Vehicle searching method and device, electronic equipment and storage medium

Abstract

The invention discloses a vehicle searching method and device, electronic equipment and a storage medium, and belongs to the technical field of computer application. The vehicle searching method comprises the following steps: receiving a vehicle searching request for a target vehicle; obtaining a geomagnetic fingerprint of the position of the target vehicle; based on the geomagnetic fingerprint of the current position of the user terminal and the geomagnetic fingerprint of the position of the target vehicle, obtaining the position coordinate of the user terminal and the position coordinate of the target vehicle by using a pre-trained positioning model; and determining a navigation path according to the position coordinates of the user terminal, the position coordinates of the target vehicle and environment sensing data acquired in real time. By applying the technical scheme provided by the invention, the navigation path can be quickly determined, and the vehicle searching efficiency is improved.

Description

Technical Field

[0001] This application belongs to the field of computer application technology, and specifically relates to a vehicle finding method, device, electronic device and storage medium. Background Technology

[0002] In daily life, after parking their vehicles and leaving to attend to personal matters, users may forget where they parked when they return, requiring them to search for their cars. This usually relies on the Global Positioning System (GPS) or mobile networks. However, with the increasing scarcity of urban parking resources and the proliferation of complex parking environments such as large underground parking lots and multi-story parking garages, problems such as blocked GPS signals and weak or nonexistent mobile network signals are common, making it difficult for users to find their vehicles.

[0003] Therefore, in environments where GPS signals are blocked, mobile network signals are weak or nonexistent, how to improve vehicle locating efficiency is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0004] The purpose of this application is to provide a vehicle locating method, device, electronic device, and storage medium to improve vehicle locating efficiency.

[0005] To solve the above-mentioned technical problems, this application provides the following technical solution: Firstly, a vehicle locating method is provided, applied to a user terminal, the method comprising: Receive a vehicle location request for the target vehicle; Obtain the geomagnetic fingerprint of the target vehicle's location; Based on the geomagnetic fingerprint of the current location of the user terminal and the geomagnetic fingerprint of the location of the target vehicle, the location coordinates of the user terminal and the location coordinates of the target vehicle are obtained using a pre-trained positioning model. The navigation path is determined based on the location coordinates of the user terminal, the location coordinates of the target vehicle, and the real-time collected environmental perception data.

[0006] Optionally, the geomagnetic fingerprint of the target vehicle's location is obtained by performing anti-interference processing on the initial geomagnetic fingerprint of the target vehicle's location using a pre-trained Generative Adversarial Network (GAN) anti-interference model. The initial geomagnetic fingerprint of the target vehicle's location is generated based on the triaxial magnetic field strength collected by the vehicle's onboard magnetometer, the magnetic field change rate corresponding to the triaxial magnetic field strength, and the triaxial rotation vector provided by the vehicle's inertial measurement unit.

[0007] Optionally, after receiving the vehicle locator request for the target vehicle, the method further includes: Obtain the geomagnetic fingerprint database of the target area, wherein the target area is the parking area of ​​the target vehicle; The process of obtaining the location coordinates of the user terminal and the target vehicle based on the geomagnetic fingerprint of the current location of the user terminal and the geomagnetic fingerprint of the location of the target vehicle, using a pre-trained positioning model, includes: The geomagnetic fingerprint of the current location of the user terminal, the geomagnetic fingerprint of the location of the target vehicle, and the geomagnetic fingerprint database of the target area are input into the pre-trained positioning model to obtain the location coordinates of the user terminal and the location coordinates of the target vehicle.

[0008] Optionally, the geomagnetic fingerprint database of the target area includes the geomagnetic fingerprint and known location coordinates of each relay node in the group cooperative network of the target area; obtaining the geomagnetic fingerprint database of the target area includes: A vehicle-finding assistance request is broadcast in the target area via a short-range communication protocol; Based on the received collaborative response messages, determine the relay nodes used to build the group collaborative network; Each relay node is allocated a communication time slot based on a time division multiple access protocol; The geomagnetic fingerprint and known location coordinates of the corresponding relay node are received in different communication time slots.

[0009] Optionally, the allocation of communication time slots for each relay node based on the time division multiple access protocol includes: Based on the time division multiple access protocol, communication time slots are allocated to each relay node according to the distance between the user terminal and each relay node, with the communication time slots of relay nodes that are closer to the user terminal being allocated earlier.

[0010] Optionally, determining the navigation path based on the location coordinates of the user terminal, the location coordinates of the target vehicle, and real-time collected environmental perception data includes: The location coordinates of the user terminal, the location coordinates of the target vehicle, and the real-time collected environmental perception data are used as state input vectors and input into a pre-trained near-end policy optimization network to obtain the navigation path output by the near-end policy optimization network. The reward function of the near-end policy optimization network targets at least one of the following objectives: Minimize the time it takes for the user terminal to reach the target vehicle; Minimize the risk of collision with dynamic obstacles; Minimize the energy cost of the communication link; Maximize the approach progress to the target vehicle.

[0011] Optionally, the method further includes: The rendered output is superimposed on the real-world image captured by the image acquisition device on the user terminal, representing the navigation path.

[0012] Optionally, the method further includes: If the distance to the target vehicle is detected to have reached a target threshold, a prompt message is output to alert the user that they are about to arrive at the location of the target vehicle.

[0013] Secondly, a vehicle locator is provided for use in a user terminal, the device comprising: The request receiving module is used to receive vehicle search requests for the target vehicle. The data acquisition module is used to acquire the geomagnetic fingerprint of the location of the target vehicle; The positioning module is used to obtain the position coordinates of the user terminal and the position coordinates of the target vehicle based on the geomagnetic fingerprint of the current location of the user terminal and the geomagnetic fingerprint of the location of the target vehicle, using a pre-trained positioning model. The path determination module is used to determine the navigation path based on the location coordinates of the user terminal, the location coordinates of the target vehicle, and real-time collected environmental perception data.

[0014] Thirdly, an electronic device is provided, comprising a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the method as described in the first aspect.

[0015] Fourthly, a readable storage medium is provided, on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method as described in the first aspect.

[0016] Fifthly, a program product is provided, the program product including instructions stored in a readable storage medium and adapted to be read and executed by a processor to cause a device having the processor to perform the steps of the method as described in the first aspect.

[0017] By applying the technical solution provided in the embodiments of this application, after receiving a vehicle search request for a target vehicle, the user terminal obtains the geomagnetic fingerprint of the target vehicle's location. Then, using a positioning model, based on the geomagnetic fingerprint of the user terminal's current location and the geomagnetic fingerprint of the target vehicle's location, the user terminal's location coordinates and the target vehicle's location coordinates are obtained. Finally, based on the user terminal's location coordinates, the target vehicle's location coordinates, and the real-time collected environmental perception data, a navigation path can be quickly determined, improving vehicle search efficiency.

[0018] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating one implementation of the vehicle-finding method in this application. Figure 2 This is a flowchart illustrating another implementation of the vehicle-finding method in this application. Figure 3 This is a schematic diagram of the structure of a vehicle-finding device according to an embodiment of this application; Figure 4 This is a schematic diagram of the structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first" and "second" are generally of the same class, not limited in number; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0023] The core of this application is to provide a vehicle-finding method. This method can be applied to a user terminal, specifically to a vehicle-finding application installed on the user terminal, and used in scenarios such as vehicle location. For example, when a user parks their vehicle, leaves, and returns after completing tasks, they may find it difficult to locate the vehicle due to complex parking conditions, lack of GPS or mobile network signals, or weak GPS or mobile network signals. In this case, the user can trigger the vehicle-finding application installed on the user terminal to execute the technical solution provided in the embodiments of this application to quickly locate the vehicle. As another example, a user has booked a ride-hailing service and needs the service to pick them up in a certain area. If the area lacks GPS or mobile network signals, or the signals are weak, the user may find it difficult to locate the ride-hailing service. In this case, the user can trigger the vehicle-finding application installed on the user terminal to execute the technical solution provided in the embodiments of this application to quickly locate the ride-hailing service. The vehicle-finding application installed on the user terminal can be integrated into the ride-hailing application.

[0024] See Figure 1 The diagram shown is a flowchart of a vehicle locating method provided in this application embodiment. The method may include the following steps: S110: Receive a vehicle search request for the target vehicle.

[0025] In this embodiment of the application, the target vehicle refers to the vehicle that the user is looking for. The user can send a vehicle search request for the target vehicle through preset icons, buttons, etc. on the user terminal.

[0026] For example, after a user parks their target vehicle in a corner parking space (signal blind spot) in Zone B on the fourth basement floor of a large shopping mall, goes shopping for 2 hours, and returns, they can open the car-finding application installed on their terminal and initiate a car-finding request for the target vehicle.

[0027] After receiving a vehicle search request for the target vehicle, the user terminal can continue to perform subsequent steps.

[0028] S120: Obtain the geomagnetic fingerprint of the target vehicle's location.

[0029] In this embodiment of the application, after the user terminal receives a vehicle search request for the target vehicle, it can obtain the geomagnetic fingerprint of the target vehicle's location.

[0030] Optionally, when the target vehicle is parked and the user terminal is still within the target vehicle's reach, the target vehicle can generate a geomagnetic fingerprint of its location. Using short-range communication technology, the geomagnetic fingerprint of its location can be transmitted to the user terminal. The user terminal can cache the geomagnetic fingerprint of the target vehicle's location locally. Upon receiving a vehicle search request for the target vehicle, the user terminal can read the geomagnetic fingerprint of the target vehicle's location from the cache.

[0031] Short-range communication technologies may include at least one of the following: Ultra Wide Band (UWB), Wireless Fidelity (Wi-Fi Direct), and Zigbee.

[0032] Optionally, after receiving a vehicle location request for the target vehicle, if the user terminal and the target vehicle can communicate via short-range communication technology, the user terminal can directly obtain the geomagnetic fingerprint of the target vehicle's location.

[0033] Geomagnetic fingerprints are the static magnetic field distribution characteristics of space formed by the interaction between the Earth's magnetic field and the surrounding building's steel structure, underground metal facilities, etc. The geomagnetic fingerprint of a target vehicle's location refers to the static magnetic field distribution characteristics of the space where the target vehicle is parked. Geomagnetic fingerprints are determined by the distribution of steel reinforcement in the building structure and possess uniqueness and long-term stability.

[0034] S130: Based on the geomagnetic fingerprint of the current location of the user terminal and the geomagnetic fingerprint of the location of the target vehicle, the location coordinates of the user terminal and the target vehicle are obtained using a pre-trained positioning model.

[0035] In this embodiment, the user terminal can generate a geomagnetic fingerprint of the current location through data acquisition. Optionally, the user terminal can acquire the triaxial magnetic field strength of the current location through a magnetometer, calculate the magnetic field change rate corresponding to the triaxial magnetic field strength, obtain the triaxial rotation vector provided by the inertial measurement unit (IMU), and generate the geomagnetic fingerprint of the current location based on formula (1): (1) in, express Axial magnetic field strength; express Axial magnetic field strength; Indicate Triaxial magnetic field strength; the unit of triaxial magnetic field strength can be microtesla (µT). ); express Rate of change of axial magnetic field; express Rate of change of axial magnetic field; express The rate of change of the axial magnetic field; the rate of change of the magnetic field can also be understood as the magnetic field gradient, and the unit can be microtesla per second (µT / s). ), used to identify dynamic interference; express Axis rotation vector; express Axis rotation vector; express Three-axis rotation vector; the unit of three-axis rotation vector can be radians per second (rad / s).

[0036] The localization model can be pre-trained in the backend system or on the server, and the trained localization model can be deployed on the user terminal.

[0037] Optionally, the localization model may include convolutional neural network (CNN) layers and bidirectional long short-term memory (BiLSTM) layers. The localization model is shown in equation (2): (2) The CNN layer parameters are as follows: convolutional kernel: 5×5; number of channels: 32→16 (feature map dimensionality reduction); pooling layer: max pooling (2,2), stride 2; BiLSTM layer parameters: Hidden units: 64 (128 units in total, bidirectional); Dropout rate: 0.3 (to prevent overfitting); The loss function for the localization model is Huber Loss (a loss function that combines mean squared error and mean absolute error) (δ=1.0), which balances the influence of outliers and enhances robustness.

[0038] The user terminal can input the geomagnetic fingerprint of its current location and the geomagnetic fingerprint of the target vehicle's location into the positioning model. The positioning model then performs positioning calculations to obtain the user terminal's location coordinates and the target vehicle's location coordinates. Achieving dual-end positioning—"where am I?" and "where is the vehicle?"—is a prerequisite for path planning.

[0039] S140: Determine the navigation path based on the location coordinates of the user terminal, the location coordinates of the target vehicle, and the real-time collected environmental perception data.

[0040] In this embodiment, the user terminal can collect environmental perception data in real time, such as through millimeter-wave radar or a mobile phone camera. After obtaining the location coordinates of the user terminal and the target vehicle through a positioning model, the user terminal's location coordinates are used as the starting point, and the target vehicle's location coordinates are used as the ending point. Based on the real-time collected environmental perception data, path planning is performed to determine the navigation path.

[0041] During the movement of the user terminal, the geomagnetic fingerprint of the current location can be obtained in real time. By inputting the geomagnetic fingerprint of the current location and the geomagnetic fingerprint of the target vehicle's location into the positioning model, the position coordinates of the user terminal and the target vehicle can be obtained. Then, based on the position coordinates of the user terminal, the position coordinates of the target vehicle, and the real-time collected environmental perception data, the navigation path can be updated in real time.

[0042] By applying the method provided in the embodiments of this application, after receiving a vehicle search request for a target vehicle, the user terminal obtains the geomagnetic fingerprint of the target vehicle's location. Then, using a positioning model, based on the geomagnetic fingerprint of the user terminal's current location and the geomagnetic fingerprint of the target vehicle's location, the user terminal's location coordinates and the target vehicle's location coordinates can be obtained. Finally, based on the user terminal's location coordinates, the target vehicle's location coordinates, and the real-time collected environmental perception data, a navigation path can be quickly determined, improving vehicle search efficiency.

[0043] In some embodiments of this application, the geomagnetic fingerprint of the target vehicle's location is obtained by performing anti-interference processing on the initial geomagnetic fingerprint of the target vehicle's location using a pre-trained Generative Adversarial Network (GAN) anti-interference model. The initial geomagnetic fingerprint of the target vehicle's location is generated based on the triaxial magnetic field strength collected by the vehicle's onboard magnetometer, the magnetic field change rate corresponding to the triaxial magnetic field strength, and the triaxial rotation vector provided by the vehicle's inertial measurement unit.

[0044] In this embodiment, after the target vehicle is parked and turned off, the triaxial magnetic field strength can be collected by the on-board magnetometer, and the magnetic field change rate corresponding to the triaxial magnetic field strength can be calculated to obtain the triaxial rotation vector provided by the on-board IMU. Based on formula (1), the initial geomagnetic fingerprint of the target vehicle's location is generated. Optionally, if the on-board magnetometer or on-board IMU malfunctions, the triaxial magnetic field strength can be collected by the magnetometer of the user terminal before the user terminal moves away from the target vehicle. The triaxial rotation vector provided by the IMU of the user terminal can then be obtained, and the data obtained by the user terminal can be transmitted to the target vehicle through short-range communication technology, using the user terminal as an auxiliary or emergency measure.

[0045] The initial geomagnetic fingerprint of the target vehicle's location may contain noise. A pre-trained Generative Adversarial Network (GAN) can be used to perform anti-interference processing (real-time denoising) on ​​this initial geomagnetic fingerprint, resulting in a clean geomagnetic fingerprint of the target vehicle's location, thus improving positioning accuracy. The input to the GAN anti-interference model is the noisy initial geomagnetic fingerprint, and the output is a clean geomagnetic fingerprint, eliminating dynamic interference such as elevator movement and metal carts. The clean geomagnetic fingerprint output by the GAN anti-interference model is also a highly stable anti-interference geomagnetic fingerprint, identifying the magnetic identity of the target vehicle's location.

[0046] GAN anti-interference models include generators ( ) and discriminator ( The generator can be trained by a backend system or server, using noisy and clean data pairs for adversarial learning during the training phase. The discriminator aims to distinguish between real and generated data, while the generator aims to generate as realistic pseudo-data as possible to deceive the discriminator. During training, historically collected clean geomagnetic data (without metal interference) can be used as positive samples, while artificially added simulated interference (such as moving metal objects or the magnetic field of an elevator) can generate negative samples. Through adversarial training, the generator learns to filter out dynamic interference and retain stable geomagnetic characteristics.

[0047] After training, the GAN robust model can be deployed in the vehicle, using only the generator during the deployment phase. Utilizing the GAN robust model ensures accurate geomagnetic fingerprint extraction even under significant environmental disturbances.

[0048] The generator loss function of the GAN anti-interference model is shown in Equation (3): (3) The discriminator loss function of the GAN anti-interference model is shown in Equation (4): (4) in, The dimension of random noise is in the reasonable range of 64 to 256, with an optimal value of 128. Within this reasonable range, the generation capacity can be reduced due to too low a dimension, or the computational burden can be increased due to too high a dimension. Represents random noise The probability distribution; Represents a real data sample. Represents real data samples The probability distribution; The reasonable range for the learning rate is 0.0001 to 0.001, with an optimal value of 0.0002, used for step size control in the Adam optimizer. Within this reasonable range, a learning rate that is too large can prevent oscillations, while a value that is too small can lead to slow convergence. The reasonable range for the number of training epochs is 100 to 300, with an optimal value of 200, to ensure that the model converges to Nash equilibrium. Within this reasonable range, the number of training epochs can reduce the risk of overfitting.

[0049] The preferred values ​​of the above parameters are only examples and can be adjusted within a reasonable range in different application scenarios.

[0050] In some embodiments of this application, after receiving a vehicle locator request for the target vehicle, the method may further include the following steps: Obtain the geomagnetic fingerprint database of the target area, which is the parking area of ​​the target vehicle; Based on the geomagnetic fingerprint of the user terminal's current location and the geomagnetic fingerprint of the target vehicle's location, and using a pre-trained positioning model, the location coordinates of the user terminal and the target vehicle can be obtained, which may include the following steps: The geomagnetic fingerprint of the user terminal's current location, the geomagnetic fingerprint of the target vehicle's location, and the geomagnetic fingerprint database of the target area are input into the pre-trained localization model to obtain the location coordinates of the user terminal and the target vehicle.

[0051] In this embodiment, after receiving a vehicle location request for a target vehicle, the user terminal can obtain a geomagnetic fingerprint database of the target area. The target area is the parking area of ​​the target vehicle, such as a parking lot where the target vehicle is parked. The geomagnetic fingerprint database of the target area can include geomagnetic fingerprints and location coordinates of multiple locations.

[0052] The user terminal can input the geomagnetic fingerprint of its current location, the geomagnetic fingerprint of the target vehicle's location, and the geomagnetic fingerprint database of the target area into a pre-trained positioning model. Through the calculation of the positioning model, the position coordinates of the user terminal and the target vehicle can be obtained from the output of the positioning model.

[0053] When using a positioning model to determine the location coordinates of a user terminal and a target vehicle, incorporating a geomagnetic fingerprint database of the target area as a reference helps improve positioning accuracy.

[0054] Optionally, the geomagnetic fingerprint database of the target area can be pre-stored in the cloud and shared with the user terminal via Vehicle-to-Everything (V2X) technology. If the target area is covered by a dedicated V2X network, the user terminal can still obtain the geomagnetic fingerprint database of the target area from the cloud via V2X technology, even if the GPS signal or mobile network signal in the target area is weak.

[0055] In some embodiments of this application, the geomagnetic fingerprint database of the target area includes the geomagnetic fingerprint and known location coordinates of each relay node in the group cooperative network of the target area; obtaining the geomagnetic fingerprint database of the target area may include the following steps: A vehicle-finding collaboration request is broadcast in the target area using a short-range communication protocol; Based on the received collaborative response messages, determine the relay nodes used to build the group collaborative network; Each relay node is allocated a communication time slot based on a time division multiple access protocol; The geomagnetic fingerprint and known location coordinates of the corresponding relay node are received in different communication time slots.

[0056] For ease of description, the above steps will be combined for explanation.

[0057] In this embodiment, after receiving a vehicle-finding request for a target vehicle, the user terminal, knowing that the user needs to find the target vehicle, can broadcast a vehicle-finding collaboration request via a short-range communication protocol. The collaboration request can carry the user terminal's identifier and task type. Other vehicles or other user terminals that receive the collaboration request, if their own resources (such as battery power, load, etc.) allow, can send a collaboration response message to the user terminal. The collaboration response message can carry device identifier, device type, signal strength, remaining battery power, etc. Based on the received collaboration response messages, the user terminal determines the relay nodes used to construct the group collaborative network. Optionally, the user terminal can determine the distance to each other device that sent the collaboration response message, as well as the device type, signal strength, and remaining battery power of each other device that sent the collaboration response message, based on the received collaboration response messages, and then determine at least one relay node based on the distance to each other device that sent the collaboration response message, as well as the device type, signal strength, and remaining battery power of each other device that sent the collaboration response message.

[0058] After the user terminal determines the relay nodes used to build the group cooperative network, it can allocate communication time slots to each relay node based on the Time Division Multiple Access (TDMA) protocol. Optionally, the length of each communication time slot... It can be 20ms, and the total latency is It must be less than or equal to 1 second. This indicates the number of relay nodes.

[0059] Optionally, based on the time division multiple access protocol, communication time slots can be allocated to each relay node according to the distance between the user terminal and each relay node, with the communication time slots of relay nodes that are closer to the user terminal being allocated earlier.

[0060] Different relay nodes are located at different distances from the user terminal. Optionally, the user terminal can determine its distance to a relay node based on the time it sends a vehicle-finding cooperation request and the time it receives a cooperation response message from the relay node. The user terminal can allocate communication time slots to each relay node based on its distance, thus differentiating the information transmitted by different relay nodes. For any given relay node, the closer it is to the user terminal, the earlier its communication time slot. Relay nodes whose distance to the user terminal is less than or equal to a preset first distance threshold can be called anchor nodes for the user terminal. Anchor nodes are prioritized in scheduling, which helps ensure the calibration accuracy of the positioning signal. The first distance threshold can be set and adjusted according to actual conditions, such as setting it to 10m.

[0061] In different communication time slots, the geomagnetic fingerprint and known location coordinates of the corresponding relay node can be received.

[0062] By using a short-range communication protocol, the geomagnetic fingerprint and known location coordinates of each relay node in the group collaborative network are obtained, reducing the dependence on parking environment infrastructure.

[0063] Optionally, the user terminal can obtain the geomagnetic fingerprint of the target vehicle's location based on the group collaborative network, that is, the target vehicle transmits the geomagnetic fingerprint of the target vehicle's location to the user terminal through the group collaborative network.

[0064] like Figure 2As shown, after the target vehicle is parked and turned off, an initial geomagnetic fingerprint of the target vehicle's location is generated based on the three-axis magnetic field strength, the corresponding rate of change of the magnetic field strength, and the three-axis rotation vector. After anti-interference processing using a GAN anti-interference model, the geomagnetic fingerprint of the target vehicle's location is obtained. When the user terminal needs to locate the target vehicle, the geomagnetic fingerprint of the target vehicle's location, as well as the geomagnetic fingerprint and known coordinates of each relay node's location, are obtained using a group cooperative network. The geomagnetic fingerprint of the user terminal's current location, the geomagnetic fingerprint of the target vehicle's location, the geomagnetic fingerprint and known coordinates of each relay node's location are input into the positioning model to obtain the user terminal's position coordinates and the target vehicle's position coordinates. Based on the user terminal's position coordinates, the target vehicle's position coordinates, and the real-time collected environmental perception data, the navigation path is determined.

[0065] In some embodiments of this application, determining a navigation path based on the location coordinates of the user terminal, the location coordinates of the target vehicle, and real-time collected environmental perception data may include the following steps: The user terminal's location coordinates, the target vehicle's location coordinates, and the real-time collected environmental perception data are used as state input vectors and input into a pre-trained near-end policy optimization network to obtain the navigation path output by the near-end policy optimization network. The reward function of the near-end policy optimization network targets at least one of the following objectives: Minimize the time it takes for the user terminal to reach the target vehicle; Minimize the risk of collision with dynamic obstacles; Minimize the energy cost of the communication link; Maximize the approach progress to the target vehicle.

[0066] In this embodiment, a proximal policy optimization (PPO) network can be pre-trained. The training process of this PPO network can be performed in a background system or on a server. After training, the trained PPO network is deployed to the user terminal. In one possible implementation, the PPO network may include an input layer, a 5×5 convolutional layer, 64 BiLSTM hidden units, and an output layer.

[0067] The user terminal inputs its location coordinates, the target vehicle's location coordinates, and real-time environmental perception data as state input vectors into the near-end policy optimization network. The near-end policy optimization network then outputs the navigation path from the user terminal to the target vehicle through computation.

[0068] The reward function of the near-end policy optimization network targets at least one of the following objectives: minimizing the time it takes for the user terminal to reach the target vehicle, minimizing the collision risk with dynamic obstacles, minimizing the energy cost of the communication link, and maximizing the approach progress to the target vehicle. The collision risk with dynamic obstacles can be determined based on environmental perception data, and the energy cost of the communication link can be determined based on the remaining power of the relay nodes in the communication link.

[0069] The reward function of the near-end policy optimization network can be a multi-objective reward function, as shown in formula (5): (5) in, This indicates the estimated time it takes for the user terminal to reach the target vehicle; This indicates the risk of collision with dynamic obstacles. If the distance to a pedestrian or cart is determined to be less than or equal to a preset second distance threshold based on environmental perception data, the risk value is set to 1; otherwise, it is set to 0. The second distance threshold can be set and adjusted according to the actual situation, such as setting it to 1.5m. This represents the energy cost of the communication link. If the remaining power of the relay node on the communication link is less than or equal to a preset percentage of power, such as 20%, then the weight corresponding to the energy cost is increased. , such as Multiply by 2; This represents a weight that emphasizes obstacle avoidance priority; for example, its value can be 0.6. This represents another weight, or can be understood as an energy consumption adjustment coefficient, which, for example, can have a value of 0.1.

[0070] The loss function (with clipping) of the near-end policy optimization network is shown in Equation (6): (6) in, , representing the probability ratio between the old and new strategies. Optimize network parameters for near-end strategies; Indicates at time step The generalized advantage estimation (GAE) is calculated using the formula shown in formula (7): (7) in, This represents the shear ratio, which can be set to 0.2 to limit the policy update magnitude and prevent drastic updates. Indicates at time step Temporal Difference Error measures the immediate reward at the current moment. Adding the value of future states The prediction, and the value of the current state. The difference between predictions. It reflects the value function. The prediction error; Indicates at time step The immediate reward can be understood as the feedback that the user terminal receives directly from the environment after performing an action. For example, it could be the amount of distance the user terminal moves one step closer to the target vehicle (positive reward), the reward for avoiding a collision, or the penalty for increased communication energy consumption (negative reward). This represents the discount factor, a hyperparameter between [0, 1]. It determines the importance of future rewards. The closer it is to 1, the more it values ​​long-term interests; The closer it is to 0, the more focused it is on immediate rewards. In a car-finding scenario, this is typically set... It should be close to 1, such as 0.95 or 0.99, because finding a car is a task that requires long-term route planning. Indicates the state The State Value Function (SFC) is a function learned by a neural network that estimates the state from the current state. Initially, calculate the expected cumulative discount reward that can be obtained by acting according to the current strategy; represents the smoothing coefficient of GAE (GAE's Lambda). A hyperparameter between [0, 1] used to control the bias-variance trade-off in the advantage estimation.

[0071] when At that time, GAE degenerates into a single-step advantage estimate (GAE). It has low variance but high bias (because it only considers one step). when At this point, GAE degenerates into Monte Carlo advantage estimation, which is the sum of all future rewards minus It has low bias but high variance (because it depends on the future reward of the entire trajectory, resulting in high noise). Choose A stable and efficient compromise is obtained between [0.8, 0.95]. The time step index, representing the summation, is an integer from 0 to infinity, indicating the time step. The next step. correspond , correspond And so on; This represents the decay factor, which varies with... The error decreases exponentially with increasing frequency, ensuring that the timing difference error is minimized in more distant future time steps. Current advantage estimation Its impact is diminishing. Control time discounts, The degree of smoothness is controlled, and the two work together to form this attenuation factor.

[0072] By optimizing the network through near-end strategies, safe, efficient, and energy-saving navigation paths can be output, enabling dynamic obstacle avoidance and resource optimization, and making it able to cope with complex crowd flow scenarios.

[0073] In some embodiments of this application, the method may further include the following steps: The rendered output is the navigation path superimposed on the real-world image captured by the image acquisition device on the user terminal.

[0074] In this embodiment of the application, after determining the navigation path based on the location coordinates of the user terminal, the location coordinates of the target vehicle, and the real-time collected environmental perception data, Simultaneous Localization and Mapping (SLAM) pose estimation can be performed. The SLAM pose estimation formula is shown in formula (8): (8) in, Indicates the current image frame; This indicates that tracking and map construction are performed using feature points from the Accelerated SegmentTest (FAST) and the Binary Robust Independent Elementary Features (BRIEF).

[0075] One possible parameter configuration for SLAM pose estimation is as follows: The keyframe interval is 0.5s to balance computational load and positioning accuracy; The loop closure detection threshold is greater than 85% to prevent false matches; The feature matching error is less than 0.3m, which is used to ensure the continuity of the navigation path.

[0076] The determined navigation path is overlaid onto the real-world image captured by the image acquisition device on the user terminal and rendered for output, thereby improving navigation effectiveness.

[0077] Optionally, augmented reality (AR) guide arrows can be displayed in the output navigation path, allowing users to follow the AR arrows to navigate around temporary obstacles and eventually reach the vicinity of the target vehicle.

[0078] Optionally, a heat map reflecting the degree of path congestion can be displayed simultaneously, which can be fed back by a collaborative network of groups.

[0079] This helps to provide users with an intuitive and immersive navigation experience.

[0080] In some embodiments of this application, the method may further include the following steps: If the distance to the target vehicle is detected to have reached the target threshold, a prompt message is output to inform the user that they are about to arrive at the location of the target vehicle.

[0081] In this embodiment, the distance between the user terminal and the target vehicle can be detected in real time. When the distance between the user terminal and the target vehicle reaches a target threshold, a prompt message can be output to indicate that the user is about to arrive at the location of the target vehicle. The target threshold can be set and adjusted according to actual conditions, such as being set to 5m.

[0082] Optionally, when the user terminal is close to the target vehicle, such as within 10m, the target vehicle can be triggered to automatically flash its lights and / or honk its horn as a warning. When the user terminal approaches the target vehicle, the vehicle door can be automatically unlocked.

[0083] This enables destination confirmation and interaction between people and vehicles.

[0084] The technical solutions provided in this application can effectively locate vehicles in scenarios without GPS or mobile network signals, such as underground garages and tunnels. By utilizing geomagnetic fingerprinting and group collaborative networks, blind spot vehicle finding with zero infrastructure dependence can be achieved, and reinforcement learning path planning can be integrated to avoid dynamic obstacles.

[0085] The embodiments of this application can reuse existing vehicle sensors when applied, without the need for parking lot modifications, resulting in lower upgrade costs. Moreover, geomagnetic fingerprints have strong stability under concrete structures, low error rates in scenarios such as heavy rain or strong light, and good anti-interference performance.

[0086] It should be noted that the vehicle-finding method provided in this application embodiment can be executed by a vehicle-finding device, or by a control module within the vehicle-finding device for executing the vehicle-finding method. This vehicle-finding device is applied to a user terminal. This application embodiment uses the execution of the vehicle-finding method by a vehicle-finding device as an example to illustrate the vehicle-finding method provided in this application embodiment.

[0087] See Figure 3 The diagram shown is a structural schematic of the vehicle finding device provided in an embodiment of this application, which includes the following modules: The request receiving module 410 is used to receive a vehicle search request for the target vehicle. The data acquisition module 420 is used to acquire the geomagnetic fingerprint of the target vehicle's location; The positioning module 430 is used to obtain the location coordinates of the user terminal and the target vehicle based on the geomagnetic fingerprint of the current location of the user terminal and the geomagnetic fingerprint of the location of the target vehicle, using a pre-trained positioning model. The path determination module 440 is used to determine the navigation path based on the location coordinates of the user terminal, the location coordinates of the target vehicle, and the environmental perception data collected in real time.

[0088] Using the apparatus provided in the embodiments of this application, after receiving a vehicle search request for a target vehicle, the user terminal obtains the geomagnetic fingerprint of the target vehicle's location. Then, using a positioning model, based on the geomagnetic fingerprint of the user terminal's current location and the geomagnetic fingerprint of the target vehicle's location, the user terminal's location coordinates and the target vehicle's location coordinates can be obtained. Finally, based on the user terminal's location coordinates, the target vehicle's location coordinates, and the real-time collected environmental perception data, a navigation path can be quickly determined, improving vehicle search efficiency.

[0089] In some embodiments of this application, the geomagnetic fingerprint of the target vehicle's location is obtained by performing anti-interference processing on the initial geomagnetic fingerprint of the target vehicle's location using a pre-trained Generative Adversarial Network (GAN) anti-interference model. The initial geomagnetic fingerprint of the target vehicle's location is generated based on the triaxial magnetic field strength collected by the vehicle's onboard magnetometer, the magnetic field change rate corresponding to the triaxial magnetic field strength, and the triaxial rotation vector provided by the vehicle's inertial measurement unit.

[0090] In some embodiments of this application, the data acquisition module 420 is further configured to: After receiving a vehicle locating request for the target vehicle, the system retrieves the geomagnetic fingerprint database of the target area, which is the parking area of ​​the target vehicle. Positioning module 430 is specifically used for: The geomagnetic fingerprint of the user terminal's current location, the geomagnetic fingerprint of the target vehicle's location, and the geomagnetic fingerprint database of the target area are input into the pre-trained localization model to obtain the location coordinates of the user terminal and the target vehicle.

[0091] In some embodiments of this application, the geomagnetic fingerprint database of the target area includes the geomagnetic fingerprint and known location coordinates of each relay node in the group cooperative network of the target area; the data acquisition module 420 is specifically used for: A vehicle-finding collaboration request is broadcast in the target area using a short-range communication protocol; Based on the received collaborative response messages, determine the relay nodes used to build the group collaborative network; Each relay node is allocated a communication time slot based on a time division multiple access protocol; The geomagnetic fingerprint and known location coordinates of the corresponding relay node are received in different communication time slots.

[0092] In some embodiments of this application, the data acquisition module 420 is specifically used for: Based on the time division multiple access protocol, communication time slots are allocated to each relay node according to the distance between the user terminal and each relay node, with the communication time slots of relay nodes that are closer to the user terminal being allocated earlier.

[0093] In some embodiments of this application, the path determination module 440 is specifically used for: The user terminal's location coordinates, the target vehicle's location coordinates, and the real-time collected environmental perception data are used as state input vectors and input into a pre-trained near-end policy optimization network to obtain the navigation path output by the near-end policy optimization network. The reward function of the near-end policy optimization network targets at least one of the following objectives: Minimize the time it takes for the user terminal to reach the target vehicle; Minimize the risk of collision with dynamic obstacles; Minimize the energy cost of the communication link; Maximize the approach progress to the target vehicle.

[0094] In some embodiments of this application, a first output module is also included, for: The rendered output is the navigation path superimposed on the real-world image captured by the image acquisition device on the user terminal.

[0095] In some embodiments of this application, a second output module is also included, for: If the distance to the target vehicle is detected to have reached the target threshold, a prompt message is output to inform the user that they are about to arrive at the location of the target vehicle.

[0096] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the relevant method embodiments, and will not be repeated here to avoid repetition.

[0097] Corresponding to the above method embodiments, this application also provides an electronic device, including a processor, a memory, and a program or instructions stored in the memory and executable on the processor. When the program or instructions are executed by the processor, they implement the various processes of the above method embodiments and achieve the same technical effects. To avoid repetition, they will not be described again here.

[0098] Figure 4 A schematic diagram of the hardware structure of an electronic device to implement an embodiment of this application.

[0099] The electronic device 500 includes, but is not limited to, components such as: radio frequency unit 501, network module 502, audio output unit 503, input unit 504, sensor 505, display unit 506, user input unit 507, interface unit 508, memory 509, and processor 510.

[0100] It should be understood that, in this embodiment, the input unit 504 may include a graphics processor 5041 and a microphone 5042. The graphics processor 5041 processes image data of still images or videos obtained by an image capture device (such as a camera) in video capture mode or image capture mode. The display unit 506 may include a display panel 5061, which may be configured in the form of a liquid crystal display, an organic light-emitting diode, or the like. The user input unit 507 includes at least one of a touch panel 5071 and other input devices 5072. The touch panel 5071 is also called a touch screen. The touch panel 5071 may include two parts: a touch detection device and a touch controller. Other input devices 5072 may include, but are not limited to, physical keyboards, function keys (such as volume control buttons, power buttons, etc.), trackballs, mice, and joysticks, which will not be described in detail here.

[0101] Those skilled in the art will understand that the electronic device 500 may also include a power supply (such as a battery) for supplying power to various components. The power supply may be logically connected to the processor 510 through a power management system, thereby enabling functions such as managing charging, discharging, and power consumption through the power management system. Figure 4The electronic device structure shown does not constitute a limitation on the electronic device. The electronic device may include more or fewer components than shown, or combine certain components, or have different component arrangements, which will not be elaborated here.

[0102] The input unit 504 is used to receive a vehicle search request for the target vehicle. Radio frequency unit 501 is used to acquire the geomagnetic fingerprint of the target vehicle's location; Processor 510 is used to obtain the position coordinates of the electronic device and the target vehicle based on the geomagnetic fingerprint of the current location of the electronic device and the geomagnetic fingerprint of the location of the target vehicle, using a pre-trained localization model. The processor 510 is also used to determine the navigation path based on the position coordinates of the electronic device, the position coordinates of the target vehicle, and the environmental perception data collected in real time.

[0103] The electronic device provided in this application, upon receiving a vehicle search request for a target vehicle, obtains the geomagnetic fingerprint of the target vehicle's location. Then, using a positioning model, based on the geomagnetic fingerprint of the electronic device's current location and the geomagnetic fingerprint of the target vehicle's location, the location coordinates of the electronic device and the target vehicle can be obtained. Finally, based on the location coordinates of the electronic device, the location coordinates of the target vehicle, and the real-time collected environmental perception data, a navigation path can be quickly determined, improving vehicle search efficiency.

[0104] This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above method embodiments and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0105] The processor is the processor in the above-described electronic device embodiment. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0106] Furthermore, it should be noted that this application also provides a program product or program, which may include instructions that can be stored in a readable storage medium. The processor of the device reads the instructions from the readable storage medium and executes the instructions, causing the device to perform the descriptions in the corresponding method embodiments above; therefore, these will not be repeated here. Additionally, the beneficial effects of using the same method will also not be repeated. For technical details not disclosed in the program product or program embodiments involved in this application, please refer to the descriptions in the method embodiments of this application.

[0107] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus 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 apparatus. 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 apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0108] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0109] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many different implementations and modifications to this application without departing from the spirit and scope of the claims. All such implementations, improvements and modifications are within the scope of protection of this application.

Claims

1. A vehicle location method, characterized in that, Applied to a user terminal, the method includes: Receive a vehicle location request for the target vehicle; Obtain the geomagnetic fingerprint of the target vehicle's location; Based on the geomagnetic fingerprint of the current location of the user terminal and the geomagnetic fingerprint of the location of the target vehicle, the location coordinates of the user terminal and the location coordinates of the target vehicle are obtained using a pre-trained positioning model. The navigation path is determined based on the location coordinates of the user terminal, the location coordinates of the target vehicle, and the real-time collected environmental perception data.

2. The method according to claim 1, characterized in that, The geomagnetic fingerprint of the target vehicle's location is obtained by using a pre-trained Generative Adversarial Network (GAN) anti-interference model to process the initial geomagnetic fingerprint of the target vehicle's location to prevent interference. The initial geomagnetic fingerprint of the target vehicle's location is generated based on the triaxial magnetic field strength collected by the vehicle's onboard magnetometer, the magnetic field change rate corresponding to the triaxial magnetic field strength, and the triaxial rotation vector provided by the vehicle's inertial measurement unit.

3. The method according to claim 1, characterized in that, After receiving the vehicle location request for the target vehicle, the method further includes: Obtain the geomagnetic fingerprint database of the target area, wherein the target area is the parking area of ​​the target vehicle; The process of obtaining the location coordinates of the user terminal and the target vehicle based on the geomagnetic fingerprint of the current location of the user terminal and the geomagnetic fingerprint of the location of the target vehicle, using a pre-trained positioning model, includes: The geomagnetic fingerprint of the current location of the user terminal, the geomagnetic fingerprint of the location of the target vehicle, and the geomagnetic fingerprint database of the target area are input into the pre-trained positioning model to obtain the location coordinates of the user terminal and the location coordinates of the target vehicle.

4. The method according to claim 3, characterized in that, The geomagnetic fingerprint database of the target area includes the geomagnetic fingerprint and known location coordinates of each relay node in the group cooperative network of the target area; obtaining the geomagnetic fingerprint database of the target area includes: A vehicle-finding assistance request is broadcast in the target area via a short-range communication protocol; Based on the received collaborative response messages, determine the relay nodes used to build the group collaborative network; Each relay node is allocated a communication time slot based on a time division multiple access protocol; The geomagnetic fingerprint and known location coordinates of the corresponding relay node are received in different communication time slots.

5. The method according to claim 4, characterized in that, The allocation of communication time slots for each relay node based on the time division multiple access protocol includes: Based on the time division multiple access protocol, communication time slots are allocated to each relay node according to the distance between the user terminal and each relay node, with the communication time slots of relay nodes that are closer to the user terminal being allocated earlier.

6. The method according to claim 1, characterized in that, The step of determining the navigation path based on the location coordinates of the user terminal, the location coordinates of the target vehicle, and real-time collected environmental perception data includes: The location coordinates of the user terminal, the location coordinates of the target vehicle, and the real-time collected environmental perception data are used as state input vectors and input into a pre-trained near-end policy optimization network to obtain the navigation path output by the near-end policy optimization network. The reward function of the near-end policy optimization network targets at least one of the following objectives: Minimize the time it takes for the user terminal to reach the target vehicle; Minimize the risk of collision with dynamic obstacles; Minimize the energy cost of the communication link; Maximize the approach progress to the target vehicle.

7. The method according to claim 1, characterized in that, The method further includes: The rendered output is superimposed on the real-world image captured by the image acquisition device on the user terminal, representing the navigation path.

8. The method according to any one of claims 1 to 7, characterized in that, The method further includes: If the distance to the target vehicle is detected to have reached a target threshold, a prompt message is output to alert the user that they are about to arrive at the location of the target vehicle.

9. A vehicle finding device, characterized in that, The device, applied to a user terminal, includes: The request receiving module is used to receive vehicle search requests for the target vehicle. The data acquisition module is used to acquire the geomagnetic fingerprint of the location of the target vehicle; The positioning module is used to obtain the position coordinates of the user terminal and the position coordinates of the target vehicle based on the geomagnetic fingerprint of the current location of the user terminal and the geomagnetic fingerprint of the location of the target vehicle, using a pre-trained positioning model. The path determination module is used to determine the navigation path based on the location coordinates of the user terminal, the location coordinates of the target vehicle, and real-time collected environmental perception data.

10. An electronic device, characterized in that, It includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the vehicle locating method as described in any one of claims 1 to 8.

11. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the vehicle-finding method as described in any one of claims 1 to 8.

Images