Indoor positioning track tracing and regional early warning method and system

Abstract

The invention discloses an indoor positioning track tracing and area early warning method and system, and relates to the technical field of indoor positioning, the method comprises the following steps: receiving a positioning signal, and synchronously collecting inertial navigation data of an inertial measurement unit; filtering processing is carried out on the positioning signal, non-line-of-sight interference and multipath effect interference components are identified, and interference compensation parameters are generated; after a positioning signal is calibrated by combining the parameters, real-time position coordinates are calculated; running a pedestrian dead reckoning algorithm on the basis of inertial navigation data to obtain continuous relative position coordinates, and correcting by taking the real-time position coordinates as reference to form initial movement track data; presetting an early warning area boundary and a triggering condition, and if the condition is met, starting an early warning response and recording an early warning event; and integrating the continuous real-time position coordinates, the initial movement track data and the early warning event records, and constructing a complete movement track file to realize track tracing. The system is matched with the method, the positioning reliability in the indoor complex environment can be improved, and continuous track tracing and accurate area early warning are supported.

Description

Technical Field

[0001] This invention relates to the field of indoor positioning technology, and in particular to an indoor positioning trajectory tracing and area early warning method and system. Background Technology

[0002] In indoor scenarios such as fire emergency response and industrial workshop management, there is an urgent need for real-time positioning of personnel or mobile devices, movement trajectory tracking, and early warning of specific areas. Existing indoor positioning technologies mostly use single-mode positioning signals such as Bluetooth and Ultra-Wideband (UWB), which calculate the target location by receiving positioning signals and combining them with positioning algorithms, and then attempt to achieve trajectory recording and area early warning functions.

[0003] However, indoor environments are complex, with walls obstructing the view and dense equipment. Positioning signals are susceptible to non-line-of-sight interference and multipath effects during propagation, leading to signal distortion. Existing technologies struggle to effectively identify and compensate for such interference, resulting in insufficient positioning accuracy. This inaccuracy directly leads to discontinuous and inaccurate recording of subsequent motion trajectories, making it impossible to create a complete and traceable trajectory archive. It also causes deviations in the triggering of regional warnings, making it difficult to achieve accurate and reliable regional warnings. Summary of the Invention

[0004] To address the technical problems existing in the prior art, this invention provides an indoor positioning trajectory tracing and regional early warning method and system.

[0005] The technical solution adopted in this invention is: The first aspect of this application provides an indoor positioning trajectory tracing and area early warning method, including the following steps: Step 1: Receive positioning signals using spread spectrum modulation technology transmitted by indoor multi-mode positioning base stations, and simultaneously collect inertial navigation data from its own inertial measurement unit; Step 2: Filter the positioning signal, identify non-line-of-sight interference components and multipath interference components, and generate interference compensation parameters; Step 3: Calibrate the filtered positioning signal by combining interference compensation parameters, and calculate its real-time position coordinates using a triangulation model; Step 4: Run the pedestrian dead reckoning algorithm based on inertial navigation data to obtain continuous relative position coordinates. Use the real-time position coordinates as a reference to correct the relative position coordinates and form preliminary motion trajectory data. Step 5: Preset the boundary range and warning triggering conditions of the indoor warning area, compare the real-time location coordinates with the boundary range in real time to determine whether the warning triggering conditions are met; if the warning triggering conditions are met, activate the warning response mechanism, and simultaneously record the warning triggering time, warning type and current real-time location coordinates to form a warning event record; Step 6: Integrate continuous real-time location coordinates and early warning event records to build a complete motion trajectory archive and achieve trajectory tracing.

[0006] The second aspect of this application provides an indoor positioning trajectory tracing and area early warning system, which applies the above-mentioned indoor positioning trajectory tracing and area early warning method, including: The signal receiving and inertial navigation data acquisition module is used to receive positioning signals using spread spectrum modulation technology transmitted by multi-mode positioning base stations deployed indoors, and simultaneously acquire inertial navigation data from its own inertial measurement unit. The signal filtering and interference compensation parameter generation module is used to filter the positioning signal, identify non-line-of-sight interference components and multipath effect interference components, and generate interference compensation parameters. The positioning signal calibration and real-time coordinate calculation module is used to calibrate the filtered positioning signal in combination with interference compensation parameters and calculate its own real-time position coordinates through a triangulation model. The pedestrian dead reckoning and preliminary trajectory generation module is used to run a pedestrian dead reckoning algorithm based on inertial navigation data, obtain continuous relative position coordinates, use real-time position coordinates as a reference to correct the relative position coordinates, and form preliminary motion trajectory data. The warning condition setting and warning response module is used to preset the boundary range and warning triggering conditions of the indoor warning area, compare the real-time location coordinates with the boundary range in real time, and determine whether the warning triggering conditions are met. If the warning triggering conditions are met, the warning response mechanism is activated, and the warning triggering time, warning type and current real-time location coordinates are recorded simultaneously to form a warning event record. The trajectory data integration and archive construction module integrates continuous real-time location coordinates, preliminary motion trajectory data, and early warning event records to construct a complete motion trajectory archive, enabling trajectory tracing. The beneficial effects of the present invention are at least one of the following: By filtering the received positioning signal, identifying non-line-of-sight interference components and multipath interference components, and generating interference compensation parameters, the positioning signal is calibrated using these parameters, and then the real-time position coordinates are calculated using a triangulation model. Simultaneously, inertial navigation data is collected to assist in correction, which helps improve the accuracy of positioning in complex indoor environments and alleviates the problem of insufficient accuracy caused by interference with the positioning signal.

[0007] Based on the real-time location coordinates obtained after calibration, the continuous relative location coordinates are obtained and corrected by the pedestrian dead reckoning algorithm to form preliminary motion trajectory data. Subsequently, this data is integrated with the early warning event records to construct a complete motion trajectory archive, which can support continuous and reliable motion trajectory tracing and meet the requirement of trajectory traceability.

[0008] By leveraging the improved real-time location coordinate accuracy and comparing it with the preset warning area boundary range to determine whether a warning has been triggered, the reliability of regional warning judgment is improved, and the situation of misjudgment or missed judgment caused by positioning deviation is reduced. Attached Figure Description

[0009] Figure 1 This is a schematic flowchart of the method of the present invention; Figure 2 This is a structural block diagram of the system of the present invention. Detailed Implementation

[0010] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0011] Considering that indoor positioning signals are susceptible to non-line-of-sight interference and multipath effects, resulting in insufficient positioning accuracy, it is impossible to achieve continuous motion trajectory tracking and accurate area early warning.

[0012] To address the aforementioned technical problems, Embodiment 1 provides an indoor positioning trajectory tracing and area early warning method, such as... Figure 1 As shown, it includes the following steps: Step 1: Receive positioning signals using spread spectrum modulation technology transmitted by indoor multi-mode positioning base stations, and simultaneously collect inertial navigation data from its own inertial measurement unit.

[0013] It should be noted that multi-mode positioning base stations refer to signal transmission devices that integrate multiple positioning modes such as Bluetooth, ultra-wideband (UWB), and Wi-Fi. They can adaptively switch positioning modes according to the complexity of the indoor environment, improving signal coverage stability. Spread spectrum modulation technology refers to the technology of spreading the spectrum of positioning signals to a wider frequency band for transmission, which can enhance the signal's anti-interference capability. For example, linear frequency modulation spread spectrum (Chirp Spread Spectrum, CSS) achieves spread spectrum by transmitting chirp signals whose frequency changes linearly with time, and has the advantages of anti-multipath and low power consumption. An inertial measurement unit (IMU) is a sensor module that integrates gyroscopes and accelerometers to collect the motion state parameters of the device. Inertial navigation data refers to the data set of angular velocity, acceleration, and other data reflecting the motion characteristics of the device output by the inertial measurement unit.

[0014] In the specific implementation process, the positioning terminal, such as the positioning wristband worn by industrial personnel, activates the multi-mode signal receiving module to scan the indoor preset multi-mode positioning base station network, obtain the equipment identification, signal type and transmission parameters of each base station, and prioritizes the identification of base stations supporting CSS technology; through initial screening of signal quality, a set of candidate base stations is determined; a signal request is sent to the candidate base stations, and the positioning signals using CSS technology are received from the base stations, extracting key information such as the start frequency, end frequency, bandwidth, and propagation time of the chirp signal; simultaneously, the inertial measurement unit built into the positioning terminal is activated, the sampling frequency is set to 100Hz, and the angular velocity around the x, y, and z axes and the acceleration data along the three axes are collected, and the collected data are stored after preliminary moving average filtering and noise reduction.

[0015] Considering the quality differences of signals from different base stations in a multi-base station network, and the fact that some base station signals are more severely interfered with, direct reception will introduce low-quality data that affects subsequent positioning accuracy. In one possible implementation, step 1 includes the following sub-steps: Sub-step 1.1: Receive signals from a multi-base station network consisting of multiple multi-mode positioning base stations.

[0016] It should be noted that multi-base station network signals refer to the set of positioning signals emitted by a network formed by multiple multi-mode positioning base stations deployed at preset intervals, which achieve full indoor coverage through the collaboration of multiple base stations.

[0017] After the positioning terminal is started, it scans the positioning base station signals in the surrounding environment through the built-in multi-mode signal receiving module, obtains basic information such as the equipment identification, signal transmission power, and signal propagation time of each base station, and classifies all the scanned signals into multi-base station network signals and stores them temporarily.

[0018] Sub-step 1.2: Calculate the link quality factor and the geometric precision factor. The link quality factor is calculated based on the received signal power, the ratio of direct component to total energy, and the peak delay difference. The geometric precision factor is calculated based on the spatial distribution of the base station.

[0019] It should be noted that the link quality factor is a parameter used to evaluate the reliability of the positioning signal transmission link, comprehensively reflecting factors such as attenuation and interference during signal transmission; a higher value indicates better link quality. The geometric accuracy factor is a parameter used to evaluate the impact of the spatial distribution of multiple base stations on positioning accuracy, reflecting the rationality of the base station layout; a lower value indicates higher positioning accuracy potential. Received signal power is the actual power of the positioning signal received by the positioning terminal and is a core indicator for measuring signal strength. The ratio of direct component to total energy is the ratio of the energy of the direct path component in the positioning signal to the total signal energy; a higher ratio indicates less influence from multipath effects. Peak delay difference is the difference in the peak arrival times of each path component in the positioning signal; a smaller difference indicates a weaker multipath effect.

[0020] In practice, the received power of each base station signal is directly measured using the power detection module built into the positioning terminal. Spectral analysis is performed on the received signal to separate the direct component energy E. d With total energy Calculate the ratio of direct energy to total energy. By analyzing the signal in the time domain, the peak arrival time of each path component is identified, and the difference between the maximum and minimum values ​​is taken as the peak delay difference. .

[0021] The link quality factor is calculated using a weighted summation formula. The formula is ,in , , These are the weighting coefficients. Example values , , , For received signal power The normalized value.

[0022] Calculate geometric precision factor : Obtain the preset coordinates of each base station Based on the current coarse position of the positioning terminal, which is estimated through the initial scan signal, the geometric precision factor is calculated according to the formula. Calculation, where , , The variances of the positioning errors in the x, y, and z directions are respectively derived from the geometric relationship between the base station coordinates and the approximate location.

[0023] Sub-step 1.3: Construct an objective function based on the link quality factor and geometric precision factor, and use the objective function to select the optimal combination of base stations.

[0024] In the specific implementation process, the objective function aims to maximize the comprehensive evaluation value, and the formula for the objective function is: in , For weighting coefficients, an example value is provided. , , This represents the average link quality factor of each base station in the combination. This is the average value of the geometric precision factor of each base station in the combination.

[0025] Select all possible combinations of three or more base stations from the multi-base station network signal, calculate the objective function value F for each combination, and select the combination with the largest F value as the optimal base station combination.

[0026] Sub-step 1.4: Receive the positioning signal transmitted by the optimal base station combination using spread spectrum modulation technology.

[0027] In the specific implementation process, the positioning terminal sends a signal reception request to each base station in the optimal base station combination. After the base station responds to the request, it sends the positioning signal according to the spread spectrum modulation technology. The positioning terminal receives the signal through the multi-mode signal receiving module, performs preliminary format parsing and synchronization processing on the signal, and extracts key parameters such as the carrier frequency, symbol rate, and spreading code of the signal and stores them temporarily.

[0028] Sub-step 1.5: Collect the angular velocity and acceleration data of its own inertial measurement unit as inertial navigation data.

[0029] For example, the positioning terminal activates its built-in inertial measurement unit, sets the sampling frequency to 100Hz, and collects the angular velocities of the terminal around the x, y, and z axes in real time. , , and acceleration along the x, y, and z axes , , The collected data is filtered and denoised (using moving average filtering), and the processed data is classified as inertial navigation data and stored in timestamp order.

[0030] Step 2: Filter the positioning signal, identify non-line-of-sight interference components and multipath interference components, and generate interference compensation parameters.

[0031] Considering that even after selecting the optimal base station signal, the signal may still exhibit random fluctuations, and that combined interference from non-line-of-sight and multipath effects is difficult to accurately identify through general filtering, while fixed thresholds cannot adapt to dynamic changes in the indoor environment, resulting in interference compensation parameters lacking specificity. To address these technical problems, in one possible implementation, step 2 includes the following sub-steps: Sub-step 2.1: Use a low-pass filtering algorithm to preprocess the positioning signal received in step 1.

[0032] In the specific implementation process, the cutoff frequency of the low-pass filtering algorithm is determined. Based on the symbol rate of the positioning signal, the cutoff frequency is set to 1.5 times the symbol rate. A first-order RC low-pass filtering algorithm is used to process the positioning signal received in step 1. The filtering formula is as follows: ,in This represents the original signal value at the current moment. This is the filtered signal value at the current moment. The filtered signal value from the previous moment. α is the filter coefficient, with a value range of 0 < α < 1. An example value is α = 0.3.

[0033] Sub-step 2.2: Analyze the fluctuation pattern of the preprocessed positioning signal using a sliding time window.

[0034] It should be noted that the fluctuation pattern refers to the characteristics of the pre-processed positioning signal, such as the amplitude variation range, variation frequency, and peak distribution, which reflect the stability of the signal.

[0035] In the specific implementation process, the length of the sliding time window is set to 100ms, the window sliding step is 10ms, the preprocessed positioning signal sequence is divided into windows, and the maximum value of the signal within each window is counted. Minimum value ,average value and standard deviation The standard deviation The calculation formula is: ; in, For the first in the window The signal value at each sampling point The number of sampling points within a window is specified, and the statistical results of each window are stored in chronological order to form a dataset of signal fluctuation patterns. Sub-step 2.3: An adaptive threshold algorithm is used in conjunction with the temporal variation characteristics of signal strength to identify non-line-of-sight interference components and multipath interference components.

[0036] It should be noted that adaptive threshold algorithms refer to algorithms that dynamically adjust the threshold based on the real-time fluctuation patterns of the signal, making them more adaptable to dynamic signal changes compared to fixed thresholds. Signal strength temporal variation characteristics refer to the trend of signal strength change over time, such as gradual, abrupt, or periodic variations. Non-line-of-sight interference components refer to signal components that cannot propagate directly due to walls or obstacles, and reach the positioning terminal after reflection and diffraction, leading to signal strength attenuation and prolonged propagation time. Multipath interference components refer to interference components formed by the superposition of signal components that have propagated through multiple different paths to the positioning terminal, causing signal waveform distortion.

[0037] In the specific implementation process, based on the fluctuation pattern dataset obtained from sub-step 2.2, the upper limit of the adaptive threshold is calculated. With lower threshold The formula is , ,in This is the threshold coefficient, with an example value. .

[0038] Analyze the temporal variation characteristics of signal strength; if the signal value at a certain moment exceeds the upper limit threshold... or below the lower threshold Furthermore, if the signal strength exhibits abrupt change characteristics, which means that the difference in signal strength between adjacent moments is greater than 0.5V, then the signal component is determined to be a non-line-of-sight interference component.

[0039] If the signal value is within the threshold range but exhibits periodic fluctuations, such as a fluctuation period between 1ms and 10ms and a fluctuation amplitude greater than 0.3V, then the signal component is determined to be a multipath interference component.

[0040] Sub-step 2.4: Dynamically adjust the interference identification threshold based on the signal fluctuation pattern to generate interference compensation parameters that are adapted to the current environment.

[0041] Interference identification threshold refers to the critical value used to determine whether a signal is an interference component, including an upper threshold and a lower threshold. Interference compensation parameters refer to parameters used to compensate for the influence of non-line-of-sight interference components and multipath interference components on the positioning signal, including interference amplitude compensation value and interference time compensation value.

[0042] The level of interference in the indoor environment changes dynamically over time. A fixed interference identification threshold cannot adapt to environmental changes, resulting in a decrease in the accuracy of interference identification. Dynamically adjusting the threshold based on the signal fluctuation pattern can improve the real-time performance and accuracy of interference identification, thereby generating compensation parameters that can accurately counteract the current interference. Therefore, it is necessary to dynamically adjust the threshold and generate interference compensation parameters.

[0043] In the specific implementation process, dynamically adjusting the interference identification threshold means re-statistically analyzing the signal fluctuation pattern every 1 second and updating the upper limit threshold according to the threshold calculation formula in sub-step 2.3. With lower threshold This ensures that the threshold matches the current signal fluctuation state.

[0044] Calculating the interference amplitude compensation value refers to the average amplitude of the statistical non-line-of-sight interference components. The average amplitude of the multipath interference component Interference amplitude compensation value (The weighting coefficients are set according to the degree of interference.)

[0045] The calculated interference time compensation value refers to the average propagation delay of the statistical non-line-of-sight interference components. Average propagation delay of multipath interference components Interference time compensation value (The weighting coefficients are set according to the degree of impact of interference on propagation time).

[0046] Generating interference compensation parameters refers to adjusting the interference amplitude compensation value. Interference time compensation value Combine them to form interference compensation parameters that are adapted to the current environment.

[0047] Step 3: Calibrate the filtered positioning signal by combining the interference compensation parameters, and calculate its real-time position coordinates using the triangulation model.

[0048] Considering that after generating interference compensation parameters and calibrating the signal, solving the coordinates solely through the traditional triangulation model does not fully exploit the high-dimensional temporal characteristics of the positioning signal, and does not consider the carrier phase offset and clock offset that may occur during the positioning signal transmission, these factors will lead to residual errors in the coordinate solution. To solve the above technical problems, in one possible implementation, step 3 includes the following sub-steps: Sub-step 3.1: Call the interference compensation parameters generated in step 2 to calibrate the amplitude and phase of the filtered positioning signal.

[0049] In the specific implementation process, the interference amplitude compensation value generated in step 2 is obtained. Amplitude of the filtered positioning signal Perform calibration; the calibration formula is: ,in This represents the calibrated signal amplitude.

[0050] Phase calibration is achieved by obtaining the interference time compensation value generated in step 2. Combined with the carrier frequency of the signal Calculate phase compensation amount Phase of the filtered positioning signal Perform calibration; the calibration formula is: ,in This represents the calibrated signal phase.

[0051] Sub-step 3.2: Use a lightweight convolutional neural network to extract high-dimensional temporal features of the calibrated positioning signal.

[0052] It should be noted that high-dimensional temporal features refer to the deep features of a positioning signal in the time dimension that cannot be extracted by traditional methods. These features contain more location-related information.

[0053] In the specific implementation process, a lightweight convolutional neural network is constructed: the network consists of an input layer, two convolutional layers, one pooling layer, and an output layer. The input layer has dimensions (1, N, 2), where N is the number of signal sampling points, and 2 corresponds to the amplitude and phase data of the signal, respectively; the first convolutional layer uses 16 3×1 convolutional kernels with ReLU activation function; the pooling layer uses 2×1 max pooling kernels; the second convolutional layer uses 32 3×1 convolutional kernels with ReLU activation function; the output layer is a fully connected layer, outputting a high-dimensional temporal feature vector with a dimension of 64.

[0054] Feature extraction refers to arranging the calibrated positioning signals in chronological order to form (1, N, 2) dimensional input data, which is then fed into a lightweight convolutional neural network. Through forward propagation, the network outputs a 64-dimensional high-dimensional temporal feature vector.

[0055] Sub-step 3.3: Weighted fusion of high-dimensional temporal features and interference compensation parameters to establish a position coordinate correction model.

[0056] In the specific implementation process, the 64-dimensional high-dimensional time series feature vector is combined with the 2-dimensional interference compensation parameters ( Each value is standardized and mapped to the 0-1 range to eliminate dimensional differences.

[0057] Assign weights to the standardized high-dimensional temporal features Assign weights to the standardized interference compensation parameters. According to the formula To integrate, among which This is the standardized high-dimensional temporal feature vector. This is the standardized interference compensation parameter vector. To fuse feature vectors.

[0058] A position coordinate correction model is established, using a linear regression model as the position coordinate correction model, and the fused feature vectors are used. As input, the coordinate correction amount ( As an example, the model formula is: ,in The weight matrix is ​​3×66. The bias vector is 3×1, determined through offline training. and The specific value.

[0059] Sub-step 3.4: Input the calibrated positioning signal into the triangulation model, combine it with the position coordinate correction model to solve the problem, and output the real-time position coordinates of the device.

[0060] Sub-step 3.4: Input the calibrated positioning signal into the triangulation model, combine it with the position coordinate correction model to solve the problem, and output the real-time position coordinates of the device.

[0061] Considering that the calibrated positioning signal can be used to calculate the distance from the positioning terminal to each base station, and that the triangulation model can calculate the initial position coordinates based on multiple sets of distance data, but still has some error, in order to solve this technical problem, in the specific implementation process, the propagation time of the calibrated positioning signal is used as a basis. According to the formula Calculate the distance from the positioning terminal to the i-th base station in the optimal base station combination, where c is the electromagnetic wave propagation speed.

[0062] Let the coordinates of the three base stations in the optimal base station combination be respectively , , The initial position coordinates of the positioning terminal are Establish a system of equations based on the distance formula: ; ; ; Solving this system of equations yields the initial position coordinates. .

[0063] The correction amount output by the model is corrected by calling the position coordinates. According to the formula ; Calculate the real-time location coordinates and output the coordinates.

[0064] During positioning signal transmission, carrier phase offset and clock offset may occur due to differences in hardware equipment, environmental interference, and other factors. These offsets accumulate and affect the accuracy of coordinate calculation. The angular velocity and acceleration timing information in the inertial navigation data can reflect the motion state of the terminal and are related to carrier phase offset and clock offset. Therefore, it is necessary to extract this information for offset prediction. In one possible implementation, the following sub-steps are included after sub-step 3.4: Sub-step 3.5: Extract the angular velocity and acceleration timing information from the inertial navigation data collected in step 1.

[0065] In the specific implementation process, the angular velocity and acceleration data within the most recent 1 second are extracted from the inertial navigation data stored in step 1, and arranged in time stamp order to form an angular velocity time sequence information sequence. (Sampling frequency 100Hz, 100 data points per second) and acceleration time sequence information. Smoothing can be performed on the two sequences, such as by using a moving average method, to remove random noise from the data.

[0066] Sub-step 3.6: Construct a dynamic prediction model for phase deviation based on angular velocity and acceleration time-series information.

[0067] In the specific implementation process, feature extraction is performed on the smoothed angular velocity time series and acceleration time series, and statistical features such as mean, variance, maximum value, minimum value, and rate of change of the sequences are calculated to form a 10-dimensional input feature vector. A Long Short-Term Memory (LSTM) network is used to construct a phase deviation dynamic prediction model, which includes an input layer, one LSTM layer, and one fully connected layer. The input layer has a dimension of 10, the LSTM layer contains 32 hidden units, and the output dimension of the fully connected layer is 2, corresponding to the predicted carrier phase deviation and clock offset values, respectively. The model is trained using offline collected inertial navigation data time series information and corresponding measured carrier phase deviation and clock offset data to optimize the model parameters and ensure that the model prediction error is less than a preset threshold.

[0068] Sub-step 3.7: Estimate the carrier phase offset and clock offset of the positioning signal in real time using the phase offset dynamic prediction model.

[0069] In the specific implementation process, the 10-dimensional input feature vector constructed in sub-step 3.6 is input into the trained phase deviation dynamic prediction model. Through forward propagation calculation, the predicted carrier phase deviation value at the current moment is output. With clock offset prediction The two predicted values ​​are temporarily stored for subsequent compensation.

[0070] Sub-step 3.8: Generate dynamic compensation factors, embed the dynamic compensation factors into the solution process of the triangulation model, and correct the real-time position coordinates of the solution output.

[0071] In the specific implementation process, based on the carrier phase offset prediction value Calculate the phase compensation factor Based on clock offset prediction value Calculate the time compensation factor ( (mean of signal propagation time).

[0072] Time compensation factor The formula used to correct the distance calculation from the positioning terminal to the base station is as follows: Phase compensation factor The phase measurement value of the positioning signal is corrected using the following formula: The corrected distance and phase data are re-input into the triangulation model, and combined with the position coordinate correction model, the calculation is recalculated according to the solution process in sub-step 3.4 to obtain the corrected real-time position coordinates.

[0073] Step 4: Run the pedestrian dead reckoning algorithm based on the inertial navigation data to obtain continuous relative position coordinates. Use the real-time position coordinates as a reference to correct the relative position coordinates and form preliminary motion trajectory data.

[0074] Considering that the pedestrian dead reckoning algorithm does not take into account the differences in the carrying methods of positioning terminals after obtaining high-precision real-time position coordinates, and that the characteristics of inertial navigation data are different under different carrying methods, using a uniform step size estimation model will lead to inaccurate relative position coordinate estimation. In one possible implementation, step 4 includes the following sub-steps: Sub-step 4.1: Extract the acceleration signal from the inertial navigation data collected in step 1.

[0075] In the specific implementation process, the acceleration signal along the direction perpendicular to the ground (z-axis) is extracted from the inertial navigation data stored in step 1. Because the acceleration signal in this direction is most significantly affected by the carrying method, the extracted acceleration signals are arranged in time stamp order to form an acceleration signal sequence. The sequence is then filtered and denoised.

[0076] Sub-step 4.2: Analyze the acceleration signal using the K-nearest neighbor algorithm to identify the carrying method, which includes handheld, arm swing, and pocket modes.

[0077] It should be noted that the K-nearest neighbors algorithm refers to an algorithm that calculates the distance between the sample to be classified and each sample in the training set, selects the K nearest samples, and determines the category of the sample to be classified based on the category of these K samples. It can be used to classify and identify the carrying method. The carrying method refers to how the terminal is carried by the user, including handheld, arm swing, and pocket modes. The motion state and acceleration signal characteristics of the terminal are different in different modes.

[0078] In the specific implementation process, the smoothed acceleration signal sequence Feature extraction is performed, calculating six features of the signal: peak value, trough value, peak interval, mean, variance, and kurtosis, forming a 6-dimensional feature vector. Model training: Offline collection of acceleration signals under three carrying methods—handheld, arm swing, and pocket—is conducted. Feature vectors are extracted to construct a training set, with 1000 training samples corresponding to each carrying method. Euclidean distance is used as the distance metric for the K-nearest neighbor algorithm, with K set to 5.

[0079] The extracted 6-dimensional feature vector is input into the trained K-nearest neighbor algorithm. The Euclidean distance between the vector and each sample in the training set is calculated. The 5 closest samples are selected, and the carrying mode with the highest proportion among these 5 samples is determined as the current carrying mode.

[0080] Sub-step 4.3: Based on the recognition results, call the adapted step size estimation model, which introduces step frequency and acceleration variance parameters.

[0081] It should be noted that the relationship between step length, step frequency, and acceleration variance differs depending on the carrying method. A unified step length estimation model cannot be adapted to all carrying methods. Based on the recognition results, an adapted model is called, and the step frequency and acceleration variance parameters are introduced to improve the accuracy of step length calculation. Therefore, it is necessary to call the adapted model.

[0082] In practical implementation, calculating step frequency refers to calculating the step frequency based on the acceleration signal sequence. peak interval According to the formula Calculate the step frequency (unit: steps / second).

[0083] According to the formula Calculate the variance of the acceleration signal, where For the first in the sequence One acceleration value, The mean of the sequence. The sequence length is given.

[0084] The corresponding step size estimation model is invoked based on the identified carrying method. The models for the three carrying methods are as follows: Step size corresponding to handheld mode ,in , , Calculate the step frequency (unit: steps / second).

[0085] Step length corresponding to swing arm mode ,in , , (Unit: m).

[0086] Step length corresponding to the pocket mode ,in , , (Unit: m).

[0087] Sub-step 4.4: Calculate the step size for each step using the step size estimation model, and deduce the heading angle by combining the angular velocity in the inertial navigation data.

[0088] In the specific implementation process, according to the step size estimation model in sub-step 4.3, the step size of each step of the user is calculated in real time. Each time an acceleration peak is detected, it is determined as a step, and the step size value of that step is output.

[0089] Extracting the angular velocity signal from the inertial navigation data in step 1, mainly using the angular velocity around the z-axis. (Reflecting the terminal's direction), according to the formula Calculate the heading angle, where The heading angle at the current moment. The heading angle at the previous moment. The angular velocity around the z-axis at the current moment. The sampling time interval (10ms) and the initial heading angle The location coordinates are determined by calibrating the real-time location coordinates with the base station coordinates in step 3.

[0090] Sub-step 4.5: Obtain continuous relative position coordinates based on step size and heading angle.

[0091] In the specific implementation process, let the relative position coordinates of the previous step be... The current step size is The heading angle is Calculate the relative position coordinates of the current step using the formula: ; ; ; In indoor positioning, the z-axis coordinate changes little and is temporarily set to remain unchanged; the relative position coordinates of each step are recorded in chronological order to form a continuous sequence of relative position coordinates.

[0092] Sub-step 4.6: Use the real-time position coordinates calculated in step 3 as a reference to correct the continuous relative position coordinates and form preliminary motion trajectory data.

[0093] In the specific implementation process, the correction period is set to 1 second, and the latest real-time position coordinates calculated in step 3 are called every second. The latest relative position coordinates in the continuous relative position coordinate sequence Correct the coordinates to the baseline, and then perform a translation correction on all relative position coordinates before correction according to the formula: ; ; ; in , , These are the relative position coordinates before correction. , , The corrected coordinates are then sorted by timestamps to form preliminary motion trajectory data.

[0094] Step 5: Preset the boundary range and warning triggering conditions of the indoor warning area, compare the real-time location coordinates with the boundary range in real time to determine whether the warning triggering conditions are met; if the warning triggering conditions are met, start the warning response mechanism, and simultaneously record the warning triggering time, warning type and current real-time location coordinates to form a warning event record.

[0095] It should be noted that the boundary range of the warning area refers to the pre-defined boundary coordinates of the indoor area requiring a warning, and the shape of the area (such as a rectangle, circle, or polygon) is defined by multiple boundary point coordinates. Warning triggering conditions refer to the rules used to determine whether a warning needs to be activated, including triggering upon entering the warning area, triggering upon leaving the warning area, and triggering upon staying in the warning area for more than a preset time. The warning response mechanism refers to the warning operation initiated when the warning triggering conditions are met, including sound alarms, vibration alarms, and light alarms. The warning event log refers to data that records key information related to the warning, including the warning trigger time, warning type, and location coordinates at the time of triggering, used for subsequent trajectory tracing and event verification. The warning type refers to the specific category of the warning, including warnings for entering the warning area, leaving the warning area, and exceeding the time limit for staying in the warning area.

[0096] In the specific implementation process, the preset warning area boundary range refers to the warning area information set by the user received by the positioning terminal through the host computer or local configuration interface. If it is a rectangular area, the coordinates of the four vertices are input. , , , If the region is circular, enter the coordinates of the center. With radius Store the boundary coordinates locally.

[0097] The preset warning trigger conditions refer to the warning trigger condition type selected by the user through the configuration interface. If "Enter Warning Zone Trigger" is selected, a warning will be triggered when the real-time location coordinates fall within the boundary range; if "Timeout Stay Trigger" is selected, an additional stay time threshold will be set. (Example value: 5 minutes) When the real-time location coordinates remain within the boundary range for a longer period of time than Timely triggering of alerts; storage of trigger condition configurations.

[0098] The positioning terminal acquires the real-time location coordinates in step 3 at a frequency of 10Hz. Determine whether the coordinates fall within the preset warning area boundary: For a rectangular region, determine Is it in and between, Is it in and between, Is it in and between( The minimum x-coordinate of the four vertices. If the maximum value is 1, and the same applies to the others, then it is determined to be within the region.

[0099] For a circular region, calculate the distance between the real-time coordinates and the coordinates of the center of the circle. ,like Then it is determined to be within the area.

[0100] Based on the preset warning trigger conditions, determine whether the trigger requirements are met. If the warning trigger conditions are met, the positioning terminal activates the warning response mechanism (such as activating the built-in speaker to emit an audible alarm or the vibration motor to generate vibration), and records the warning trigger time (accurate to milliseconds), warning type (such as entering the warning area), and current real-time location coordinates. This information is associated and stored to form a warning event record. If the conditions are not met, return to step 1 to continue monitoring.

[0101] Step 6: Integrate continuous real-time location coordinates and early warning event records to build a complete motion trajectory archive and achieve trajectory tracing.

[0102] In one possible implementation, step 6 includes the following sub-steps: Sub-step 6.1: Collect the continuous real-time position coordinates calculated in step 3, the preliminary motion trajectory data formed in step 4, and the early warning event records generated in step 5.

[0103] In the specific implementation process, the positioning terminal collects the continuous real-time position coordinates of step 3 from the signal processing module, the preliminary motion trajectory data of step 4 from the trajectory generation module, and the early warning event records of step 5 from the early warning module in the order of timestamps. The three types of data are aggregated into the local data storage area and sorted by timestamps to ensure the temporal continuity of the data.

[0104] Sub-step 6.2: Encrypt the collected data using an encryption algorithm.

[0105] In the specific implementation process, the AES-256 encryption algorithm can be used. The AES-256 encryption algorithm refers to the encryption algorithm in the Advanced Encryption Standard (AES) with a key length of 256 bits. It has high encryption strength and can effectively protect data security and prevent data leakage.

[0106] The positioning terminal has a built-in AES-256 encryption module that generates a random 256-bit encryption key and stores it in a secure storage area. It encrypts the collected continuous real-time location coordinates, preliminary motion trajectory data and early warning event records in groups. The encryption process for each group of data is as follows: the data is filled in, and byte substitution, row shifting, column mixing and round key addition operations are performed according to the encryption round of the AES-256 algorithm. The encrypted data is then output and stored.

[0107] Sub-step 6.3: Construct a time-series index based on the timestamps in the encrypted data.

[0108] In the specific implementation process, the timestamp of each record in the encrypted data is extracted, and the timestamps are divided into multiple time intervals according to the time sequence. An index item is created for each time interval. The index item contains the start time of the interval, the end time of the interval, and the storage address of the corresponding encrypted data. All index items are arranged in chronological order to form a time-series index, which is stored in the header area of ​​the encrypted data.

[0109] Sub-step 6.4: Divide the spatial regions based on the location information in the encrypted data and construct a spatial dimension grid index.

[0110] It should be noted that spatial region division refers to dividing the overall indoor positioning area into multiple non-overlapping small grid regions according to fixed dimensions, thereby achieving fine-grained management of the space. Grid indexing refers to an index structure built based on the divided spatial grid. Each grid corresponds to a set of encrypted storage addresses containing the location data within that grid, used for quickly locating trajectory data within a specific spatial region.

[0111] In trajectory tracing scenarios, in addition to querying by time, it is often necessary to query movement trajectories within a specific spatial area (such as the movement of people in a server room or corridor). Without a spatial index, all encrypted data must be traversed to filter out information for the target area, resulting in extremely low query efficiency. By constructing a grid index, spatial location and data storage address can be directly mapped, significantly improving the query speed of the spatial dimension and complementing the time-series index.

[0112] Sub-step 6.5: Integrate temporal index and grid index to construct a complete motion trajectory archive and realize trajectory tracing.

[0113] In the specific implementation process, all index entries of the time-series index and grid index are traversed, and the time interval and grid number corresponding to the same encrypted data block are associated to construct a fusion index table. Each record in the fusion index table contains three parts: time interval identifier, grid number, and encrypted storage address. The fusion index table, encrypted continuous real-time position coordinates, preliminary motion trajectory data, and early warning event records are integrated according to a fixed structure: [fusion index table] + [encrypted trajectory data block] + [encrypted early warning event record] to form a complete motion trajectory archive. The archive can be stored in the local flash memory of the positioning terminal or uploaded to a server for centralized management in an encrypted manner.

[0114] When it is necessary to trace a trajectory, the user enters query conditions (such as time range). spatial region The system performs the following operations: parsing the query conditions to determine the target time interval and target grid range; filtering out records that match both the target time interval and target grid in the fusion index table and obtaining the corresponding encrypted storage address; reading the encrypted data according to the storage address and decrypting it using a preset 256-bit key; restoring the continuous real-time position coordinates, preliminary motion trajectory data, and early warning event records, and displaying the complete trajectory and associated early warning events after sorting by timestamp.

[0115] Example 2 An indoor positioning trajectory tracking and area early warning system, applying the aforementioned indoor positioning trajectory tracking and area early warning method, such as... Figure 2 As shown, it includes: The signal receiving and inertial navigation data acquisition module is used to receive positioning signals using spread spectrum modulation technology transmitted by multi-mode positioning base stations deployed indoors, and simultaneously acquire inertial navigation data from its own inertial measurement unit. The signal filtering and interference compensation parameter generation module is used to filter the positioning signal, identify non-line-of-sight interference components and multipath effect interference components, and generate interference compensation parameters. The positioning signal calibration and real-time coordinate calculation module is used to calibrate the filtered positioning signal in combination with interference compensation parameters and calculate its own real-time position coordinates through a triangulation model. The pedestrian dead reckoning and preliminary trajectory generation module is used to run a pedestrian dead reckoning algorithm based on inertial navigation data, obtain continuous relative position coordinates, use real-time position coordinates as a reference to correct the relative position coordinates, and form preliminary motion trajectory data. The warning condition setting and warning response module is used to preset the boundary range and warning triggering conditions of the indoor warning area, compare the real-time location coordinates with the boundary range in real time, and determine whether the warning triggering conditions are met. If the warning triggering conditions are met, the warning response mechanism is activated, and the warning triggering time, warning type and current real-time location coordinates are recorded simultaneously to form a warning event record. The trajectory data integration and archive construction module is used to integrate continuous real-time location coordinates, preliminary motion trajectory data and early warning event records to construct a complete motion trajectory archive and realize trajectory tracing.

[0116] The embodiments described above are merely illustrative of specific implementations of the present invention, and while the descriptions are detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention.

Claims

1. A method for indoor positioning trajectory tracing and area early warning, characterized in that, Includes the following steps: Step 1: Receive positioning signals using spread spectrum modulation technology transmitted by indoor multi-mode positioning base stations, and simultaneously collect inertial navigation data from its own inertial measurement unit; Step 2: Filter the positioning signal, identify non-line-of-sight interference components and multipath interference components, and generate interference compensation parameters; Step 3: Calibrate the filtered positioning signal by combining interference compensation parameters, and calculate its real-time position coordinates using a triangulation model; Step 4: Run the pedestrian dead reckoning algorithm based on inertial navigation data to obtain continuous relative position coordinates. Use the real-time position coordinates as a reference to correct the relative position coordinates and form preliminary motion trajectory data. Step 5: Preset the boundary range and warning triggering conditions of the indoor warning area, and compare the real-time location coordinates with the boundary range in real time to determine whether the warning triggering conditions are met; If the warning triggering conditions are met, the warning response mechanism will be activated, and the warning triggering time, warning type, and current real-time location coordinates will be recorded simultaneously to form a warning event record; Step 6: Integrate continuous real-time location coordinates and early warning event records to build a complete motion trajectory archive and achieve trajectory tracing.

2. The indoor positioning trajectory tracing and area early warning method according to claim 1, characterized in that, Step 1 includes the following sub-steps: Sub-step 1.1: Receive signals from a multi-base station network consisting of multiple multi-mode positioning base stations; Sub-step 1.2: Calculate the link quality factor and the geometric precision factor. The link quality factor is calculated based on the received signal power, the ratio of direct component to total energy, and the peak delay difference. The geometric precision factor is calculated based on the spatial distribution of the base station. Sub-step 1.3: Construct an objective function based on the link quality factor and geometric precision factor, and use the objective function to select the optimal base station combination; Sub-step 1.4: Receive the positioning signal transmitted by the optimal base station combination using spread spectrum modulation technology; Sub-step 1.5: Collect the angular velocity and acceleration data of its own inertial measurement unit as inertial navigation data.

3. The indoor positioning trajectory tracing and area early warning method according to claim 2, characterized in that, Step 2 includes the following sub-steps: Sub-step 2.1: Preprocess the positioning signal received in step 1 using a low-pass filtering algorithm; Sub-step 2.2: Analyze the fluctuation pattern of the preprocessed positioning signal using a sliding time window; Sub-step 2.3: Using an adaptive threshold algorithm combined with the temporal variation characteristics of signal strength, identify non-line-of-sight interference components and multipath effect interference components; Sub-step 2.4: Dynamically adjust the interference identification threshold based on the signal fluctuation pattern to generate interference compensation parameters that are adapted to the current environment.

4. The indoor positioning trajectory tracing and area early warning method according to claim 3, characterized in that, Step 3 includes the following sub-steps: Sub-step 3.1: Call the interference compensation parameters generated in step 2 to calibrate the amplitude and phase of the filtered positioning signal; Sub-step 3.2: Use a lightweight convolutional neural network to extract high-dimensional temporal features of the calibrated positioning signal; Sub-step 3.3: Weighted fusion of high-dimensional temporal features and interference compensation parameters to establish a position coordinate correction model; Sub-step 3.4: Input the calibrated positioning signal into the triangulation model, combine it with the position coordinate correction model to solve the problem, and output the real-time position coordinates of the device.

5. The indoor positioning trajectory tracing and area early warning method according to claim 4, characterized in that, Sub-step 3.4 is followed by the following sub-steps: Sub-step 3.5: Extract the angular velocity and acceleration timing information from the inertial navigation data collected in step 1; Sub-step 3.6: Construct a dynamic prediction model for phase deviation based on angular velocity and acceleration time-series information; Sub-step 3.7: Estimate the carrier phase offset and clock offset of the positioning signal in real time using the phase offset dynamic prediction model; Sub-step 3.8: Generate dynamic compensation factors, embed the dynamic compensation factors into the solution process of the triangulation model, and correct the real-time position coordinates of the solution output.

6. The indoor positioning trajectory tracing and area early warning method according to claim 5, characterized in that, Step 4 includes the following sub-steps: Sub-step 4.1: Extract the acceleration signal from the inertial navigation data acquired in step 1; Sub-step 4.2: Analyze the acceleration signal using the K-nearest neighbor algorithm to identify the carrying method, which includes handheld, arm swing, and pocket modes; Sub-step 4.3: Based on the recognition results, call the adapted step size estimation model, which introduces step frequency and acceleration variance parameters; Sub-step 4.4: Calculate the step size for each step using the step size estimation model, and deduce the heading angle by combining the angular velocity in the inertial navigation data; Sub-step 4.5: Obtain continuous relative position coordinates based on step size and heading angle; Sub-step 4.6: Use the real-time position coordinates calculated in step 3 as a reference to correct the continuous relative position coordinates and form preliminary motion trajectory data.

7. The indoor positioning trajectory tracing and area early warning method according to claim 6, characterized in that, Step 6 includes the following sub-steps: Sub-step 6.1: Collect the continuous real-time position coordinates calculated in step 3, the preliminary motion trajectory data formed in step 4, and the early warning event records generated in step 5; Sub-step 6.2: Encrypt the collected data using an encryption algorithm; Sub-step 6.3: Construct a time-series index based on the timestamps in the encrypted data; Sub-step 6.4: Divide the spatial regions based on the location information in the encrypted data and construct a spatial dimension grid index; Sub-step 6.5: Integrate temporal index and grid index to construct a complete motion trajectory archive and realize trajectory tracing.

8. An indoor positioning trajectory tracking and area early warning system, characterized in that, An indoor positioning trajectory tracing and area early warning method according to any one of claims 1-7 includes: The signal receiving and inertial navigation data acquisition module is used to receive positioning signals using spread spectrum modulation technology transmitted by multi-mode positioning base stations deployed indoors, and simultaneously acquire inertial navigation data from its own inertial measurement unit. The signal filtering and interference compensation parameter generation module is used to filter the positioning signal, identify non-line-of-sight interference components and multipath effect interference components, and generate interference compensation parameters. The positioning signal calibration and real-time coordinate calculation module is used to calibrate the filtered positioning signal in combination with interference compensation parameters and calculate its own real-time position coordinates through a triangulation model. The pedestrian dead reckoning and preliminary trajectory generation module is used to run a pedestrian dead reckoning algorithm based on inertial navigation data, obtain continuous relative position coordinates, use real-time position coordinates as a reference to correct the relative position coordinates, and form preliminary motion trajectory data. The warning condition setting and warning response module is used to preset the boundary range and warning triggering conditions of the indoor warning area, compare the real-time location coordinates with the boundary range in real time, and determine whether the warning triggering conditions are met. If the warning triggering conditions are met, the warning response mechanism is activated, and the warning triggering time, warning type and current real-time location coordinates are recorded simultaneously to form a warning event record. The trajectory data integration and archive construction module is used to integrate continuous real-time location coordinates, preliminary motion trajectory data and early warning event records to construct a complete motion trajectory archive and realize trajectory tracing.

Images