Smartphone positioning and navigation method and system in non-networked indoor scene

Abstract

The present application relates to the technical field of non-networked indoor autonomous positioning and navigation, and particularly to a smartphone positioning and navigation method and system in a non-networked indoor scene, which comprises the following steps: a smartphone receives data returned by a Bluetooth AOA positioning base station through point-to-point communication, and obtains an AOA position estimation result by solving a single-base-station AOA fusion positioning model; the smartphone collects real-time data of its own IMU and magnetometer, recursively estimates the user's coordinates and heading angle by a pedestrian dead reckoning (PDR) method, and autonomously switches among multi-base-station centroid positioning, AOA fusion positioning and PDR positioning according to the actual AOA base station deployment in the indoor space, so that the position result can effectively cover the entire space scene; a binary raster map is obtained by constructing an indoor map model, position post-processing of the positioning coordinates is performed based on map matching, pedestrian coordinates are obtained, the map model is dataized, map display is realized, and finally pedestrian position display is realized. The present application can reliably operate in a network-lacking scene.

Description

Technical Field

[0001] This invention relates to a method and system for smartphone positioning and navigation in non-networked indoor scenarios, belonging to the field of non-networked indoor positioning and navigation technology. Background Technology

[0002] As urbanization deepens, the building area of ​​large indoor public places such as shopping malls, museums, and hospitals continues to expand, and their spatial layouts are becoming increasingly complex, leading to a rapid increase in pedestrian demand for location-based services (LBS). However, due to factors such as the complexity of indoor environments, high deployment costs of positioning devices, and user privacy protection, a universally applicable indoor positioning and navigation solution has not yet been formed. Solutions often need to be customized for specific scenarios. Even with the rapid development of IoT, AI, and big data technologies, pedestrian positioning technology in indoor public places remains difficult to widely apply, significantly limiting the development of a smart society.

[0003] Current mainstream indoor positioning technologies include Wi-Fi, ultra-wideband (UWB), Bluetooth (BLE), radio frequency identification (RFID), LED lighting, machine vision, and mobile phone built-in sensors (IMU, geomagnetic sensors), but all of these technologies have significant limitations:

[0004] Technologies that rely on specific positioning tags: Wireless positioning technologies such as UWB, RFID, and LED light require users to be equipped with specific signal receiving or transmitting tags to achieve positioning, which significantly reduces the visitor experience (e.g., patent CN120091268A).

[0005] Machine vision technologies require high-precision cameras to achieve large-space coverage without blind spots, and the complexity of multi-camera joint visual positioning algorithms is high, resulting in high system costs and the risk of user privacy leakage (such as patent CN119958570A).

[0006] AR (Augmented Reality) positioning technology relies on the user's smartphone visual sensor and requires the phone camera to be turned on, which also has the problems of privacy exposure and poor user experience (such as patent CN119672264A).

[0007] Inertial navigation technology: Although some studies have improved positioning accuracy by improving algorithms (such as patents CN117232507A and CN113029148A), the inherent integral accumulation error of this technology makes it unsuitable for indoor positioning alone.

[0008] In the context of location and navigation needs for visitors to indoor public places, wireless positioning technologies based on Wi-Fi and Bluetooth BLE have become the mainstream choice because users' smartphones are generally equipped with corresponding transceiver modules, eliminating the need for additional location tags and avoiding privacy exposure. However, this type of technology still faces key bottlenecks:

[0009] Insufficient positioning accuracy and stability: In the past few decades, positioning technologies based on Wi-Fi and Bluetooth have mostly relied on Received Signal Strength Indication (RSSI) algorithms. However, RSSI values ​​are easily affected by indoor obstacles (furniture, walls, people's activities) and electromagnetic interference, making it difficult to meet the requirements for positioning accuracy and stability. Even though researchers have proposed a variety of improved algorithms (such as patents CN108716918A and CN118632344A), the results still cannot meet the requirements for practical applications.

[0010] Offline database construction and maintenance challenges: Wi-Fi fingerprint positioning based on RSSI requires extensive offline database construction work in complex indoor spaces (such as patent CN117998581A), and changes in pedestrian flow and spatial layout adjustments can easily lead to fingerprint database failure, significantly reducing positioning accuracy.

[0011] The drawbacks of passive positioning: Current indoor wireless positioning based on Wi-Fi and Bluetooth BLE is passive, relying on external base stations to transmit signals. The location of the target is calculated by the base station and the backend server. This mode requires the establishment of a local area network to collect observation data from multiple base stations and transmit it to the backend, resulting in high initial installation costs and difficult construction and maintenance (e.g., patents CN120091268A, CN108716918A, CN118632344A, CN117998581A, CN115942247A). Furthermore, because the effective coverage radius of Wi-Fi and Bluetooth base stations is limited, a large number of base stations need to be deployed to improve positioning accuracy, further increasing equipment and construction costs.

[0012] In summary, existing pedestrian positioning solutions for indoor public places generally suffer from four core problems: First, the positioning accuracy is not high, and wireless signals are easily interfered with by obstruction in complex environments, resulting in large errors and poor stability; second, the equipment cost is high, with the number of positioning base stations deployed being directly proportional to the accuracy, leading to large equipment investments; third, deployment and maintenance are complex, involving local area network networking and cabling, resulting in high initial installation costs and maintenance difficulties; and fourth, privacy risks are prominent, as the server backend calculates user location and device data, which can easily lead to privacy leaks (e.g., Wi-Fi positioning requires searching for a large number of hotspots, posing a risk of public network privacy leaks). Summary of the Invention

[0013] To address the problems in the prior art, this invention provides a method and system for smartphone positioning and navigation in non-networked indoor scenarios.

[0014] The technical solution adopted by this invention to solve its technical problem is:

[0015] Smartphone positioning and navigation methods in offline indoor scenarios include:

[0016] S1, the smartphone receives the azimuth, elevation and RSSI data returned by the Bluetooth AOA positioning base station, and obtains the AOA position estimation result by solving the single base station AOA fusion positioning model;

[0017] S2, the smartphone collects real-time data from its own IMU and magnetometer, and recursively estimates the user's coordinates and heading angle using the PDR method based on pedestrian trajectory;

[0018] S3, based on the actual AOA base station deployment in the indoor space, autonomously switches between multi-base station centroid positioning, AOA fusion positioning, and PDR positioning to ensure that its location results can effectively cover the entire spatial scene;

[0019] S4 uses a binary raster map constructed from an indoor map model to perform post-processing on the location coordinates based on map matching to obtain pedestrian coordinates. By digitizing the map model, the map is displayed, ultimately showing the pedestrian's location.

[0020] Furthermore, in S1, the AOA fusion positioning algorithm based on heading projection is used to fuse the AOA estimated position and heading angle data to obtain the AOA fusion positioning solution position coordinates;

[0021] The AOA fusion positioning algorithm for heading projection fuses the estimated position and heading angle data, including:

[0022] Data preprocessing involves preprocessing the azimuth, elevation, and RSSI signals from the base station.

[0023] AOA fusion positioning based on heading projection integrates the current heading angle calculated from IMU and magnetometer data, and uses geometric projection transformation to effectively constrain the AOA estimated coordinates of a single base station.

[0024] Further, S1 includes:

[0025] S11, the smartphone receives Bluetooth data packets returned by the positioning base station and obtains azimuth, elevation and RSSI data;

[0026] S12, preprocess the data and compensate for the installation angle errors caused by the base station installation;

[0027] S13 performs azimuth rate filtering and elevation mean filtering on the corrected azimuth and elevation angles, rate filtering and mean filtering on the RSSI data, removes pulse spikes and high-frequency jitter caused by multipath, and outputs smoothed azimuth, elevation and RSSI data.

[0028] S14. The smoothed data is fed into the AOA positioning model to calculate the estimated position coordinates.

[0029] Furthermore, the preprocessing procedure for installation angle compensation includes:

[0030] S121, Smartphone acquires angle data from stationary fixed test points. Perform single-point AOA calculation to obtain the positioning result;

[0031] S122, Remove outlier angle values ​​that are outside the map range;

[0032] S123, regarding the retained angle data Gaussian filtering is used to remove outliers to reduce data fluctuations, resulting in... ;

[0033] S124, Use the calibration system to obtain the true angle value of the fixed test point. , combined The installation angle compensation value of the base station was obtained using the least squares method. ;

[0034] S125, Measure the angle data of n fixed points for each base station and calculate its compensation value. The final installation compensation angles for each base station were obtained using the interquartile range method. .

[0035] Furthermore, the angular rate filtering processing method includes:

[0036] S131, Set two angular rate thresholds Different processing methods are applied to data in different rate ranges.

[0037]

[0038] Perform outlier identification and processing;

[0039] S132 uses a sliding window mean filter for pitch angle data to make the pitch angle data output smooth.

[0040] Further, S2 includes:

[0041] S21. The smartphone acquires the three-axis acceleration data output by the IMU, integrates the vertical acceleration to obtain the velocity curve, and uses the peak detection function combined with the time threshold and the peak gap threshold to identify the number of steps.

[0042] S22. Using the Weinberg step size estimation model, the forward acceleration peak value is... Valley value Substitute the values ​​into the model and calculate the step size;

[0043] S23. The smartphone acquires the initial heading calculated by the accelerometer and magnetometer, and then calculates the relative heading value from the triaxial angular velocity data, thereby obtaining the current pedestrian heading angle. ;

[0044] S24, from the initial coordinates and step length and heading angle Calculate the pedestrian's position coordinates.

[0045] Furthermore, the full-space fusion positioning in S3, which autonomously switches between multiple positioning strategies for positioning coordinates, includes:

[0046] S31. When a smartphone starts location navigation, it receives data sent by multiple activated location base stations.

[0047] S32. Preprocess all base station RSSI values ​​to obtain It can autonomously switch to the optimal positioning strategy based on the size of the preprocessed RSSI data;

[0048] S33. Multi-base station centroid localization calculates the centroid of multiple base station AOA positioning points, and removes the maximum outliers as the final positioning result.

[0049] Furthermore, the location post-processing in S4 includes: applying location smoothing and map matching to the location results of all positioning strategies to obtain a more accurate smooth location estimate.

[0050] A system for implementing the above-described smartphone positioning and navigation method in offline indoor scenarios, the system comprising:

[0051] Lighting power supply, providing power for the long-term continuous operation of the positioning base station;

[0052] The Bluetooth AOA positioning base station acquires phase data and calculates the Bluetooth signal angle of arrival of the mobile phone. It returns the calculated data to the mobile phone via Bluetooth communication and has a communication module that supports Bluetooth 4.2 and above protocols.

[0053] User smartphones equipped with Bluetooth 4.2 or higher communication modules calculate position based on the fusion of their own inertial navigation sensor suite and magnetometer data.

[0054] Furthermore, the Bluetooth AOA positioning base station uses eight receiving antenna units and utilizes orthogonal demodulation technology to acquire phase data and calculate the signal angle of arrival. The calculated signal angle of arrival includes azimuth, elevation, and RSSI data.

[0055] Furthermore, the Bluetooth AOA positioning base station is installed on the ceiling at the center line of the indoor road. At least one Bluetooth AOA positioning base station needs to be deployed at road corners and forks. The Bluetooth AOA positioning base stations are deployed at set intervals.

[0056] The beneficial effects of this invention are:

[0057] 1. This invention utilizes a novel offline array antenna Bluetooth AOA positioning base station as the main measuring device. The base station can communicate point-to-point with a smartphone via Bluetooth without needing to connect to a network, and the smartphone calculates and displays the location coordinates. This invention simplifies the base station deployment scheme and reduces the initial system installation cost and workload. Moreover, this technology does not rely on any network environment (internal or public network), ensuring reliable operation even in scenarios with insufficient network coverage, while also protecting user privacy.

[0058] 2. This invention is for pedestrian positioning and navigation in public indoor spaces. The system hardware only includes a small number of Bluetooth AOA positioning base stations and a smartphone carried by the pedestrian. Users do not need to purchase additional hardware to achieve autonomous positioning and navigation on the terminal, which significantly reduces the system hardware investment cost.

[0059] 3. This invention fully utilizes the geometric accuracy of AOA measurement and positioning technology (compared to traditional RSSI measurement), and performs fusion positioning based on the AOA positioning model, thereby improving the accuracy of position estimation. Simultaneously, through the fusion of multi-source information from AOA, IMU, and magnetometer data, it achieves autonomous switching between AOA and PDR positioning. Continuous and stable position estimation in large spaces can be achieved with only a small number of positioning base stations, reducing the number of base stations required and significantly lowering positioning costs in indoor scenarios.

[0060] 4. In the design of the positioning algorithm, this invention focuses on low computational complexity, abandons the popular deep neural network method, and replaces it with improved filtering and geometric transformation to minimize system latency and ensure real-time updates of positioning information. Attached Figure Description

[0061] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0062] Figure 1 This is a schematic diagram of the hardware architecture of a smartphone positioning and navigation system based on point-to-point communication in the smartphone positioning and navigation method for non-networked indoor scenarios of the present invention.

[0063] Figure 2 This is a schematic diagram of the installation and deployment of the Bluetooth AOA positioning base station in the smartphone positioning and navigation method for non-networked indoor scenarios of the present invention;

[0064] Figure 3 This is a structural diagram of the high-precision smartphone positioning algorithm based on multi-source fusion and map matching in the smartphone positioning and navigation method for non-networked indoor scenarios of the present invention;

[0065] Figure 4 This is a flowchart of the AOA-based position calculation method in the smartphone positioning and navigation method for non-networked indoor scenarios of the present invention;

[0066] Figure 5 This is a flowchart of the installation angle compensation process in the smartphone positioning and navigation method for non-networked indoor scenarios of the present invention.

[0067] Figure 6 This is a flowchart of the PDR (Pedestrian Track Calculation) method in the smartphone positioning and navigation method for non-networked indoor scenarios of the present invention.

[0068] Figure 7 This is a schematic diagram of AOA fusion positioning based on heading projection in the smartphone positioning and navigation method for non-networked indoor scenes of the present invention;

[0069] Figure 8 This is a flowchart of the full-space fusion positioning process for autonomous switching of multiple positioning strategies in the smartphone positioning and navigation method in non-networked indoor scenarios of the present invention.

[0070] Figure 9 for Figure 3 A magnified view of part a in the middle. Detailed Implementation

[0071] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0072] This invention addresses the problems and bottlenecks of existing technologies for pedestrian positioning in indoor public places, and implements a smartphone positioning and navigation method and system that can be used in non-networked public indoor locations. The system employs a new offline Bluetooth AOA positioning base station. The user's smartphone communicates with the base station in real-time, point-to-point communication using Bluetooth 4.2 or higher protocol to obtain AOA (Angle of Arrival) angle data. Simultaneously, it integrates data from the phone's IMU and magnetometer. The mobile terminal autonomously calculates the pedestrian's location based on multi-source fusion, allowing the user to obtain their location offline.

[0073] The technology of this invention belongs to an autonomous positioning method based on wireless signals. It features low equipment cost, simple deployment, and achieves high-precision user autonomous positioning and navigation without requiring networking or external network environment, while eliminating the issue of user privacy leakage. The complete implementation method is further described below.

[0074] like Figure 1As shown, the present invention provides a smartphone positioning and navigation system for non-networked indoor scenarios. The system hardware includes a lighting power supply, a Bluetooth AOA positioning base station, and a user smartphone carrying a Bluetooth 4.2 or higher communication module.

[0075] The lighting power supply provides power for the long-term continuous operation of the positioning base station. The Bluetooth AOA positioning base station is a new type of array antenna positioning base station. It uses 8 receiving antenna units, uses orthogonal demodulation (IQ) technology to obtain phase data and calculate the signal angle of arrival (AOA), and has a communication module that supports Bluetooth 4.2 and above protocols.

[0076] Smartphones refer to mobile phones carried by users in public places, requiring a communication module that supports Bluetooth 4.2 or higher. The positioning base station and the smartphone communicate point-to-point in real-time via Bluetooth to acquire and calculate the azimuth angle (Azi), pitch angle (Ele), and RSSI data of the Bluetooth signal arrival point. This raw data is then returned to the phone via Bluetooth. The phone uses this Azimuth data, combined with its own inertial navigation sensor suite (IMU) and magnetometer data, to calculate its position.

[0077] The installation and deployment diagram of the Bluetooth AOA positioning base station is as follows: Figure 2 As shown in the diagram, B131.... represents parking space or shop number, #1.... represents Bluetooth AOA positioning base station, and xoy is the local two-dimensional positioning coordinate system. Bluetooth AOA positioning base stations are installed in a single or double row (depending on road width) on the ceiling at the center line of the indoor road. Generally, at least one Bluetooth AOA positioning base station needs to be deployed at road corners and forks; in other locations, the interval between Bluetooth AOA positioning base stations should be less than 8 meters (different values ​​are taken depending on the complexity of the deployment environment).

[0078] To achieve high-precision indoor pedestrian positioning in non-networked scenarios, this invention fully utilizes the built-in Bluetooth communication module, inertial navigation sensor kit, magnetometer, and offline map of smartphones to develop a high-precision positioning algorithm based on multi-source fusion and map matching, which is deployed on smartphones.

[0079] like Figure 3 and Figure 9 As shown, this invention provides a smartphone positioning and navigation method for offline indoor scenarios, including:

[0080] S1, the smartphone receives the azimuth, elevation and RSSI data returned by the Bluetooth AOA positioning base station, and obtains the AOA position estimation result by solving the single base station AOA fusion positioning model; in order to improve the positioning accuracy of single base station AOA, the AOA fusion positioning algorithm with heading projection is used to fuse the AOA estimated position and heading angle data to obtain the AOA fusion positioning solution position coordinates.

[0081] S2. Simultaneously, the smartphone collects real-time data from its own IMU and magnetometer, and recursively estimates the user's coordinates and heading angle using the PDR method based on pedestrian trajectory.

[0082] S3. Finally, based on the actual deployment of AOA base stations in the indoor space, the system autonomously switches between multi-base station centroid positioning, AOA fusion positioning, and PDR positioning to ensure that the location results can effectively cover the entire spatial scene. Multi-base station centroid positioning is a method to obtain the user's location at the initial moment when the user starts positioning and navigation. It calculates the centroid of the positions solved by multiple AOA base stations as the user's coordinates.

[0083] S4 uses a binary raster map constructed from an indoor map model to perform post-processing on the location coordinates based on map matching to obtain pedestrian coordinates. Simultaneously, by digitizing the map model, the map is displayed, ultimately achieving the display of pedestrian locations.

[0084] This positioning algorithm has low computational complexity and low computing resource requirements. It does not affect the functionality of smartphones. External devices only need a Bluetooth positioning base station with communication coverage. The total system cost is extremely low and the installation and deployment are simple. Moreover, the smartphone can perform positioning calculations autonomously, protecting user privacy.

[0085] For step S1, as follows Figure 4 As shown, the AOA-based location calculation method flowchart demonstrates all the steps involved in a single Bluetooth AOA positioning base station and a smartphone collaboratively completing a full Bluetooth AOA positioning calculation. Specifically, it includes:

[0086] The smartphone receives Bluetooth data packets returned by the positioning base station to obtain azimuth, elevation, and RSSI data. It then enters a data preprocessing stage, compensating for azimuth and elevation angle errors caused by base station installation. The corrected azimuth and elevation angles undergo azimuth rate filtering and elevation mean filtering, while the RSSI data undergoes rate filtering and mean filtering to remove pulse spikes and high-frequency jitter caused by multipath propagation, outputting smoothed azimuth, elevation, and RSSI data. The smoothed data is then fed into the AOA positioning model to calculate estimated position coordinates.

[0087] (1) Regarding the above installation angle compensation, due to installation errors, the zero-degree angle direction of the base station may not be completely consistent with the X-axis direction of the coordinate axis. This results in a certain error between the measured value and the true value in the data. Therefore, it is necessary to calculate the compensation angle to eliminate the installation error. The zero-degree direction of the pitch angle also has an installation angle error with respect to the horizontal normal.

[0088] The preprocessing procedure for installation angle compensation is as follows: Figure 5 As shown. The smartphone acquires angle data from a stationary, fixed test point. Single-point AOA calculation is performed to obtain the positioning results, and outlier angle values ​​outside the map range are removed. Then, the retained angle data... Gaussian filtering is used to remove outliers to reduce data fluctuations, resulting in... The true angle value of this fixed test point is obtained using a calibration system. , combined The installation angle compensation value of the base station was obtained using the least squares method. For each base station, measure the angle data of n fixed points and calculate its compensation value. The final installation compensation angles for each base station were obtained using the interquartile range method. .

[0089] (2) Regarding angular rate filtering and mean filtering, due to multipath interference from indoor obstacles, people, etc., the azimuth angle returned by the Bluetooth AOA base station to the smartphone is... Pitch angle Both RSSI data and other data contain various interference signals such as abnormal pulse spikes, continuous oscillations, and Gaussian noise. Therefore, the raw data must be preprocessed before it can be input into the AOA positioning model.

[0090] For azimuth data, this system uses angular rate filtering to identify and process outliers. The rate of change of the observed azimuth values ​​is calculated. (Angular velocity)

[0091] (2-1)

[0092] in, It is the difference between the current angle value and the previous angle value. For time intervals.

[0093] Set two angular velocity thresholds Different processing methods are applied to data in different rate ranges.

[0094] (2-2)

[0095] in, The current observation value for the data point. This is the filtered result for the current data point. These are the filter coefficients.

[0096] For pitch angle data, a sliding window mean filter is used to smooth the pitch angle data output.

[0097] (2-3)

[0098] in, The pitch angle data is within the window, and L represents the length of the sliding window.

[0099] For RSSI data, outlier identification and processing are first performed as shown in Equation 2-2, and then mean filtering is performed as shown in Equation 2-3 to smooth the RSSI data.

[0100] (3) Regarding the single-base station AOA positioning model, the single-base station AOA positioning model can calculate two-dimensional coordinates (x, y) parallel to the ground plane. For user positioning and navigation in public places, these two-dimensional coordinates are sufficient. The z-axis height of the signal source (smartphone) is set to 1.2m, which is consistent with the height at which most pedestrians hold their phones. Assume the positioning base station coordinates are (x1, y2). 1, The source coordinates are (x, y, z) (z = 1.2 m). The azimuth angle between the source and the base station is... Pitch angle is Based on geometric relationships, a system of equations can be obtained.

[0101] (2-4)

[0102] Solving this problem yields the two-dimensional coordinates of the information source.

[0103] (2-5)

[0104] (4) Regarding the AOA fusion positioning model for heading projection, due to environmental interference with radio waves, the AOA positioning model (such as...) Figure 4 The results of this method exhibit instability such as random jumps and back-and-forth movements. Therefore, this invention employs an AOA fusion positioning algorithm based on heading angle projection to mitigate position instability and improve positioning accuracy. For example... Figure 7 As shown, the AOA fusion positioning model of heading projection improves the stability and accuracy of positioning by combining the AOA positioning results of a single base station with pedestrian heading angle data.

[0105] The specific steps include:

[0106] S31. Assume the positioning base station is located at A, with coordinates (x1, y1, z1). The location of the base station at time t-1 is known. Coordinates are At time t, base station A obtains the observed value. and .

[0107] S32, Regarding the observed values and After data preprocessing, the coordinates of point N at time t can be calculated using Equation 2-5.

[0108]

[0109] S33. From step S23, the heading angle at time t−1 can be calculated as follows: Then the position at time t must be at the point where The vertex is and the angle of departure is . On the rays.

[0110] S34. Typically, the Bluetooth radio wave signal path is highly unstable due to environmental factors, causing the actual positioning point N to often not be on its flight path, but rather exhibiting a jumping pattern. Therefore, a projection method is used to project the AOA positioning result N onto the aforementioned ray, with the projected point N' being an estimated value of the position coordinates.

[0111]

[0112] For step S2, pedestrian heading estimation and track calculation specifically include:

[0113] By utilizing the inertial navigation sensor suite (IMU) and magnetometer built into a smartphone, three-axis acceleration, angular velocity, and magnetic vector can be collected. The pedestrian position coordinates can be recursively obtained using the pedestrian trajectory estimation PDR method, and real-time estimates of the pedestrian's heading angle can also be obtained. Figure 6 The flowchart for calculating PDR (Pedestrian Track Determination) for pedestrian tracks is shown below, with the specific steps as follows:

[0114] S21. The smartphone acquires the three-axis acceleration data output by the IMU, integrates the vertical acceleration to obtain the velocity curve, and uses the peak detection function combined with the time threshold and the peak gap threshold to identify the number of steps.

[0115] S22. Using the Weinberg step size estimation model, the forward acceleration peak value is... Valley value Substitute into the model and calculate the step size.

[0116]

[0117] in, Let H be the step size estimated for the k-th step, and H be the parameters of the Weinberg step size model.

[0118] S23. The smartphone acquires the initial heading calculated by the accelerometer and magnetometer, and then calculates the relative heading value from the triaxial angular velocity data, thereby obtaining the current pedestrian heading angle. .

[0119] S24, from the initial coordinates and step length and heading angle Calculate pedestrian position coordinates

[0120]

[0121] Regarding the autonomous switching positioning method using multiple positioning strategies in step S3, the above content achieves AOA fusion positioning (AOA fusion positioning for short) and PDR positioning based on pedestrian trajectory estimation. However, since the two positioning technologies have their own applicable ranges, simply using the aforementioned modules cannot achieve positioning and navigation covering the entire area. Bluetooth AOA base station communication coverage radius is limited, and poor Bluetooth signal quality in many positioning areas leads to positioning errors. Achieving full spatial coverage through dense deployment of positioning base stations would result in a geometric increase in system costs; the inherent integral accumulation error problem in PDR positioning causes its positioning error to increase significantly over time.

[0122] Based on this, the present invention fully combines the advantages of each module's positioning technology, performs post-processing based on the position coordinates output by each module, and forms a complete fusion positioning process covering the entire space, thereby eliminating the technical barriers existing in single positioning technologies and improving the accuracy of indoor positioning. For example... Figure 8 The diagram shows the full-space fusion positioning process with autonomous switching of multiple positioning strategies. The specific steps are as follows:

[0123] a1. When a smartphone initiates location navigation, it receives data (azimuth, elevation, and RSSI values) from n activated location base stations.

[0124] a2. Preprocess all base station RSSI values ​​to obtain (i=1, ..., n). For i=1, ..., n, check each base station one by one: if for all i=1, ..., n, ... < This indicates that no base station has met the activation conditions, so multi-base station AOA centroid positioning is enabled. The positioning results are output and displayed after position smoothing and map matching.

[0125] a3. For all i=1, ...,n, if there exists ≥ (Preset RSSI threshold) indicates that the received signal strength of base station i meets the activation condition, and the AOA fusion positioning of base station i is enabled (i.e., AOA fusion positioning of heading projection). The positioning result is output and displayed after position smoothing and map matching.

[0126] a4. While displaying the AOA fusion positioning result of the i-th base station, continuously judge its Size, once < If the signal strength received by base station i is insufficient, it will become inactive. In this case, PDR positioning is activated, and the current location coordinates of base station i are used as the initial value (x) for PDR calculation. 0, The location coordinates are output after the positioning result is smoothed and matched with the map (y0).

[0127] The a3~a4 loop repeats repeatedly, and the user's location calculation autonomously switches between AOA fusion positioning and PDR positioning, continuously outputting the positioning coordinates.

[0128] Figure 8 The specific implementation methods of multi-base station AOA centroid localization, position smoothing, and map matching techniques in the full-space fusion positioning algorithm shown are described below.

[0129] (1) Multi-base station AOA centroid positioning

[0130] If the user is not within the high-precision coverage area of ​​any AOA positioning base station when initiating location navigation, a multi-base station AOA centroid positioning strategy will be used to obtain the user's current location. The smartphone simultaneously receives data from two or more positioning base stations, and < (for i) When the system uses multi-base station AOA centroid positioning, it employs this method.

[0131] First, the two-dimensional position coordinates of each of the n base stations are calculated using a single-base station AOA positioning model. (i=1,…,n), calculate their average value to obtain the initial centroid coordinates.

[0132]

[0133] Secondly, calculation Euclidean distance between (i=1,…,n) and the centroid

[0134] , i=1,…,n

[0135] Finally, delete Location points, only those with a distance less than a threshold are retained. The location points. For the m retained location points... Calculate the mean of (i=1,…,m) and use it as the final centroid localization estimate.

[0136]

[0137] (2) Location smoothing and map matching

[0138] Position smoothing: The calculated position coordinates are smoothed using an improved low-pass filter. The rate of change of the position coordinates is calculated. (Rate), set a threshold for the rate of change of position coordinates. Low-pass filtering is applied to only points whose rate of change of position exceeds a threshold. Taking the x-axis coordinate as an example, the filtering formula is as follows:

[0139]

[0140] in, The coordinates are before filtering. These are the filtered coordinates. These are the filter coefficients.

[0141] Map matching: The indoor spatial map model is digitized and converted into a binary raster map. Using the binary raster map as a constraint, the coordinates of the location points at obstacles are mapped to the nearest obstacle edge using the vertical projection method. Finally, the estimated position coordinates are output after map matching.

[0142] This invention relates to a smartphone positioning and navigation method and system based on point-to-point communication in non-networked indoor scenarios. It utilizes a novel offline array antenna Bluetooth AOA positioning base station as the main measuring device. The system hardware consists of only a few such base stations and a smartphone carried by a pedestrian. The Bluetooth AOA positioning base station communicates point-to-point with the user's smartphone via Bluetooth. The base station is responsible for calculating the angle of arrival of the signal source (user's smartphone) and returning data packets, while the smartphone receives the data packets and calculates its own position coordinates. An AOA fusion positioning algorithm based on heading projection is used to achieve high-precision position calculation within the base station's coverage area. Indoor map information is incorporated, and a full-space fusion positioning method based on post-processing of the position is used to achieve the optimal selection of multiple fusion positioning strategies.

[0143] Although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A smartphone positioning and navigation method for offline indoor scenarios, characterized in that: include: S1, the smartphone receives the azimuth, elevation and RSSI data returned by the Bluetooth AOA positioning base station, and obtains the AOA position estimation result by solving the single base station AOA fusion positioning model; S2, the smartphone collects real-time data from its own IMU and magnetometer, and recursively estimates the user's coordinates and heading angle using the PDR method based on pedestrian trajectory; S3, based on the actual AOA base station deployment in the indoor space, autonomously switches between multi-base station centroid positioning, AOA fusion positioning, and PDR positioning to ensure that its location results can effectively cover the entire spatial scene; S4 uses a binary raster map constructed from an indoor map model to perform post-processing on the location coordinates based on map matching to obtain pedestrian coordinates. By digitizing the map model, the map is displayed, ultimately showing the pedestrian's location. In S1, the AOA fusion positioning algorithm based on heading projection is used to fuse the AOA estimated position and heading angle data to obtain the AOA fusion positioning solution position coordinates. The AOA fusion positioning algorithm for heading projection fuses the estimated position and heading angle data, including: Data preprocessing involves preprocessing the azimuth, elevation, and RSSI signals from the base station. AOA fusion positioning based on heading projection integrates the current heading angle calculated from IMU and magnetometer data, and uses geometric projection transformation to effectively constrain the AOA estimated coordinates of a single base station; S1 includes: S11, the smartphone receives Bluetooth data packets returned by the positioning base station and obtains azimuth, elevation and RSSI data; S12, preprocess the data and compensate for the installation angle errors caused by the base station installation; S13 performs azimuth rate filtering and elevation mean filtering on the corrected azimuth and elevation angles, rate filtering and mean filtering on the RSSI data, removes pulse spikes and high-frequency jitter caused by multipath, and outputs smoothed azimuth, elevation and RSSI data. S14. The smoothed data is fed into the AOA positioning model to calculate the estimated position coordinates.

2. The smartphone positioning and navigation method in a non-networked indoor scene according to claim 1, characterized in that: The preprocessing procedure for installation angle compensation includes: S121, Smartphone acquires angle data from stationary fixed test points. Perform single-point AOA calculation to obtain the positioning result; S122, Remove outlier angle values ​​that are outside the map range; S123, regarding the retained angle data Gaussian filtering is used to remove outliers to reduce data fluctuations, resulting in... ; S124, Use the calibration system to obtain the true angle value of the fixed test point. , combined The installation angle compensation value of the base station was obtained using the least squares method. ; S125, Measure the angle data of n fixed points for each base station and calculate its compensation value. The final installation compensation angles for each base station were obtained using the interquartile range method. .

3. The smartphone positioning and navigation method in a non-networked indoor scene according to claim 1, characterized in that: The angular rate filtering processing method includes: S131, Set two angular rate thresholds Different processing methods are applied to data in different rate ranges. ； Perform outlier identification and processing; S132 uses a sliding window mean filter for pitch angle data to make the pitch angle data output smooth.

4. The smartphone positioning and navigation method in a non-networked indoor scene according to claim 1, characterized in that: The full-space fusion positioning in S3, which autonomously switches between multiple positioning strategies for positioning coordinates, includes: a1. When a smartphone starts location navigation, it receives data sent by multiple activated location base stations. a2. Preprocess all base station RSSI values ​​to obtain It can autonomously switch to the optimal positioning strategy based on the size of the preprocessed RSSI data; a3. Multi-base station centroid localization: Calculate the centroid of multiple base station AOA positioning points, and remove the maximum outliers as the final positioning result.

5. The smartphone positioning and navigation method in a non-networked indoor scene according to claim 1, characterized in that: The location post-processing in S4 includes: applying location smoothing and map matching to the location results of all positioning strategies to obtain a more accurate smooth location estimate.

6. A system for implementing the smartphone positioning and navigation method in a non-networked indoor scene as described in any one of claims 1-5, characterized in that: The system includes: Lighting power supply, providing power for the long-term continuous operation of the positioning base station; The Bluetooth AOA positioning base station acquires phase data and calculates the Bluetooth signal angle of arrival of the mobile phone. It returns the calculated data to the mobile phone via Bluetooth communication and has a communication module that supports Bluetooth 4.2 and above protocols. User smartphones equipped with Bluetooth 4.2 or higher communication modules calculate position based on the fusion of their own inertial navigation sensor suite and magnetometer data.

7. The system according to claim 6, characterized in that: The Bluetooth AOA positioning base station uses 8 receiving antenna units and uses orthogonal demodulation technology to obtain phase data and calculate the signal angle of arrival. The calculated signal angle of arrival includes azimuth, elevation and RSSI data.

8. The system according to claim 7, characterized in that: The Bluetooth AOA positioning base station is installed on the ceiling at the center line of the indoor road, and at least one Bluetooth AOA positioning base station is deployed at road corners and forks. The Bluetooth AOA positioning base stations are deployed at set intervals.

Images