A positioning device position planning method and device, electronic equipment and medium

Abstract

The application provides a positioning device position planning method and device, electronic equipment and medium, the method comprises: acquiring indoor three-dimensional map and wireless signal data, acquiring environment feature data according to the indoor three-dimensional map, inputting the environment feature data and the wireless signal data into the trained positioning device position prediction model for processing, outputting the initial positioning device position, calculating the signal coverage intensity, the positioning device uniformity and the positioning device deployment cost in the indoor three-dimensional map according to the initial positioning device position, and then calculating the positioning device fitness, adjusting the initial positioning device position according to the positioning device fitness, and obtaining the target positioning device position. The application utilizes accurate three-dimensional map data and positioning device position prediction model to realize automatic planning of the positioning device position, can obtain the best positioning device arrangement scheme through optimization, and ensures the comprehensiveness and uniformity of the positioning signal coverage.

Description

Technical Field

[0001] This invention relates to the field of positioning device deployment, and in particular to a positioning device location planning method, apparatus, electronic device and medium. Background Technology

[0002] In the current deployment of indoor positioning systems, the commonly used technical solution is to first use drawing software to create an indoor floor plan, and then have technicians manually design the physical installation locations of the positioning devices based on experience. Since practical applications often involve the planning of dozens or even hundreds of positioning devices, the existing planning method is inefficient and requires a large amount of manpower. Furthermore, the current planning method requires technicians to combine manual observation of the indoor environment to determine whether the placement of the positioning devices can provide complete coverage, making it impossible to confirm whether the placement of the positioning devices is optimal. This leads to situations where some locations have no signal coverage or poor signal quality, or where too many positioning devices cover a certain area, increasing coverage costs.

[0003] Current technologies heavily rely on the experience of technicians and on-site surveys to plan the installation locations of positioning equipment. This approach is prone to irrational site planning, potentially leading to incomplete signal coverage or multipath effects due to building obstructions, thus reducing the accuracy of the positioning system. Secondly, positioning accuracy is affected by the discrepancy between drawings and the actual indoor environment, as well as the difficulty in precisely installing equipment according to the planned locations. The combined effect of these two errors significantly reduces the final positioning accuracy. Furthermore, current technology lacks effective automation tools and methods to simplify the deployment and maintenance process, which not only increases deployment complexity and cost but also affects system stability and maintainability. Summary of the Invention

[0004] In view of the above problems, embodiments of this application propose a positioning device location planning method, apparatus, electronic device and medium.

[0005] In a first aspect of this application, a positioning device location planning method is provided, characterized in that the method includes:

[0006] Acquire indoor 3D maps and wireless signal data;

[0007] Environmental feature data are obtained based on the indoor 3D map;

[0008] The environmental feature data and the wireless signal data are input into the trained positioning device location prediction model for processing, and the initial positioning device location is output.

[0009] Calculate the signal coverage strength, positioning device uniformity, and positioning device deployment cost in the indoor 3D map based on the initial positioning device location;

[0010] The adaptability of the positioning device is calculated based on the signal coverage strength, the uniformity of the positioning device, and the deployment cost of the positioning device.

[0011] The initial positioning device position is adjusted according to the positioning device adaptability to obtain the target positioning device position.

[0012] Optionally, the indoor 3D map is acquired via a vehicle, which includes a lidar and a camera. Acquiring the indoor 3D map and wireless signal data includes:

[0013] Obtain the preset driving route;

[0014] During the driving process along the described route, three-dimensional point cloud data is collected by the lidar and two-dimensional image data is collected by the camera.

[0015] An initial 3D map is synthesized based on the 3D point cloud data and the 2D image data;

[0016] Feature extraction is performed on the three-dimensional point cloud data to obtain three-dimensional feature data;

[0017] The initial 3D map is processed based on the 3D feature data to obtain the indoor 3D map;

[0018] The wireless signal data is obtained based on the preset attribute information of the positioning device.

[0019] Optionally, the step of extracting features from the three-dimensional point cloud data to obtain three-dimensional feature data includes:

[0020] Voxel data and original point location data are obtained from the three-dimensional point cloud data;

[0021] Feature extraction is performed on the voxel data to obtain candidate region features;

[0022] Feature extraction is performed on the original point data to obtain key point features;

[0023] The three-dimensional feature data is determined based on the candidate region features and the key point features.

[0024] Optionally, the location prediction model for the positioning device is trained in the following manner:

[0025] Obtain a standard indoor 3D map, wherein the standard indoor 3D map includes a number of standard positioning devices arranged at the locations of standard indoor positioning devices;

[0026] Obtain standard environmental feature data based on the standard indoor 3D map;

[0027] Obtain standard wireless signal data based on the preset attribute information of the standard positioning device;

[0028] At least one of the standard environmental feature data, the standard wireless signal data, and the standard indoor positioning device location is preprocessed, and the preprocessing includes one or more of the following: data cleaning, missing data imputation, normalization, and classification coding.

[0029] Based on the preprocessed standard environmental feature data, standard wireless signal data, and standard indoor positioning device location, obtain training environmental feature data, training wireless signal data, training indoor positioning device location, and verification environmental feature data and verification wireless signal data.

[0030] The training environment feature data, the training wireless signal data, and the location of the indoor positioning device are used as training samples.

[0031] The training environment feature data and the training wireless signal data are used as inputs to the positioning device location prediction model, and the location of the positioning device in the training room is used as the output of the positioning device location prediction model. The positioning device location prediction model is trained using the training samples to obtain the positioning device location prediction model to be verified.

[0032] The verification environment feature data and the verification wireless signal data are used as verification samples.

[0033] The verification environment feature data and the verification wireless signal data are input into the location prediction model of the positioning device to be verified to obtain the location of the indoor positioning device output by the location prediction model of the positioning device to be verified.

[0034] Calculate the loss function of the positioning device location prediction model based on the location of the positioning device in the training room and the location of the positioning device in the verification room.

[0035] The model parameters of the positioning device location prediction model to be verified are adjusted using the loss function of the positioning device location prediction model. The steps of using the training environment feature data and the training wireless signal data as inputs to the positioning device location prediction model, using the location of the positioning device in the training room as the output of the positioning device location prediction model, and using the training samples to train the positioning device location prediction model are repeated until a preset stopping condition is reached, thus obtaining a trained positioning device location prediction model.

[0036] Optionally, the positioning device location prediction model is composed of a sequential model, which includes an input layer, several hidden layers, and an output layer. The step of inputting the environmental feature data and the wireless signal data into the trained positioning device location prediction model for processing and outputting the initial positioning device location includes:

[0037] The environmental feature data and the wireless signal data are input into the input layer for processing to obtain the first positioning device feature output by the input layer.

[0038] The first positioning device feature is input into the hidden layer for processing to obtain the second positioning device feature output by the hidden layer.

[0039] The second positioning device feature is input into the output layer for processing to obtain the initial positioning device position output by the output layer.

[0040] Optionally, the step of calculating the signal coverage strength, positioning device uniformity, and positioning device deployment cost in the indoor 3D map based on the initial positioning device location includes:

[0041] The signal radius of the positioning device is obtained based on the preset attribute information of the positioning device;

[0042] The signal coverage intensity in the indoor 3D map is calculated based on the initial positioning device location and the signal radius of the positioning device.

[0043] The uniformity of the positioning devices and the deployment cost of the positioning devices are calculated based on the initial positioning device locations.

[0044] Optionally, adjusting the initial positioning device position based on the positioning device adaptability to obtain the target positioning device position includes:

[0045] The initial positioning device position is subjected to random perturbation processing a preset number of times to obtain several sets of perturbed positioning device positions;

[0046] Based on the location of each group of disturbance positioning devices, calculate the disturbance signal coverage intensity, disturbance positioning device uniformity, and disturbance positioning device deployment cost in the indoor 3D map;

[0047] The perturbation positioning device fitness rate corresponding to each group of perturbation positioning device locations is calculated based on the perturbation signal coverage intensity, the perturbation positioning device uniformity, and the perturbation positioning device deployment cost.

[0048] The fitness of the disturbance positioning devices corresponding to several sets of disturbance positioning device locations is compared, and the set of disturbance positioning device locations corresponding to the largest disturbance positioning device fitness is taken as the target positioning device location.

[0049] In a second aspect of this application, a positioning device location planning apparatus is also provided, characterized in that the apparatus comprises:

[0050] The data acquisition module is used to acquire indoor 3D maps and wireless signal data;

[0051] An environmental feature data extraction module is used to obtain environmental feature data based on the indoor 3D map.

[0052] The initial positioning device location acquisition module is used to input the environmental feature data and the wireless signal data into the trained positioning device location prediction model for processing, and output the initial positioning device location.

[0053] The optimization parameter calculation module is used to calculate the signal coverage strength, positioning device uniformity, and positioning device deployment cost in the indoor 3D map based on the initial positioning device location.

[0054] The fitness calculation module is used to calculate the fitness of the positioning device based on the signal coverage strength, the uniformity of the positioning device, and the deployment cost of the positioning device.

[0055] The target positioning device location acquisition module is used to adjust the initial positioning device position according to the positioning device adaptability to obtain the target positioning device position.

[0056] In a third aspect of this application, an electronic device is also provided, characterized in that it includes a processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein the computer program, when executed by the processor, implements the positioning device location planning method as described above.

[0057] In a fourth aspect of this application, a computer-readable storage medium is also provided, characterized in that a computer program is stored on the computer-readable storage medium, and the computer program, when executed by a processor, implements the positioning device location planning method as described above.

[0058] The embodiments of this application have the following advantages:

[0059] In this embodiment, firstly, indoor 3D maps and wireless signal data are acquired. Environmental feature data is then obtained from the indoor 3D map, providing more accurate and reliable map data support for the planning of positioning device locations. This method not only reduces reliance on manual surveys but also improves the rationality of location planning, effectively avoiding problems such as incomplete signal coverage and multipath effects. Secondly, the environmental feature data and wireless signal data are input into a trained positioning device location prediction model for processing, outputting the initial positioning device location. This enables automatic planning of positioning device locations within the current indoor 3D map. Through accurate 3D map data and advanced algorithms, automatic planning can more precisely determine the optimal installation location of the positioning device, effectively avoiding errors and deviations that may occur in manual planning and improving the deployment efficiency of the positioning device. Finally, based on the initial positioning device location, the signal coverage strength, positioning device uniformity, and positioning device deployment cost in the indoor 3D map are calculated. The positioning device fitness is then calculated based on these factors, and the initial positioning device location is adjusted according to the positioning device fitness to obtain the target positioning device location. Deep learning network technology can be used to verify and optimize the signal coverage of positioning networks in indoor environments. This not only ensures the comprehensiveness and uniformity of signal coverage but also takes into account the economic efficiency of deployment costs, thereby achieving optimal resource allocation and ensuring the reliability and stability of the positioning network. Attached Figure Description

[0060] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.

[0061] Figure 1 This is a flowchart illustrating the steps of a positioning device location planning method according to an embodiment of this application;

[0062] Figure 2 This is a schematic diagram of the structure of a positioning device location planning apparatus provided in an embodiment of this application. Detailed Implementation

[0063] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the various embodiments of the present invention will be described in detail below with reference to the accompanying drawings. However, those skilled in the art will understand that many technical details are presented in the various embodiments of the present invention to facilitate a better understanding of the invention. However, the technical solutions claimed in the present invention can be implemented even without these technical details and with various changes and modifications based on the following embodiments. The division of the various embodiments below is for ease of description and should not constitute any limitation on the specific implementation of the present invention. The various embodiments can be combined with and referenced by each other without contradiction.

[0064] Reference Figure 1 The diagram shows a flowchart of the steps of a positioning device location planning method provided in an embodiment of this application.

[0065] The method may specifically include the following steps:

[0066] Step 101: Obtain indoor 3D map and wireless signal data.

[0067] It should be noted that an indoor 3D map can be a digital map that details the structure and layout of an indoor space. It accurately reproduces the physical environment and spatial characteristics of the interior through 3D modeling technology. The creation of indoor 3D maps typically involves using technologies such as laser scanning, photogrammetry, or computer-aided design (CAD) to obtain high-precision spatial data. The accuracy and detail of indoor 3D maps are crucial for improving the performance of indoor positioning systems. They provide necessary data support for automatically planning and optimizing the location of positioning devices, ensuring the accuracy and reliability of positioning services. In this embodiment, an indoor 3D map can be obtained using SLAM (Simultaneous Localization and Mapping) technology.

[0068] Among them, SLAM (Simultaneous Localization and Mapping) technology allows robots to simultaneously build a map of their environment and determine their own location while moving in unknown environments. SLAM technology has wide applications in fields such as robot navigation, autonomous vehicles, augmented reality (AR), and virtual reality (VR).

[0069] The basic principles of SLAM include the following key steps:

[0070] Sensor data collection: Environmental data is collected using sensors such as laser scanners, cameras, ultrasonic sensors, and inertial measurement units (IMUs).

[0071] Data association: Matching newly collected sensor data with known map data to determine the relationships between data points.

[0072] Position estimation: Using sensor data and data correlation results, the robot's current position is estimated through algorithms (such as Kalman filtering, particle filtering, etc.).

[0073] Map Update: Update and improve the environment map based on robot position estimates and sensor data.

[0074] In this embodiment of the application, the indoor 3D map obtained by SLAM technology can be a 3D map of the indoor environment constructed when a vehicle moves in an unknown environment.

[0075] It should be noted that wireless signal data can be the fundamental information needed to achieve accurate location estimation. Wireless signal data is crucial in indoor positioning systems; the following is a detailed explanation of these characteristics and how they are acquired:

[0076] Frequency: The frequency of a wireless signal refers to the number of times the signal oscillates per second, measured in Hertz (Hz). Commonly used wireless communication technologies for indoor positioning, such as Zigbee, Bluetooth, and Wi-Fi, typically operate in the 2.4 GHz band.

[0077] Wavelength: Wavelength is the distance a signal travels in one complete cycle, and it is inversely proportional to frequency. Wavelength can be calculated using the formula λ = c / f, where λ is the wavelength, c is the speed of light, and f is the frequency.

[0078] Transmit power: This refers to the energy intensity of the signal transmitted by a wireless device, usually measured in milliwatts (mW) or decibel-milliwatts (dBm). Transmit power is determined by the hardware specifications of the wireless device and can be set and adjusted by consulting the device's technical manual or through the software interface.

[0079] The methods for collecting this wireless signal characteristic data typically include:

[0080] Hardware query: Directly read parameters such as frequency, wavelength, and transmit power through the device's hardware interface.

[0081] Software configuration: Use the software tools or APIs provided by the device to obtain and set these parameters.

[0082] Signal analysis: Using specialized signal analysis tools or software to measure and record the characteristics of wireless signals.

[0083] By precisely controlling and optimizing the transmission power and frequency of wireless signals, multipath effects and signal interference can be reduced, thereby improving the performance of the positioning system.

[0084] In this embodiment of the application, wireless signal data can be used as input to the location prediction model of the positioning device to obtain the location of the positioning device.

[0085] In this embodiment of the application, indoor 3D maps and wireless signal data can be acquired to facilitate subsequent location planning for positioning devices.

[0086] Step 102: Obtain environmental feature data based on the indoor 3D map.

[0087] It should be noted that environmental feature data is a key parameter in indoor 3D maps, including but not limited to the size and shape features of the indoor space and the distribution of obstacles. Its specific meaning can be explained as follows:

[0088] Indoor space dimensions include the length, width, and height. These basic dimensions determine the size of the space and directly affect the propagation distance and coverage of wireless signals. Larger spaces may require more positioning equipment to ensure comprehensive signal coverage. Indoor space dimensions can be obtained through architectural drawings, laser scanning, or direct measurement using measuring tools such as a tape measure.

[0089] Shape features can be the specific shape of an interior space. If the shape of an interior space is complex or irregular, special feature description methods may be needed to quantify its complexity. For example, the ratio of the space's perimeter to its area can be used as a simple shape feature indicator. The shape features of an interior space can be accurately captured using 3D modeling software or laser scanning technology.

[0090] Obstacle distribution can include the type, number, and size of obstacles, such as walls, furniture, and equipment. Obstacles can affect the propagation path of wireless signals, causing signal blockage, reflection, or diffraction. Identifying and quantifying the characteristics of these obstacles is crucial for optimizing the placement of positioning devices. Obstacle distribution can be identified and recorded through indoor 3D maps, sensor data, or manual surveys.

[0091] In the design and deployment of indoor positioning systems, comprehensively considering these environmental factors can help optimize the layout of positioning devices, improve the uniformity of signal coverage, and enhance positioning accuracy. For example, in irregularly shaped indoor spaces, a denser layout of positioning devices may be needed to compensate for uneven signal propagation; while in environments with many obstacles, the positions of devices may need to be adjusted or the number of devices increased to overcome the effects of signal obstruction. Accurate environmental characteristic data can enable more intelligent and efficient indoor positioning solutions.

[0092] In this embodiment of the application, environmental feature data can be used as input to the location prediction model of the positioning device to obtain the location of the positioning device.

[0093] In this embodiment of the application, environmental feature data can be obtained based on an indoor 3D map.

[0094] This application uses indoor 3D maps to obtain environmental feature data, providing more accurate and reliable map data support for the planning of positioning device locations. This method not only reduces reliance on manual surveys but also improves the rationality of location planning, effectively avoiding problems such as incomplete signal coverage and multipath effects.

[0095] Step 103: Input the environmental feature data and the wireless signal data into the trained positioning device location prediction model for processing, and output the initial positioning device location.

[0096] It should be noted that the positioning device location prediction model can be an algorithmic model used to determine the optimal installation location of indoor positioning devices. This model can be based on machine learning algorithms, utilizing indoor environmental feature data and wireless signal propagation characteristics to predict the layout of positioning devices.

[0097] In this embodiment, environmental feature data and wireless signal data can be input into a trained positioning device location prediction model for processing, and the initial positioning device location can be output. The initial positioning device location can be several three-dimensional coordinates on an indoor three-dimensional map, representing the initial placement position of several positioning devices.

[0098] This application processes environmental feature data and wireless signal data into a trained positioning device location prediction model to output the initial positioning device location. It can automatically plan the positioning device location in the current indoor 3D map. Through accurate 3D map data and advanced algorithms, automatic planning can more accurately determine the best installation location of the positioning device, effectively avoiding errors and deviations that may occur in manual planning and improving the deployment efficiency of the positioning device.

[0099] Step 104: Calculate the signal coverage strength, positioning device uniformity, and positioning device deployment cost in the indoor 3D map based on the initial positioning device location.

[0100] It should be noted that in indoor 3D maps, signal coverage strength, uniformity of positioning devices, and deployment cost of positioning devices are key indicators for evaluating and optimizing the performance of indoor positioning systems. The following is a detailed explanation of these indicators:

[0101] Signal coverage strength refers to the strength and uniformity of wireless signals within an indoor space. Ideal signal coverage should provide a stable and sufficiently strong signal throughout the entire indoor space to ensure the accuracy and reliability of location services. The signal strength at each location can be determined through simulation or actual measurement, and the signal coverage uniformity throughout the space can be calculated.

[0102] The uniformity of positioning devices refers to the evenness of their distribution within an indoor space. A uniform distribution ensures comprehensive signal coverage and avoids signal blind spots. This uniformity can be assessed by calculating the distance between devices or their density.

[0103] The deployment cost of positioning equipment can include equipment purchase cost, installation cost, and maintenance cost. When optimizing the layout of positioning equipment, cost-effectiveness needs to be considered to ensure that performance requirements are met while minimizing overall cost. Deployment costs can be assessed by calculating the number of devices, installation difficulty, and expected maintenance costs.

[0104] In this embodiment of the application, the signal coverage strength in the indoor 3D map, the uniformity of the positioning devices in the indoor 3D map, and the deployment cost of the positioning devices in the indoor 3D map can be calculated based on the initial positioning device location.

[0105] Step 105: Calculate the adaptability of the positioning device based on the signal coverage strength, the uniformity of the positioning device, and the deployment cost of the positioning device.

[0106] It should be noted that in indoor positioning systems, positioning device adaptability can be a comprehensive evaluation index, reflecting the overall performance of the positioning device layout in terms of signal coverage strength, device uniformity, and deployment cost. A higher positioning device adaptability score indicates better overall performance of the layout scheme in terms of signal coverage, device uniformity, and cost-effectiveness.

[0107] In this embodiment of the application, the adaptability of the positioning device can be calculated based on the signal coverage strength, uniformity of positioning devices, and deployment cost of positioning devices in the indoor 3D map.

[0108] Step 106: Adjust the initial positioning device position according to the positioning device adaptability to obtain the target positioning device position.

[0109] It should be noted that the target positioning device location can be an optimized layout of positioning devices within an indoor 3D map. This layout achieves optimal signal coverage, uniform device distribution, and cost-effectiveness. In this embodiment, the target positioning device location can be the optimal layout of positioning devices within an indoor 3D map.

[0110] In this embodiment, the initial positioning device position can be adjusted based on the positioning device adaptability to obtain the target positioning device position. The adjustment of the initial positioning device position based on the positioning device adaptability can be achieved through iterative optimization. The target positioning device position can be a set of two-dimensional coordinates on an indoor 3D map, representing the optimal placement of several positioning devices. The installation height of the positioning devices can be selected based on the actual installation environment.

[0111] In this embodiment, firstly, indoor 3D maps and wireless signal data are acquired. Environmental feature data is then obtained from the indoor 3D map, providing more accurate and reliable map data support for the planning of positioning device locations. This method not only reduces reliance on manual surveys but also improves the rationality of location planning, effectively avoiding problems such as incomplete signal coverage and multipath effects. Secondly, the environmental feature data and wireless signal data are input into a trained positioning device location prediction model for processing, outputting the initial positioning device location. This enables automatic planning of positioning device locations within the current indoor 3D map. Through accurate 3D map data and advanced algorithms, automatic planning can more precisely determine the optimal installation location of the positioning device, effectively avoiding errors and deviations that may occur in manual planning and improving the deployment efficiency of the positioning device. Finally, based on the initial positioning device location, the signal coverage strength, positioning device uniformity, and positioning device deployment cost in the indoor 3D map are calculated. The positioning device fitness is then calculated based on these factors, and the initial positioning device location is adjusted according to the positioning device fitness to obtain the target positioning device location. Deep learning network technology can be used to verify and optimize the signal coverage of positioning networks in indoor environments. This not only ensures the comprehensiveness and uniformity of signal coverage but also takes into account the economic efficiency of deployment costs, thereby achieving optimal resource allocation and ensuring the reliability and stability of the positioning network.

[0112] In one optional embodiment of this application, the indoor 3D map is acquired by a vehicle, which includes a lidar and a camera. Step 101 further includes the following sub-steps:

[0113] S111: Obtain the preset driving route;

[0114] S112: During the driving process according to the driving route, three-dimensional point cloud data is collected by the lidar and two-dimensional image data is collected by the camera;

[0115] S113: Synthesize an initial three-dimensional map based on the three-dimensional point cloud data and the two-dimensional image data;

[0116] S114: Perform feature extraction on the three-dimensional point cloud data to obtain three-dimensional feature data;

[0117] S115: Process the initial three-dimensional map according to the three-dimensional feature data to obtain the indoor three-dimensional map;

[0118] S116: Obtain the wireless signal data according to the preset attribute information of the positioning device.

[0119] It should be noted that the vehicle can be a vehicle that includes a computing unit, which can process the data collected by sensors installed on the vehicle. The data processing can be performed on the vehicle or on a cloud server. In specific implementations, the computing unit can be a computer or other devices.

[0120] It should be noted that the driving route can be a preset route for the vehicle in the current indoor environment. The current indoor environment includes, but is not limited to, an underground parking lot. In this embodiment, the vehicle can travel according to the driving route, and the vehicle can travel according to the driving route in autonomous driving mode.

[0121] It's worth noting that LiDAR and cameras are two commonly used sensor technologies with wide applications in autonomous driving, robot navigation, augmented reality (AR), virtual reality (VR), and indoor positioning. The following is a detailed introduction to these two technologies:

[0122] LiDAR (Light Detection and Ranging) determines the distance to a target by emitting laser pulses and measuring their return time. It can generate high-precision 3D environmental maps, including the shape, size, and location of objects. LiDAR offers centimeter-level distance measurement accuracy, can acquire large amounts of data in a short time for rapid environmental awareness, and operates stably under various lighting conditions.

[0123] Cameras generate images by capturing light and converting it into electrical signals. Cameras can be equipped with image processing chips, enabling real-time image analysis and recognition.

[0124] In the embodiments of this application, lidar can provide accurate distance and three-dimensional structural information, while cameras can provide rich visual details and color information. The combination of the two can achieve more comprehensive and reliable environmental perception.

[0125] It should be noted that 3D point cloud data and 2D image data are two different types of data, and they have wide applications in fields such as computer vision, robotics, and geographic information systems (GIS).

[0126] 3D point cloud data can be a dataset composed of a large number of 3D points, each of which typically contains its coordinates (X, Y, Z) in 3D space as well as possible other attributes (such as color, intensity, etc.). It can be acquired through LiDAR, structured light scanning, stereo vision systems, or other 3D scanning technologies.

[0127] Two-dimensional image data can be a two-dimensional array of pixels, where each pixel contains its position (X, Y) in the image and color information (such as RGB values). It can be acquired through a camera, scanner, or other image acquisition device. In this embodiment, an initial three-dimensional map can be generated using three-dimensional point cloud data and two-dimensional image data.

[0128] It should be noted that the initial 3D map can be generated by fusing 3D point cloud data and 2D image data from different sensors using perceptual fusion technology to produce a high-precision indoor 3D map. Specifically, fusing data from different sensors using perceptual fusion technology can refer to fusing multimodal data at the feature level in the LiDAR branch, but fusing at the dataset and feature level in the image branch. In this embodiment, the initial 3D map is the initial indoor 3D map obtained through data fusion.

[0129] It should be noted that the three-dimensional feature data can be the three-dimensional feature data in an indoor three-dimensional map. Taking an indoor three-dimensional map as an example, the three-dimensional feature data can be corners, columns, garage floor signs, etc. This application does not limit which three-dimensional features are specifically included in the three-dimensional feature data.

[0130] In this embodiment of the application, the indoor 3D map can be a more accurate indoor 3D map obtained by post-processing the initial 3D map. Post-processing includes, but is not limited to, denoising and feature extraction.

[0131] It should be noted that the preset attribute information of the positioning device can be the inherent attributes of the positioning device to be deployed. Below are some common preset attribute information for positioning devices:

[0132] The device type can be the type of positioning device, such as Wi-Fi access point, Bluetooth beacon, ultra-wideband (UWB) tag, RFID reader, etc.

[0133] The communication protocol can be the one used by the positioning device, such as IEEE 802.11 (Wi-Fi), Bluetooth, Zigbee, LoRa, etc.

[0134] Transmission power can be the power of the signal transmitted by the positioning device, usually measured in milliwatts (mW) or decibel milliwatts (dBm).

[0135] The frequency range can be the frequency range in which the positioning device operates, such as 2.4GHz, 5GHz, etc.

[0136] Signal coverage can be the physical space range that a device can effectively cover, which is usually related to transmission power and environmental factors.

[0137] Before deploying positioning equipment, understanding and recording these pre-defined attributes is crucial, as they directly impact the equipment's installation, configuration, and performance. For example, transmit power and frequency range affect signal propagation and coverage. By comprehensively considering these attributes, the layout of positioning equipment can be planned and optimized more effectively, ensuring the system provides accurate and reliable positioning services.

[0138] In this embodiment, the vehicle can obtain wireless signal data based on preset attribute information of the positioning device. The wireless signal data may include basic parameters such as the frequency, wavelength, and transmission power of the wireless signal.

[0139] In this embodiment, the vehicle can acquire a preset driving route. While driving along the route, it collects 3D point cloud data using a LiDAR and 2D image data using a camera. Then, it synthesizes an initial 3D map based on the 3D point cloud data and the 2D image data. Feature extraction is performed on the 3D point cloud data to obtain 3D feature data. This 3D feature data is then used to process the initial 3D map to obtain an indoor 3D map. Furthermore, wireless signal data can be acquired based on preset attribute information from the positioning device.

[0140] In practical implementation, in an indoor environment used for model training, an autonomous vehicle starts from a preset starting point and prepares to begin the scanning process. Using sensors onboard the vehicle, such as LiDAR and cameras, it automatically collects indoor environmental data. Then, using perception fusion technology, the data collected by different sensors is fused to generate a high-precision indoor 3D map, ensuring the map's accuracy and completeness. Finally, the generated map undergoes post-processing, including noise reduction, feature extraction (such as corners, pillars, garage floor signs, etc.), and map optimization, to improve the map's usability and accuracy.

[0141] In one optional embodiment of this application, step S114 further includes the following sub-steps:

[0142] S1141: Obtain voxel data and original point location data based on the three-dimensional point cloud data;

[0143] S1142: Perform feature extraction on the voxel data to obtain candidate region features;

[0144] S1143: Perform feature extraction on the original point data to obtain key point features;

[0145] S1144: Determine the three-dimensional feature data based on the candidate region features and the key point features.

[0146] It should be noted that voxel data can be a data representation in three-dimensional space, similar to pixels in a two-dimensional image, but extended to three-dimensional space. A voxel is an abbreviation for "Volumetric Pixel," representing a basic unit in three-dimensional space. Each voxel typically contains its position information (X, Y, Z coordinates) in three-dimensional space, as well as possible other attributes (such as color, density, transparency, etc.). In the embodiments of this application, voxel data can be used to obtain candidate region features.

[0147] Raw point cloud data refers to unprocessed 3D point cloud data directly acquired through 3D scanning technologies (such as LiDAR, structured light scanning, stereo vision, etc.). This data contains surface information of the scanned object or environment, and each point typically includes its coordinates (X, Y, Z) in 3D space, as well as possible other attributes (such as color, reflectance intensity, timestamp, etc.). In this embodiment, voxel data can be used to obtain key point features.

[0148] Region proposal features refer to a series of features used in computer vision and image processing to identify and extract regions in an image that may contain objects of interest (such as people, animals, vehicles, etc.). These features are typically used to generate region proposals, which are potential regions in an image that may contain target objects. In this embodiment, the region proposal features may be potential regions containing target objects such as corners, pillars, and garage floor signs.

[0149] Keypoint features, in image processing and computer vision, refer to a set of attributes used to describe salient and stable local features in an image. These keypoints are typically points in an image that are not easily affected by changes in lighting, scale, rotation, and viewpoint, such as corner points, edge points, or textured areas. In this embodiment, keypoint features can be points in an indoor 3D map that are not easily affected by changes in lighting, scale, rotation, and viewpoint.

[0150] In this embodiment of the application, voxel data and original point data can be obtained from 3D point cloud data, then feature extraction can be performed on the voxel data to obtain candidate region features, feature extraction can be performed on the original point data to obtain key point features, and finally 3D feature data can be determined based on the candidate region features and key point features.

[0151] In practical implementation, a 3D target detection algorithm model based on fusion perception 3D point cloud data can be used to extract features from the 3D point cloud data.

[0152] Point cloud data is processed by both a global feature extraction model and a local feature extraction model. In the global feature extraction model, the 3D point cloud data is voxelized. The resulting voxel data is then subjected to downsampling feature extraction followed by upsampling feature extraction to obtain candidate region features.

[0153] In the local feature extraction model, raw point data is obtained from 3D point cloud data. Keypoint sampling is performed on the raw point data to obtain keypoint sampling results, which are then input into a point cloud learning network (multilayer perceptron) to obtain keypoint features.

[0154] The candidate region features and key point features are input into the detection model for feature fusion to generate candidate regions containing the target object. Then, the target object is classified, and the candidate bounding box is used for regression processing to obtain the position of the target object in the 3D map. The detected target object and its position are output.

[0155] Global feature extraction models, in image processing and computer vision, refer to a class of models used to extract global features from the entire image. These models typically focus on the overall attributes of the image, such as color distribution, texture structure, and shape information, rather than local details. In practical implementations, the entire 3D point cloud data can be processed.

[0156] Local feature extraction models refer to a class of models used in image processing and computer vision to extract features from local regions of an image. These models focus on small regions or local structures in an image, such as corners, edges, and textures. In specific implementations, keypoints in 3D point cloud data can be processed.

[0157] Detection models, in computer vision and image processing, are used to identify and locate specific target objects in images or videos. These models typically need to perform two tasks simultaneously: target classification (determining whether an object of a specific category exists in an image) and target localization (determining the object's position and size within the image). In practical implementations, they can be used to detect target objects such as corners, pillars, and garage floor signs in indoor 3D maps.

[0158] In one optional embodiment of this application, the location prediction model of the positioning device is trained in the following manner:

[0159] S201: Obtain a standard indoor 3D map, wherein the standard indoor 3D map includes a number of standard positioning devices arranged at the locations of the standard indoor positioning devices;

[0160] S202: Obtain standard environmental feature data based on the standard indoor 3D map;

[0161] S203: Obtain standard wireless signal data based on the preset attribute information of the standard positioning device;

[0162] S204: Preprocess at least one of the standard environmental feature data, the standard wireless signal data, and the standard indoor positioning device location, wherein the preprocessing is one or more of the following: data cleaning, missing data imputation, normalization, and classification coding.

[0163] S205: Based on the preprocessed standard environmental feature data, standard wireless signal data, and standard indoor positioning device location, obtain training environmental feature data, training wireless signal data, training indoor positioning device location, and verification environmental feature data and verification wireless signal data;

[0164] S206: Use the training environment feature data, the training wireless signal data, and the location of the training indoor positioning device as training samples;

[0165] S207: The training environment feature data and the training wireless signal data are used as inputs to the positioning device location prediction model, the location of the positioning device in the training room is used as the output of the positioning device location prediction model, and the positioning device location prediction model is trained using the training samples to obtain the positioning device location prediction model to be verified.

[0166] S208: Use the verification environment feature data and the verification wireless signal data as verification samples;

[0167] S209: Input the verification environment feature data and the verification wireless signal data into the location prediction model of the positioning device to be verified to obtain the location of the indoor positioning device output by the location prediction model of the positioning device to be verified.

[0168] S210: Calculate the loss function of the positioning device position prediction model based on the location of the positioning device in the training room and the location of the positioning device in the verification room;

[0169] S211: The model parameters of the positioning device location prediction model to be verified are adjusted using the loss function of the positioning device location prediction model, and the steps of using the training environment feature data and the training wireless signal data as inputs to the positioning device location prediction model, using the location of the positioning device in the training room as the output of the positioning device location prediction model, and using the training samples to train the positioning device location prediction model are repeated until the preset stopping condition is reached, so as to obtain the trained positioning device location prediction model.

[0170] It should be noted that a standard indoor 3D map can be a number of optimal indoor positioning network environments that have been pre-configured and debugged by wireless positioning experts in multiple indoor environments used for model training. The standard indoor 3D map can be used to provide training and testing data required for subsequent machine learning.

[0171] In this embodiment, the vehicle can acquire a standard indoor 3D map, which includes several standard positioning devices deployed at the locations of standard indoor positioning devices. The acquisition of the standard indoor 3D map can be achieved by the vehicle traveling along a preset route in an optimal indoor positioning network environment. During this route travel, 3D point cloud data is collected using LiDAR, and 2D image data is collected using a camera. An initial 3D map is then synthesized from the 3D point cloud data and the 2D image data. Feature extraction is performed on the 3D point cloud data to obtain 3D feature data. This 3D feature data is then used to process the initial 3D map to obtain the indoor 3D map.

[0172] In this embodiment, standard environmental feature data can be obtained based on a standard indoor 3D map. Standard environmental feature data includes, but is not limited to, the size, shape features, and obstacle distribution of the indoor space within an optimal indoor positioning network environment.

[0173] In this embodiment, standard wireless signal data can be obtained based on preset attribute information of a standard positioning device. The standard wireless signal data can include basic parameters such as frequency, wavelength, and transmission power of the wireless signal in an optimal indoor positioning network environment.

[0174] In this embodiment, information such as the location of standard indoor positioning devices can be obtained based on a standard indoor 3D map. Specifically, this can be data on the deployment of wireless positioning points, including the number of positioning devices, the location of standard indoor positioning devices, their unique identifiers, and signal strength. The specific data collection method is as follows: Using a high-precision indoor map collected by an autonomous vehicle, after coordinate transformation, the deployed positioning devices can be identified, thereby obtaining the coordinate information of the positioning device locations. This coordinate information is then bound and mapped to pre-organized positioning device coding information. Simultaneously, a wireless signal acquisition terminal follows the autonomous vehicle, collecting wireless signals at various indoor locations. The data collected by the wireless acquisition terminal includes the unique identifier of the positioning device and the signal strength from the device to the acquisition terminal's location. Since the position of the acquisition terminal is relative to and fixed with the position of the autonomous vehicle, the relative indoor position of the acquisition terminal at different collection times can be obtained subsequently through coordinate transformation.

[0175] In this context, a data acquisition terminal refers to a device or system used to collect and transmit data. These terminals are typically located at the source of the data and are responsible for collecting raw data and sending it to a central processing system or cloud server via a network. Data acquisition terminals can be various types of devices, including sensors, data loggers, mobile devices, and industrial control systems, depending on the application scenario and data type. In this embodiment, the data acquisition terminal can collect the signal strength from the positioning device to its location.

[0176] It's important to note that data cleaning, missing data imputation, normalization, and classification coding are several key steps in data preprocessing, crucial for improving data quality and ensuring the performance of machine learning models. The following is a detailed description of each step:

[0177] Data cleaning refers to the process of identifying and correcting errors and inconsistencies in a dataset. This includes removing duplicate records, correcting erroneous values, and handling outliers. The purpose of data cleaning is to ensure the quality of the data, making it more suitable for analysis and modeling.

[0178] Missing data imputation refers to the process of handling missing values ​​in a dataset. Missing values ​​may be due to data collection errors, equipment malfunctions, or other reasons. Methods for handling missing values ​​include: deleting records containing missing values: if there are only a few missing values, these records can be deleted directly. Mean / median imputation: imputing missing values ​​with the mean or median of the variable. Mode imputation: imputing missing values ​​with the mode (the most frequent value) of the variable. Interpolation: estimating missing values ​​using linear interpolation or other interpolation methods. Machine learning methods: predicting missing values ​​using regression models or other machine learning algorithms.

[0179] Normalization refers to scaling data to a common range so that values ​​of different features are comparable. This is crucial for many machine learning algorithms, especially distance-based algorithms such as K-nearest neighbors. Common normalization methods include: min-max normalization: scaling data to the range [0,1] or [-1,1]; and Z-score standardization: transforming data into a distribution with a mean of 0 and a standard deviation of 1.

[0180] Categorical encoding refers to converting categorical variables into numerical forms so that machine learning algorithms can process them. Categorical variables are typically in text or label form, such as color, gender, country, etc. Common categorical encoding methods include:

[0181] One-hot encoding: Creates a binary feature for each category, where only one feature is 1 and the rest are 0.

[0182] Label Encoding: Assign a unique integer value to each category.

[0183] Ordinal Encoding: Assigning integer values ​​to ordered categorical variables while maintaining the order of the categories.

[0184] Binary encoding: Converts categories into binary representations, reducing the number of features.

[0185] These preprocessing steps help improve data quality and consistency, thereby enhancing the performance and reliability of machine learning models. In practical applications, the appropriate preprocessing method can be determined based on the characteristics of the data and the analysis objectives.

[0186] In the embodiments of this application, at least one of standard environmental feature data, standard wireless signal data, and standard indoor positioning device location can be preprocessed. The preprocessing includes one or more of data cleaning, missing data imputation, normalization, and classification coding.

[0187] In practical implementation, data cleaning can be understood as follows: if the signal strength at a beacon location is significantly higher or lower than the average signal strength of that beacon at the current location when collected multiple times, and its absolute Z-Score is greater than a certain threshold (e.g., 3), then the signal strength data at that beacon location may be considered an outlier. These outliers may be due to measurement errors or antenna directivity. The Z-Score method can be used to identify outliers and remove them from the dataset. Z-score is an outlier detection method based on a standard score. The Z-score represents the distance of a data point from the mean. It is obtained by calculating the difference between each data point and the mean and dividing it by the standard deviation. If the Z-score is greater than a certain threshold (e.g., 3), the data point is considered an outlier. The advantage of this method is that it does not depend on the distribution shape of the data, but only considers the relative distance between the data point and the mean.

[0188] Data cleaning can be performed using Python's Pandas library to process a DataFrame containing RSSI (Received Signal Strength Indicator) data. First, a sample dataset is created containing some NaN values, representing missing values ​​in the data. Next, the mean of the RSSI column is calculated, and this mean is used to fill in all the NaN values ​​in the DataFrame. This is done to ensure that the dataset is not affected by missing values ​​in subsequent analysis or machine learning model training. Finally, the padded DataFrame will no longer contain any NaN values; all missing data points have been replaced with the mean of the column.

[0189] In the specific implementation, the data imputation process can be manifested as follows: for missing training data, imputation can be selected; for numerical features, mean imputation can be used in this application. Data imputation can use Python's NumPy library to process a NumPy array containing RSSI (Received Signal Strength Indicator) signal strength. First, an array named `data` is defined, containing the actual acquired signal quality data. Next, the mean (mean_value) and standard deviation (std_dev) of this array are calculated. Then, the Z-score for each data point is calculated, which is obtained by subtracting the mean from each data point and then dividing by the standard deviation. The Z-score is a standardization method that helps identify outliers in the data. Finally, all data points with a Z-score greater than 3 are identified; these data points are considered outliers and stored in an array named `outliers`. In this way, extreme values ​​that may be caused by measurement errors or other reasons can be identified and handled.

[0190] In practical implementation, normalization is beneficial because sensor data may have different dimensions and numerical ranges. Scaling all features to a uniform range (typically 0 to 1, or -1 to 1) is helpful. Normalization helps the model converge faster and avoids some features having a disproportionate impact on model training due to their large numerical range. Normalization can be performed using Python's Pandas and scikit-learn libraries to process DataFrames containing RSSI (Received Signal Strength Indicator) data. First, create a sample dataset containing a column named 'RSSI'. Then, import the MinMaxScaler class from the scikit-learn library for data normalization. Create a MinMaxScaler object to normalize the 'RSSI' column in the DataFrame, scaling the data to a range of 0 to 1. The normalized data is stored in a new column 'Normalized_RSSI'. Finally, print the normalized DataFrame to view the results. This operation ensures that data with different ranges have equal weight in the machine learning model, improving model performance.

[0191] In practical implementation, categorical encoding can encode categorical data. If the dataset contains categorical data (such as the type of indoor space or the material of obstacles), it needs to be converted into numerical data. This application can use one-hot encoding to convert each category into a new binary column, where only the column corresponding to the category is 1, and the rest are 0.

[0192] In practical implementation, features useful for optimizing the indoor positioning network can be extracted and constructed based on a standard indoor 3D map or several standard positioning devices deployed at the locations of standard indoor positioning devices. These features may include, but are not limited to:

[0193] Standard environmental characteristic data, such as:

[0194] Indoor space dimensions: Extract the length, width, and height of the indoor environment, as these dimensions can directly affect signal propagation and coverage.

[0195] Shape features: If the shape of an interior space is irregular, features can be constructed to describe the complexity of its shape, such as using the ratio of the space's perimeter to its area.

[0196] Obstacle distribution: Identify and quantify the type, number, and size of obstacles, as these characteristics have a significant impact on signal propagation obstruction and reflection.

[0197] Standard wireless signal data, for example:

[0198] Signal frequency: The frequency of a signal affects its wavelength and penetration ability, and can be used as one of its characteristics.

[0199] Signal wavelength: Wavelength is frequency-dependent and can be used to evaluate the propagation characteristics of signals in different materials.

[0200] Signal power: The power of signal transmission directly affects the signal coverage and strength.

[0201] Standard indoor 3D maps, for example:

[0202] Map feature extraction: Features are extracted from the high-precision indoor 3D map generated by SLAM technology, including corners, columns, floor signs, lights, desks and chairs, etc. These features can be used as reference points for signal propagation.

[0203] Standard indoor positioning device location, specific location coordinates of the indoor positioning device, and beacon spacing: construct the distance characteristics between beacons, as the spacing affects signal overlap and coverage uniformity.

[0204] Simulation results of signal propagation models, for example:

[0205] Signal coverage: Based on the signal propagation model, the coverage area of ​​each positioning device is extracted.

[0206] Signal strength distribution: By analyzing the signal strength at different locations, we can construct the characteristics of the signal strength distribution.

[0207] Signal attenuation: Extract the attenuation rate of the signal during propagation.

[0208] Through the feature construction steps described above, a rich and representative feature set can be generated. One or more of these features can be used as input to a machine learning model to predict or optimize the layout of indoor positioning networks. These features include not only direct measurements of the raw data but also derived features obtained through analysis and processing of the raw data, thus providing the model with a more comprehensive perspective to understand and solve the indoor positioning network optimization problem.

[0209] In practical implementation, Python's TensorFlow, TensorFlow Data Validation (TFDV), Pandas, and scikit-learn libraries can be used to process PandasDataFrames containing preliminary indoor environment layout data. First, a sample DataFrame is created, containing data such as the length and width of the indoor space, signal frequency and power, and obstacle materials. Then, this Pandas DataFrame is converted into a TensorFlow dataset for further processing and analysis.

[0210] To better understand a dataset, you can use Python's TFDV library to infer its schema and visualize statistics. This is an optional step that can help you better understand the structure and distribution of the data.

[0211] In the data preprocessing step, a Python function `preprocess_data` can be defined. This function is responsible for cleaning missing values, feature normalization, and encoding categorical data. For missing values, the mean is used for imputation; for numerical features, `StandardScaler` is used for normalization; and for categorical features, one-hot encoding is used. Finally, the processed data is converted back to a TensorFlow dataset and transformed into a format suitable for model training, with a batch size of 32. Through these steps, the preprocessed dataset can be obtained.

[0212] These steps ensure that the dataset has been properly cleaned and transformed before being input into the neural network model, thereby improving the model's performance and reliability.

[0213] In this embodiment, training environment feature data, training wireless signal data, training indoor positioning device location, and verification environment feature data and verification wireless signal data can be obtained based on preprocessed standard environmental feature data, standard wireless signal data, and standard indoor positioning device location. The training environment feature data, training wireless signal data, and training indoor positioning device location are a subset of the standard environment feature data, standard wireless signal data, and standard indoor positioning device location, while the verification environment feature data and verification wireless signal data are a subset of the standard environment feature data and standard wireless signal data. This can be understood as dividing the standard dataset into training and verification sets according to a certain proportion.

[0214] In this embodiment, training environment feature data, training wireless signal data, and training indoor positioning device location can be used as training samples. The training environment feature data and training wireless signal data can be used as inputs to the positioning device location prediction model, and the training indoor positioning device location can be used as the output of the positioning device location prediction model. The positioning device location prediction model can be trained using the training samples to obtain the positioning device location prediction model to be verified.

[0215] In this embodiment, verification environmental feature data and verification wireless signal data can be used as verification samples. The verification environmental feature data and verification wireless signal data are input into the location prediction model of the positioning device to be verified to obtain the location of the indoor positioning device output by the location prediction model of the positioning device to be verified.

[0216] In this embodiment, the loss function of the positioning device location prediction model is calculated based on the location of the positioning device in the training room and the location of the positioning device in the verification room. The Adam optimizer and mean squared error can be selected as the loss function, and mean absolute error is added as the evaluation index.

[0217] It should be noted that the Adam (Adaptive Moment Estimation Optimizer) is a widely used optimization algorithm in the field of deep learning. It combines the advantages of the Momentum and RMSprop algorithms, aiming to provide a fast and stable model training convergence process. Adam adjusts the size of gradient updates by calculating an adaptive learning rate for each parameter, making it more flexible in handling gradients of different parameters. Specifically, Adam maintains the first moment (i.e., the exponentially weighted average of the gradients) and the second moment (i.e., the exponentially weighted average of the squared gradients) for each parameter and uses these moments to adjust the learning rate for each parameter. This adaptability makes Adam perform well in handling sparse gradients and non-stationary objectives, and therefore it has been widely used in various deep learning tasks. In the embodiments of this application, the Adam optimizer can be used to obtain the loss function of a positioning device location prediction model.

[0218] Mean Squared Error (MSE) is a commonly used loss function, widely applied in regression problems, to measure the difference between model predictions and actual values. MSE quantifies this difference by calculating the square of the difference between the predicted and actual values ​​and then averaging these squares. In this embodiment, MSE can be used to obtain the loss function for a positioning device location prediction model.

[0219] Mean Absolute Error (MAE) is a commonly used evaluation metric, particularly suitable for regression problems, used to measure the average absolute difference between model predictions and actual values. MAE quantifies this difference by calculating the absolute values ​​of the differences between predicted and actual values ​​and then averaging these absolute values. In this embodiment, MAE can be used to evaluate the loss function of a positioning device location prediction model.

[0220] In this embodiment, the model parameters of the positioning device location prediction model to be validated can be adjusted using the loss function of the positioning device location prediction model. The process involves re-executing the steps of using training environment feature data and training wireless signal data as input to the positioning device location prediction model, using the indoor positioning device location as the output, and training the positioning device location prediction model with training samples until a preset stopping condition is reached, thus obtaining a trained positioning device location prediction model. The preset stopping condition can be the number of training rounds, which is not limited in this application.

[0221] This application effectively acquires the layout of indoor positioning devices by obtaining standard indoor 3D maps and related data, performing preprocessing, and training the model. Based on standard environmental feature data and wireless signal data, preprocessing steps such as data cleaning, missing data imputation, normalization, and classification coding improve data quality and consistency. Subsequently, through training and validation processes, model parameters are continuously adjusted until a preset stopping condition is reached, resulting in a trained positioning device location prediction model. This model can automatically process and analyze large amounts of data, improving the deployment efficiency and accuracy of positioning devices, while model training and validation ensure the reliability and accuracy of the prediction results. Furthermore, it can adapt to different indoor environments and device configurations, exhibiting strong versatility and flexibility.

[0222] In one optional embodiment of this application, the positioning device location prediction model is composed of a sequential model, which includes an input layer, several hidden layers, and an output layer. Step 103 further includes the following sub-steps:

[0223] S301: Input the environmental feature data and the wireless signal data into the input layer for processing to obtain the first positioning device feature output by the input layer;

[0224] S302: Input the first positioning device feature into the hidden layer for processing to obtain the second positioning device feature output by the hidden layer;

[0225] S303: Input the second positioning device feature into the output layer for processing to obtain the initial positioning device position output by the output layer.

[0226] It should be noted that a sequential model is a common model structure in deep learning, especially when using frameworks such as Keras or TensorFlow. This model consists of a series of sequentially stacked network layers, each taking the output of the previous layer as input, progressively processing the data until the final output is obtained. Sequential models are widely used in various neural network architectures due to their simple structure, ease of understanding and implementation. In the embodiments of this application, a sequential model can be used to construct a positioning device location prediction model.

[0227] The input layer, which can be the first layer of a neural network, is responsible for receiving raw data or features. In sequential models, the input layer typically does not perform any computations but instead defines the shape and structure of the network input. For example, when processing image data, the input layer might define the image's width, height, and number of color channels. The main function of the input layer is to provide a clear input interface for the network, ensuring that data can be correctly passed to subsequent hidden layers. In this embodiment, the positioning device location prediction model may include an input layer.

[0228] Hidden layers are network layers located between the input and output layers. They are the core components of neural networks for complex data processing and feature extraction. Each hidden layer consists of multiple neurons (or nodes), which process the input data through activation functions and pass the results to the next layer. The number of hidden layers and the number of neurons in each layer are key parameters in designing neural networks, determining the network's depth and width, and thus affecting the model's learning ability and performance. In this embodiment, the positioning device location prediction model may include several hidden layers. The specific number of hidden layers can be set according to actual needs.

[0229] The output layer, often the last layer of a neural network, is responsible for generating the model's final output. The structure and activation function of the output layer are typically designed based on the type of task. For example, in classification tasks, the output layer might contain the same number of neurons as the number of classes and use a softmax activation function to generate a probability distribution; in regression tasks, the output layer might have only one neuron and use a linear activation function to directly output the predicted value. The output layer's result is the model's final conclusion after processing the input data and forms the basis for evaluating model performance and making decisions. In this embodiment, the location prediction model for a positioning device may include an output layer.

[0230] In this embodiment, environmental feature data and wireless signal data can be input to the input layer for processing to obtain the first positioning device feature output by the input layer. Then, the first positioning device feature is input to the hidden layer for processing to obtain the second positioning device feature output by the hidden layer. Finally, the second positioning device feature is input to the output layer for processing to obtain the initial positioning device position output by the output layer.

[0231] The first positioning device feature and the second positioning device feature can be multidimensional arrays or matrices that embody environmental feature data and wireless signal data.

[0232] In the implementation, a Sequential model can be defined using Python, with several Dense layers added as hidden layers. Each hidden layer is followed by a Dropout layer to reduce overfitting. The input layer parameters are set to the number of features, and the number of neurons in the output layer is set to the number of beacon locations predicted. The ReLU activation function can be used as the activation function for the hidden layers, and a linear activation function is chosen for the output layer to address the regression problem in this application. Next, the model is compiled, selecting the Adam optimizer and mean squared error as the loss function, and adding mean absolute error as the evaluation metric. Finally, the `model.fit` method is used to train the model.

[0233] In this embodiment, the model can be further optimized by adjusting parameters such as network structure, number of hidden layers, number of neurons, activation function, optimizer, and loss function in the positioning device location prediction model, combined with the test results of the test data.

[0234] In one optional embodiment of this application, step 104 further includes the following sub-steps:

[0235] S401: Obtain the signal radius of the positioning device according to the preset attribute information of the positioning device;

[0236] S402: Calculate the signal coverage strength in the indoor 3D map based on the initial positioning device location and the signal radius of the positioning device;

[0237] S403: Calculate the uniformity of the positioning device and the deployment cost of the positioning device based on the initial positioning device location.

[0238] The signal radius of a positioning device refers to the maximum distance over which the device can effectively cover and provide positioning services. This radius depends on various factors, including the device's signal transmission power, receiving sensitivity, environmental interference and obstacles, and the type of positioning technology used. In this embodiment, the signal radius of the positioning device can be obtained based on the device's attribute information.

[0239] Signal coverage strength refers to the level of signal quality that a wireless signal can provide within a specific area. It is commonly used to evaluate the performance of wireless communication systems (such as Wi-Fi, Bluetooth, and mobile networks), particularly in positioning and communication applications. In this embodiment, signal coverage strength can be determined by calculating the average proportion of predicted coverage exceeding a threshold.

[0240] The uniformity of positioning devices refers to the evenness of their distribution within a target area. A high uniformity layout means the devices are more evenly distributed in space, ensuring more comprehensive signal coverage, reducing blind spots, and improving the quality and reliability of positioning services. In this embodiment, the uniformity of positioning devices can be obtained by calculating the variance of their locations or their density distribution in different areas. A smaller variance indicates higher uniformity, and a more uniform density also indicates higher uniformity.

[0241] The deployment cost of positioning equipment refers to the economic costs and resource consumption required to deploy positioning equipment within a target area. Deployment costs include the cost of the equipment itself, installation costs, and maintenance costs. Optimizing deployment costs means minimizing costs while ensuring the quality of positioning services. In this embodiment, the deployment cost of positioning equipment can be obtained by calculating the Euclidean distance between each positioning device's location and the average location; the smaller the distance, the lower the cost.

[0242] In this embodiment, the signal radius of the positioning device can be obtained based on the preset attribute information of the positioning device, the signal coverage intensity in the indoor three-dimensional map can be calculated based on the initial positioning device position and the signal radius of the positioning device, and the uniformity of the positioning device and the deployment cost of the positioning device can be calculated based on the initial positioning device position.

[0243] In one optional embodiment of this application, step 106 further includes the following sub-steps:

[0244] S501: The initial positioning device position is subjected to random perturbation processing a preset number of times to obtain several sets of perturbed positioning device positions;

[0245] S502: Calculate the disturbance signal coverage intensity, disturbance positioning device uniformity, and disturbance positioning device deployment cost in the indoor 3D map based on the location of each group of disturbance positioning devices;

[0246] S503: Calculate the adaptability of the disturbance positioning device corresponding to each group of disturbance positioning device locations based on the disturbance signal coverage strength, the uniformity of the disturbance positioning device, and the deployment cost of the disturbance positioning device.

[0247] S504: Compare the fitness of the disturbance positioning devices corresponding to several groups of disturbance positioning device positions, and take the group of disturbance positioning device positions corresponding to the largest disturbance positioning device fitness as the target positioning device position.

[0248] It should be noted that random perturbation processing can be random perturbation, a technique commonly used in optimization algorithms, especially in local search and heuristic algorithms. Its basic idea is to explore other possible solutions in the solution space by making small random changes to the current solution, hoping to find a better solution. In the embodiments of this application, perturbation processing can be a random, minor modification to the coordinate values ​​in the initial two-dimensional position of the positioning device.

[0249] In this embodiment, the initial positioning device location can be subjected to a preset number of random disturbances to obtain several sets of disturbed positioning device locations. Then, based on each set of disturbed positioning device locations, the disturbance signal coverage intensity, the uniformity of the disturbed positioning devices, and the deployment cost of the disturbed positioning devices in the indoor 3D map are calculated. The specific calculation method can refer to the method described above for calculating signal coverage intensity, positioning device uniformity, and positioning device deployment cost. The preset number of disturbances can be set according to the actual situation.

[0250] In this embodiment, the fitness of each group of disturbance positioning devices can be calculated based on the disturbance signal coverage strength, the uniformity of disturbance positioning devices, and the deployment cost of disturbance positioning devices. Then, the fitness values ​​of several groups of disturbance positioning devices are compared, and the group of disturbance positioning devices with the highest fitness value is taken as the target positioning device location. The highest fitness value can be simply the largest numerical value of the disturbance positioning device fitness value.

[0251] In this embodiment, firstly, indoor 3D maps and wireless signal data are acquired. Environmental feature data is then obtained from the indoor 3D map, providing more accurate and reliable map data support for the planning of positioning device locations. This method not only reduces reliance on manual surveys but also improves the rationality of location planning, effectively avoiding problems such as incomplete signal coverage and multipath effects. Secondly, the environmental feature data and wireless signal data are input into a trained positioning device location prediction model for processing, outputting the initial positioning device location. This enables automatic planning of positioning device locations within the current indoor 3D map. Through accurate 3D map data and advanced algorithms, automatic planning can more precisely determine the optimal installation location of the positioning device, effectively avoiding errors and deviations that may occur in manual planning and improving the deployment efficiency of the positioning device. Finally, based on the initial positioning device location, the signal coverage strength, positioning device uniformity, and positioning device deployment cost in the indoor 3D map are calculated. The positioning device fitness is then calculated based on these factors, and the initial positioning device location is adjusted according to the positioning device fitness to obtain the target positioning device location. Deep learning network technology can be used to verify and optimize the signal coverage of positioning networks in indoor environments. This not only ensures the comprehensiveness and uniformity of signal coverage but also takes into account the economic efficiency of deployment costs, thereby achieving optimal resource allocation and ensuring the reliability and stability of the positioning network.

[0252] Reference Figure 2 The diagram shows a structural schematic of a positioning device location planning apparatus according to an embodiment of this application. The apparatus includes:

[0253] Data acquisition module 201 is used to acquire indoor 3D maps and wireless signal data;

[0254] Environmental feature data extraction module 202 is used to obtain environmental feature data based on the indoor 3D map;

[0255] The initial positioning device location acquisition module 203 is used to input the environmental feature data and the wireless signal data into the trained positioning device location prediction model for processing, and output the initial positioning device location.

[0256] The optimization parameter calculation module 204 is used to calculate the signal coverage strength, positioning device uniformity, and positioning device deployment cost in the indoor three-dimensional map based on the initial positioning device location.

[0257] The fitness calculation module 205 is used to calculate the fitness of the positioning device based on the signal coverage strength, the uniformity of the positioning device, and the deployment cost of the positioning device.

[0258] The target positioning device location acquisition module 206 is used to adjust the initial positioning device location according to the positioning device adaptability to obtain the target positioning device location.

[0259] In one optional embodiment of this application, the indoor 3D map is acquired via a vehicle, the vehicle including a lidar and a camera, and the data acquisition module 201 includes:

[0260] The route acquisition module is used to acquire preset routes;

[0261] The two-dimensional image data acquisition module is used to acquire three-dimensional point cloud data through the lidar and two-dimensional image data through the camera during the driving process according to the driving route.

[0262] An initial 3D map synthesis module is used to synthesize an initial 3D map based on the 3D point cloud data and the 2D image data;

[0263] A three-dimensional feature data extraction module is used to extract features from the three-dimensional point cloud data to obtain three-dimensional feature data;

[0264] An indoor 3D map acquisition module is used to process the initial 3D map based on the 3D feature data to acquire the indoor 3D map;

[0265] A wireless signal data acquisition module is used to acquire the wireless signal data according to the preset attribute information of the positioning device.

[0266] In one optional embodiment of this application, the environmental feature data extraction module 202 includes:

[0267] A 3D point cloud data processing module is used to acquire voxel data and raw point data based on the 3D point cloud data;

[0268] The candidate region feature acquisition module is used to extract features from the voxel data and acquire candidate region features.

[0269] The key point feature acquisition module is used to extract features from the original point data and acquire key point features;

[0270] A three-dimensional feature data acquisition module is used to determine the three-dimensional feature data based on the candidate region features and the key point features.

[0271] In one optional embodiment of this application, the location prediction model of the positioning device is trained in the following manner:

[0272] Obtain a standard indoor 3D map, wherein the standard indoor 3D map includes a number of standard positioning devices arranged at the locations of standard indoor positioning devices;

[0273] Obtain standard environmental feature data based on the standard indoor 3D map;

[0274] Obtain standard wireless signal data based on the preset attribute information of the standard positioning device;

[0275] At least one of the standard environmental feature data, the standard wireless signal data, and the standard indoor positioning device location is preprocessed, and the preprocessing includes one or more of the following: data cleaning, missing data imputation, normalization, and classification coding.

[0276] Based on the preprocessed standard environmental feature data, standard wireless signal data, and standard indoor positioning device location, obtain training environmental feature data, training wireless signal data, training indoor positioning device location, and verification environmental feature data and verification wireless signal data.

[0277] The training environment feature data, the training wireless signal data, and the location of the indoor positioning device are used as training samples.

[0278] The training environment feature data and the training wireless signal data are used as inputs to the positioning device location prediction model, and the location of the positioning device in the training room is used as the output of the positioning device location prediction model. The positioning device location prediction model is trained using the training samples to obtain the positioning device location prediction model to be verified.

[0279] The verification environment feature data and the verification wireless signal data are used as verification samples.

[0280] The verification environment feature data and the verification wireless signal data are input into the location prediction model of the positioning device to be verified to obtain the location of the indoor positioning device output by the location prediction model of the positioning device to be verified.

[0281] Calculate the loss function of the positioning device location prediction model based on the location of the positioning device in the training room and the location of the positioning device in the verification room.

[0282] The model parameters of the positioning device location prediction model to be verified are adjusted using the loss function of the positioning device location prediction model. The steps of using the training environment feature data and the training wireless signal data as inputs to the positioning device location prediction model, using the location of the positioning device in the training room as the output of the positioning device location prediction model, and using the training samples to train the positioning device location prediction model are repeated until a preset stopping condition is reached, thus obtaining a trained positioning device location prediction model.

[0283] In one optional embodiment of this application, the positioning device location prediction model is composed of a sequential model, which includes an input layer, several hidden layers, and an output layer. The initial positioning device location acquisition module 203 includes:

[0284] The first positioning device feature acquisition module is used to input the environmental feature data and the wireless signal data into the input layer for processing, and obtain the first positioning device feature output by the input layer.

[0285] The second positioning device feature acquisition module is used to input the first positioning device feature into the hidden layer for processing, and obtain the second positioning device feature output by the hidden layer.

[0286] The initial positioning device location acquisition submodule is used to input the second positioning device features into the output layer for processing, and obtain the initial positioning device location output by the output layer.

[0287] In one optional embodiment of this application, the optimization parameter calculation module 204 includes:

[0288] A positioning device signal radius acquisition module is used to acquire the positioning device signal radius based on preset attribute information of the positioning device.

[0289] A signal coverage strength acquisition module is used to calculate the signal coverage strength in the indoor 3D map based on the initial positioning device location and the signal radius of the positioning device.

[0290] The uniformity and deployment cost acquisition module is used to calculate the uniformity of the positioning device and the deployment cost of the positioning device based on the initial positioning device location.

[0291] In one optional embodiment of this application, the target positioning device location acquisition module 206 includes:

[0292] The disturbance positioning device location acquisition module is used to perform random disturbance processing on the initial positioning device location a preset number of times to obtain several sets of disturbance positioning device locations;

[0293] The disturbance positioning device location processing module is used to calculate the disturbance signal coverage intensity, disturbance positioning device uniformity, and disturbance positioning device deployment cost in the indoor three-dimensional map based on the location of each group of disturbance positioning devices.

[0294] The disturbance positioning device fitness calculation module is used to calculate the disturbance positioning device fitness corresponding to each group of disturbance positioning device locations based on the disturbance signal coverage intensity, the disturbance positioning device uniformity, and the disturbance positioning device deployment cost.

[0295] The target positioning device location acquisition submodule is used to compare the fitness of the disturbance positioning devices corresponding to several sets of disturbance positioning device locations, and take the set of disturbance positioning device locations corresponding to the largest disturbance positioning device fitness as the target positioning device location.

[0296] As the apparatus embodiment is basically similar to the method embodiment, it is described in a relatively simple manner. For relevant details, please refer to the description of the method embodiment.

[0297] An embodiment of this application also provides an electronic device, which may include a processor, a memory, and a computer program stored in the memory and capable of running on the processor. When the computer program is executed by the processor, it implements the positioning device location planning method as described above.

[0298] An embodiment of this application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the positioning device location planning method as described above.

[0299] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation portals are provided for users to choose to authorize or refuse.

[0300] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0301] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, embodiments of the present invention can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of the present invention can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0302] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0303] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0304] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0305] Although preferred embodiments of the present invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present invention.

[0306] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes the aforementioned element.

[0307] The above provides a detailed description of the positioning device location planning method, apparatus, electronic device, and medium. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A positioning device location planning method, characterized in that, The method includes: Acquire indoor 3D maps and wireless signal data; Environmental feature data are obtained based on the indoor 3D map; The environmental feature data and the wireless signal data are input into the trained positioning device location prediction model for processing, and the initial positioning device location is output. Calculate the signal coverage strength, positioning device uniformity, and positioning device deployment cost in the indoor 3D map based on the initial positioning device location; The adaptability of the positioning device is calculated based on the signal coverage strength, the uniformity of the positioning device, and the deployment cost of the positioning device. The initial positioning device position is adjusted according to the positioning device adaptability to obtain the target positioning device position; The positioning device location prediction model is composed of a sequential model, which includes an input layer, several hidden layers, and an output layer. The process of inputting the environmental feature data and the wireless signal data into the trained positioning device location prediction model for processing and outputting the initial positioning device location includes: The environmental feature data and the wireless signal data are input into the input layer for processing to obtain the first positioning device feature output by the input layer. The first positioning device feature is input into the hidden layer for processing to obtain the second positioning device feature output by the hidden layer. The second positioning device feature is input into the output layer for processing to obtain the initial positioning device position output by the output layer.

2. The method according to claim 1, characterized in that, The indoor 3D map is acquired via a vehicle, which includes a lidar unit and a camera. The acquisition of the indoor 3D map and wireless signal data includes: Obtain the preset driving route; During the driving process along the described route, three-dimensional point cloud data is collected by the lidar and two-dimensional image data is collected by the camera. An initial 3D map is synthesized based on the 3D point cloud data and the 2D image data; Feature extraction is performed on the three-dimensional point cloud data to obtain three-dimensional feature data; The initial 3D map is processed based on the 3D feature data to obtain the indoor 3D map; The wireless signal data is obtained based on the preset attribute information of the positioning device.

3. The method according to claim 2, characterized in that, The step of extracting features from the three-dimensional point cloud data to obtain three-dimensional feature data includes: Voxel data and original point location data are obtained from the three-dimensional point cloud data; Feature extraction is performed on the voxel data to obtain candidate region features; Feature extraction is performed on the original point data to obtain key point features; The three-dimensional feature data is determined based on the candidate region features and the key point features.

4. The method according to claim 1, characterized in that, The location prediction model for the positioning device is trained in the following manner: Obtain a standard indoor 3D map, wherein the standard indoor 3D map includes a number of standard positioning devices arranged at the locations of standard indoor positioning devices; Obtain standard environmental feature data based on the standard indoor 3D map; Obtain standard wireless signal data based on the preset attribute information of the standard positioning device; At least one of the standard environmental feature data, the standard wireless signal data, and the standard indoor positioning device location is preprocessed, and the preprocessing includes one or more of the following: data cleaning, missing data imputation, normalization, and classification coding. Based on the preprocessed standard environmental feature data, standard wireless signal data, and standard indoor positioning device location, obtain training environmental feature data, training wireless signal data, training indoor positioning device location, and verification environmental feature data and verification wireless signal data. The training environment feature data, the training wireless signal data, and the location of the indoor positioning device are used as training samples. The training environment feature data and the training wireless signal data are used as inputs to the positioning device location prediction model, and the location of the positioning device in the training room is used as the output of the positioning device location prediction model. The positioning device location prediction model is trained using the training samples to obtain the positioning device location prediction model to be verified. The verification environment feature data and the verification wireless signal data are used as verification samples. The verification environment feature data and the verification wireless signal data are input into the location prediction model of the positioning device to be verified to obtain the location of the indoor positioning device output by the location prediction model of the positioning device to be verified. Calculate the loss function of the positioning device location prediction model based on the location of the positioning device in the training room and the location of the positioning device in the verification room. The model parameters of the positioning device location prediction model to be verified are adjusted using the loss function of the positioning device location prediction model. The steps of using the training environment feature data and the training wireless signal data as inputs to the positioning device location prediction model, using the location of the positioning device in the training room as the output of the positioning device location prediction model, and using the training samples to train the positioning device location prediction model are repeated until a preset stopping condition is reached, thus obtaining a trained positioning device location prediction model.

5. The method according to claim 1, characterized in that, The calculation of signal coverage strength, positioning device uniformity, and positioning device deployment cost in the indoor 3D map based on the initial positioning device location includes: The signal radius of the positioning device is obtained based on the preset attribute information of the positioning device; The signal coverage intensity in the indoor 3D map is calculated based on the initial positioning device location and the signal radius of the positioning device. The uniformity of the positioning devices and the deployment cost of the positioning devices are calculated based on the initial positioning device locations.

6. The method according to claim 5, characterized in that, The step of adjusting the initial positioning device position according to the positioning device adaptability to obtain the target positioning device position includes: The initial positioning device position is subjected to random perturbation processing a preset number of times to obtain several sets of perturbed positioning device positions; Based on the location of each group of disturbance positioning devices, calculate the disturbance signal coverage intensity, disturbance positioning device uniformity, and disturbance positioning device deployment cost in the indoor 3D map; The perturbation positioning device fitness rate corresponding to each group of perturbation positioning device locations is calculated based on the perturbation signal coverage intensity, the perturbation positioning device uniformity, and the perturbation positioning device deployment cost. The fitness of the disturbance positioning devices corresponding to several sets of disturbance positioning device locations is compared, and the set of disturbance positioning device locations corresponding to the largest disturbance positioning device fitness is taken as the target positioning device location.

7. A positioning device for planning the location of a positioning equipment, characterized in that, The device includes: The data acquisition module is used to acquire indoor 3D maps and wireless signal data; An environmental feature data extraction module is used to obtain environmental feature data based on the indoor 3D map. The initial positioning device location acquisition module is used to input the environmental feature data and the wireless signal data into the trained positioning device location prediction model for processing, and output the initial positioning device location. The optimization parameter calculation module is used to calculate the signal coverage strength, positioning device uniformity, and positioning device deployment cost in the indoor 3D map based on the initial positioning device location. The fitness calculation module is used to calculate the fitness of the positioning device based on the signal coverage strength, the uniformity of the positioning device, and the deployment cost of the positioning device. The target positioning device location acquisition module is used to adjust the initial positioning device location according to the positioning device adaptability to obtain the target positioning device location; The positioning device location prediction model is composed of a sequential model, which includes an input layer, several hidden layers, and an output layer. The initial positioning device location acquisition module includes: The first positioning device feature acquisition module is used to input the environmental feature data and the wireless signal data into the input layer for processing, and obtain the first positioning device feature output by the input layer. The second positioning device feature acquisition module is used to input the first positioning device feature into the hidden layer for processing, and obtain the second positioning device feature output by the hidden layer. The initial positioning device location acquisition submodule is used to input the second positioning device features into the output layer for processing, and obtain the initial positioning device location output by the output layer.

8. An electronic device, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein the computer program, when executed by the processor, implements the positioning device location planning method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, which, when executed by a processor, implements the positioning device location planning method as described in any one of claims 1-6.

Images