypersurface factor fusion model uses two-way thrust iteration for non-linear collaborative solution, it also corrects the propagation characteristics of the wireless signal path in real time according to the phase distribution of the tunable metamaterial reflector, and weights different observation residuals to improve the positioning accuracy of the bracelet in corners or locally occluded areas.

6. The cross-region precise indoor positioning method according to claim 5, wherein The two-way thrust iteration includes: Forward solution: Solve the factor graph for the current moment observations to obtain a preliminary estimate of the bracelet position. Backward traction: Trace back the bracelet position at the previous moment within a short time window, and further correct the preliminary estimate of the bracelet position according to the human center coordinates identified by the voxel lidar in the subsequent frames.

7. A cross-region precise indoor positioning method according to claim 1, characterized in that, The steps of mapping the continuous coordinates of the bracelet to the digital twin map of the prison and identifying cross-region behaviors based on the coordinate changes across the region boundary include: Calibrate the geometric boundaries of the cells, corridors, and activity halls in the digital twin map. When the continuous coordinates of the bracelet enter another geometric boundary region from one geometric boundary region, determine that the prisoner has completed cross-region movement and generate a region switching instruction.

8. An indoor precise cross-region positioning method according to claim 7, characterized in that, The process of adjusting the wireless signal coverage based on the phase distribution of the tunable metamaterial reflector includes: When a cross-region behavior is detected, send an instruction to the tunable metamaterial reflector to change the phase combination of the phased array units, so as to increase the signal weight of the base station in the newly entered region. When the bracelet is close to a corner or an area with many obstacles, further fine-tune the phase distribution of the phased array units to keep the observation error of the bracelet position within the expected range.

9. A cross-region precise indoor positioning method according to claim 1, characterized in that The process of keeping the continuous coordinates of the bracelet updated during cross-region also includes: Use the quantum inertial measurement unit for short-term autonomous navigation. When the voxel lidar or the tunable metamaterial reflector is in a temporary switching state resulting in insufficient observation data, preferentially use the pose increment of the quantum inertial measurement unit for positioning transition.

10. The cross-region precise indoor positioning method according to claim 1, wherein The steps of further calibrating the phase distribution of the tunable metamaterial reflector by combining the dynamic obstacle information identified by the voxel lidar after obtaining the continuous coordinates of the bracelet to expand and maintain the prison positioning system include: When the voxel lidar detects new obstacles or after the passage is renovated, update the structural information of the corresponding region in the digital twin map. Reset the phase distribution of the tunable metamaterial reflector according to the updated structural information, so that the 6G micro base station measurement data still has a low multipath error in the new environment, thereby maintaining the effective coverage of the prison positioning system.

Images