Digital key terminal calibration method and apparatus, storage medium, electronic device

Abstract

This invention provides a calibration method and apparatus, storage medium, and electronic device for a digital key terminal. The method includes: determining a target digital key terminal and a target vehicle to be calibrated; acquiring the actual vehicle Bluetooth signal characteristics of the target digital key terminal within a calibration cavity, wherein the sealed outer shell structure of the calibration cavity is made of electromagnetic shielding material; invoking a pre-trained honey badger algorithm backpropagation neural network (HBA-BPNN) hybrid neural network model, wherein the HBA-BPNN hybrid neural network model is used to output calibration parameters for the target vehicle based on the Bluetooth signal characteristics of the digital key terminal within the calibration cavity; and calculating the actual vehicle calibration parameters of the target digital key terminal for the target vehicle using the actual vehicle Bluetooth signal characteristics and the HBA-BPNN hybrid neural network model. This embodiment solves the technical problem of low efficiency in calibrating Bluetooth signals for digital keys in the prior art.

Description

Technical Field

[0001] This invention relates to the field of vehicle technology, and more specifically, to a calibration method and apparatus for a digital key terminal, a storage medium, and an electronic device. Background Technology

[0002] In related technologies, with the rapid development of smart cars, digital keys replacing traditional physical keys has become an inevitable trend. Currently, the mainstream digital key solution on the market is the digital key itself, which receives signals from the mobile phone through an anchor point installed in the car, identifies the position of the mobile phone relative to the vehicle based on the received signal strength (RSSI), and thus performs automatic unlocking or locking functions on the vehicle.

[0003] However, significant differences exist in the Bluetooth transmission power, antenna design, and internal structure of different brands and models of mobile phones, resulting in substantial variations in Bluetooth RSSI at the same distance. This heterogeneity of mobile phone Bluetooth signals means that, even with the same positioning parameters, the distance thresholds for the positioning areas (welcome area, unlock area, locking area, and in-vehicle area) output by different phones can vary greatly, severely impacting the continuity and consistency of the user experience. For example, a user of brand A mobile phone can unlock the car at 5 meters, while a user of brand B mobile phone may need to walk to 2 meters to trigger unlocking.

[0004] To address this issue, existing technical solutions primarily employ real-vehicle calibration: for different phone models, calibration data is collected one by one in a real vehicle environment, normalized parameters are generated and uploaded to the vehicle manufacturer's backend. When a user uses the digital key, the vehicle downloads the corresponding parameters based on the phone model for positioning. This solution has the following inherent drawbacks: 1. Huge calibration workload: There are numerous phone models on the market, with dozens of new models released every year. Calibrating each phone individually in a real vehicle would consume significant manpower and time. 2. Inaccurate parameters due to differences in calibration environment: During real-vehicle calibration, factors such as ambient temperature, vehicle parking posture, and surrounding electromagnetic interference are difficult to completely replicate across different batches of calibration, leading to parameter deviations for the same phone in different batches, thus affecting positioning consistency. 3. Inability to adapt to new models: After a new phone model is released, a service gap exists because a period of waiting for the manufacturer to complete calibration is required to provide a good experience.

[0005] No efficient and accurate solution has yet been found to address the aforementioned issues in the relevant technologies. Summary of the Invention

[0006] This invention provides a calibration method and apparatus for a digital key terminal, a storage medium, and an electronic device to solve technical problems in related technologies.

[0007] According to an embodiment of the present invention, a calibration method for a digital key terminal is provided, comprising: determining a target digital key terminal to be calibrated and a target vehicle; acquiring the actual vehicle Bluetooth signal characteristics of the target digital key terminal in a calibration cavity, wherein the sealed shell structure of the calibration cavity is made of electromagnetic shielding material; calling a pre-trained honey badger algorithm backpropagation neural network (HBA-BPNN) hybrid neural network model, wherein the HBA-BPNN hybrid neural network model is used to output calibration parameters for the target vehicle based on the Bluetooth signal characteristics of the digital key terminal in the calibration cavity; and calculating the actual vehicle calibration parameters of the target digital key terminal for the target vehicle using the actual vehicle Bluetooth signal characteristics and the HBA-BPNN hybrid neural network model.

[0008] Optionally, obtaining the actual vehicle Bluetooth signal characteristics of the target digital key terminal in the calibration cavity includes: fixing the target digital key terminal at the center position of the calibration cavity; after the Bluetooth module of the target digital key terminal is started, acquiring the received signal strength (RSSI) collected by multiple acquisition anchor points in multiple directions to obtain multi-directional RSSI data, wherein each acquisition anchor point is arranged in one direction inside the calibration cavity, and each acquisition anchor point collects one channel of RSSI data; preprocessing the multi-directional RSSI data to obtain the actual vehicle Bluetooth signal characteristics of the target digital key terminal in the calibration cavity.

[0009] Optionally, before calling the pre-trained HBA-BPNN hybrid neural network model, the method further includes: identifying multiple sample digital key terminals; obtaining sample calibration parameters of the multiple sample digital key terminals for the target vehicle, and obtaining sample Bluetooth signal features of the multiple sample digital key terminals in the calibration cavity; training an initial network model using the sample calibration parameters and the sample Bluetooth signal features to obtain the HBA-BPNN hybrid neural network model.

[0010] Optionally, obtaining the sample calibration parameters of the multiple sample digital key terminals for the target vehicle includes: setting multiple calibration points for the target vehicle in an ideal real-world environment for each of the multiple sample digital key terminals, wherein each calibration point corresponds to a calibration orientation relative to the target vehicle; obtaining the sample signal strengths emitted by the sample digital key terminals collected by the vehicle anchor points of the target vehicle, and fitting them to obtain the sample calibration parameters, wherein the sample calibration parameters include the vehicle anchor point gain and the reference distance signal strength.

[0011] Optionally, the HBA-BPNN hybrid neural network model is trained using the sample calibration parameters and the sample Bluetooth signal features, including: for each sample digital key terminal, training data is generated by pairing the sample calibration parameters and the sample Bluetooth signal features to obtain a training dataset; the initial network model is initially optimized using the honey badger algorithm with the training dataset to obtain the initial parameters of the initial network model; and the initial network model is finely tuned using the backpropagation algorithm to obtain the HBA-BPNN hybrid neural network model.

[0012] Optionally, the initial optimization of the initial network model using the honey badger algorithm on the training dataset includes: splitting the training dataset into a training set, a validation set, and a test set; initializing the population by encoding all parameter terms of the initial network model as individual position vectors of HBA, where each individual represents a set of initial weights and initial thresholds, wherein the parameter terms include weights and thresholds; using the training set to perform global optimization of the parameter terms using the honey badger algorithm; using the validation set to calculate the prediction error of the fitness function and using the prediction error as the evaluation criterion to verify the optimization results; searching for the optimal initial value in the parameter space using the honey badger algorithm's mining mode and honey mode; and after reaching a preset condition or termination condition, outputting the optimal individual as the initial parameters of the BP network and using the initial parameters to perform initial optimization of the initial network model.

[0013] Optionally, the initial network model is fine-tuned using the backpropagation algorithm to obtain the HBA-BPNN hybrid neural network model, including: setting training hyperparameters, using the initial parameters as the starting point of the BP network, and calculating the predicted value using the sample Bluetooth signal features in the training dataset through forward propagation; calculating the loss value between the calibration parameters of paired samples in the training dataset and the predicted value using a loss function; calculating the gradient based on the loss value through backpropagation, and updating the weights of the initial network model using an optimizer to obtain an updated network model; verifying the model accuracy of the updated network model using the test set; if the model accuracy meets the preset accuracy requirement, terminating the training, and outputting the updated network model as the HBA-BPNN hybrid neural network model.

[0014] Optionally, calculating the vehicle calibration parameters of the target digital key terminal for the target vehicle using the actual vehicle Bluetooth signal features and the HBA-BPNN hybrid neural network model includes: normalizing the actual vehicle Bluetooth signal features to obtain model input data; inputting the model input data into the HBA-BPNN hybrid neural network model for forward propagation calculation to output layer normalized intermediate calibration parameters; denormalizing the intermediate calibration parameters to restore them to actual parameter values, and outputting the actual parameter values ​​as the vehicle calibration parameters of the target digital key terminal for the target vehicle.

[0015] Optionally, after calculating the vehicle calibration parameters of the target digital key terminal for the target vehicle using the real vehicle Bluetooth signal characteristics and the HBA-BPNN hybrid neural network model, the method further includes: associating the vehicle calibration parameters and the terminal identifier of the target digital key terminal and sending them to the target vehicle; collecting the real-time RSSI of the target digital key terminal through the vehicle anchor point of the target vehicle; determining the real-time area of ​​the target vehicle where the target digital key terminal is currently located using the vehicle calibration parameters and the real-time RSSI; and controlling the target vehicle to generate a vehicle control signal matching the real-time area.

[0016] According to another embodiment of the present invention, a calibration device for a digital key terminal is provided, comprising: a first determining module for determining a target digital key terminal to be calibrated and a target vehicle; a first acquiring module for acquiring the actual vehicle Bluetooth signal characteristics of the target digital key terminal in a calibration cavity, wherein the sealed shell structure of the calibration cavity is made of electromagnetic shielding material; a calling module for calling a pre-trained honey badger algorithm backpropagation neural network (HBA-BPNN) hybrid neural network model, wherein the HBA-BPNN hybrid neural network model is used to output calibration parameters for the target vehicle based on the Bluetooth signal characteristics of the digital key terminal in the calibration cavity; and a calculation module for calculating the actual vehicle calibration parameters of the target digital key terminal for the target vehicle using the actual vehicle Bluetooth signal characteristics and the HBA-BPNN hybrid neural network model.

[0017] Optionally, the first acquisition module includes: a fixing unit for fixing the target digital key terminal at the center position of the calibration cavity; an acquisition unit for acquiring received signal strength (RSSI) data from multiple acquisition anchor points at multiple directions after the Bluetooth module of the target digital key terminal is activated, thereby obtaining multi-directional RSSI data, wherein each acquisition anchor point is arranged at a position inside the calibration cavity, and each acquisition anchor point acquires one channel of RSSI data; and a processing unit for preprocessing the multi-directional RSSI data to obtain the actual vehicle Bluetooth signal characteristics of the target digital key terminal in the calibration cavity.

[0018] Optionally, the device further includes: a second determining module, configured to determine multiple sample digital key terminals before calling the pre-trained HBA-BPNN hybrid neural network model; a second acquiring module, configured to acquire the sample calibration parameters of the multiple sample digital key terminals for the target vehicle, and acquire the sample Bluetooth signal features of the multiple sample digital key terminals in the calibration cavity; and a training module, configured to train an initial network model using the sample calibration parameters and the sample Bluetooth signal features to obtain the HBA-BPNN hybrid neural network model.

[0019] Optionally, the second acquisition module includes: a setting unit, configured to set multiple calibration points of the target vehicle in an ideal real-world environment for each of the multiple sample digital key terminals, wherein each calibration point corresponds to a calibration orientation relative to the target vehicle; and a fitting unit, configured to acquire the sample signal strengths emitted by the sample digital key terminals collected by the vehicle anchor points of the target vehicle, and fit them to obtain sample calibration parameters, wherein the sample calibration parameters include vehicle anchor point gain and reference distance signal strength.

[0020] Optionally, the training module includes: a generation unit, used to generate training data for each sample digital key terminal by pairing the sample calibration parameters and the sample Bluetooth signal features to obtain a training dataset; an optimization unit, used to perform initial optimization of the initial network model using the training dataset and the honey badger algorithm to obtain the initial parameters of the initial network model; and a tuning unit, used to perform fine tuning of the initial network model using the backpropagation algorithm to obtain the HBA-BPNN hybrid neural network model.

[0021] Optionally, the optimization unit includes: a splitting subunit for splitting the training dataset into a training set, a validation set, and a test set; an encoding subunit for initializing the population, encoding all parameter terms of the initial network model into individual position vectors of HBA, where each individual represents a set of initial weights and initial thresholds, wherein the parameter terms include weights and thresholds; an optimization subunit for globally optimizing the parameter terms using the training set and the honey badger algorithm, calculating the prediction error of the fitness function using the validation set, and validating the optimization results using the prediction error as an evaluation criterion; a search subunit for searching for optimal initial values ​​in the parameter space using the honey badger algorithm's mining mode and honey mode; and an output subunit for outputting the optimal individual as the initial parameters of the BP network after reaching a preset condition or termination condition, and using the initial parameters to perform initial optimization of the initial network model.

[0022] Optionally, the tuning unit includes: a first calculation subunit, used to set training hyperparameters, using the initial parameters as the starting point of the BP network, and using the sample Bluetooth signal features in the training dataset for forward propagation to calculate the predicted value; a second calculation subunit, used to use a loss function to calculate the loss value between the paired sample calibration parameters in the training dataset and the predicted value; an update subunit, used to backpropagate the gradient based on the loss value, and use an optimizer to update the weights of the initial network model to obtain an updated network model; a verification subunit, used to use the test set to verify the model accuracy of the updated network model; and an output subunit, used to terminate training if the model accuracy meets a preset accuracy requirement, and output the updated network model as the HBA-BPNN hybrid neural network model.

[0023] Optionally, the calculation module includes: a processing unit for normalizing the features of the real vehicle Bluetooth signal to obtain model input data; a first output unit for inputting the model input data into the HBA-BPNN hybrid neural network model for forward propagation calculation and outputting intermediate calibration parameters with normalized layers; and a second output unit for denormalizing the intermediate calibration parameters to restore them to actual parameter values ​​and outputting the actual parameter values ​​as the real vehicle calibration parameters of the target digital key terminal for the target vehicle.

[0024] Optionally, the device further includes: a sending module, configured to send the vehicle calibration parameters and the terminal identifier of the target digital key terminal to the target vehicle after the calculation module calculates the vehicle calibration parameters of the target digital key terminal for the target vehicle using the real vehicle Bluetooth signal features and the HBA-BPNN hybrid neural network model; a collection module, configured to collect the real-time RSSI of the target digital key terminal through the vehicle anchor point of the target vehicle; a determination module, configured to determine the real-time area of ​​the target vehicle where the target digital key terminal is currently located using the vehicle calibration parameters and the real-time RSSI; and a generation module, configured to control the target vehicle to generate a vehicle control signal matching the real-time area.

[0025] According to another aspect of the embodiments of this application, a storage medium is also provided, the storage medium including a stored program that executes the above steps when the program is run.

[0026] According to another aspect of the embodiments of this application, an electronic device is also provided, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; wherein: the memory is used to store computer programs; and the processor is used to execute the steps in the above method by running the programs stored in the memory.

[0027] This application also provides a computer program product containing instructions that, when run on a computer, cause the computer to perform the steps in the above-described method.

[0028] The beneficial effects of this invention are: 1. Significantly reduces calibration workload: Traditional solutions require on-vehicle calibration for each mobile phone model; this invention only requires one on-vehicle calibration for each of the 20 basic mobile phones in each vehicle model, and the device can then complete the calibration of all other mobile phones. Specifically, the implementation involves using the signal radiation data of the 20 basic mobile phones and calibration parameters that achieve ideal positioning output in on-vehicle testing. An HBA-BPNN hybrid neural network model is used to train a calibration parameter fitting algorithm model. Finally, the signal radiation data of the mobile phone to be calibrated is used to obtain the calibration parameters through this fitting model.

[0029] 2. Eliminate environmental interference: The actual vehicle calibration only needs to be performed once on 20 basic mobile phones, which can be carried out in the optimal environment; all subsequent mobile phone data are collected through standardized devices, avoiding the impact of environmental differences on the calibration results.

[0030] 3. AI-driven precise mapping: Using the HBA-BPNN hybrid neural network model, with paired data from 20 basic mobile phones as training samples, it learns the nonlinear mapping relationship from device RSSI features to real vehicle parameters, and has good generalization ability.

[0031] 3. Six-dimensional omnidirectional signal acquisition: Six acquisition anchor points are set up to capture the spatial distribution characteristics of mobile phone Bluetooth signals from all directions, providing rich input features for AI models.

[0032] 4. Excellent scalability: After a new mobile phone model is launched, there is no need for real vehicle calibration. Calibration can be completed by collecting data once in the device. For new car models, only 20 basic mobile phones need to be calibrated to establish a mapping model. Attached Figure Description

[0033] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings: Figure 1 This is a hardware structure block diagram of a computer according to an embodiment of the present invention; Figure 2 This is a flowchart of a calibration method for a digital key terminal according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the calibration cavity in an embodiment of the present invention; Figure 4 This is a system architecture diagram according to an embodiment of the present invention; Figure 5 This is a structural block diagram of a calibration device for a digital key terminal according to an embodiment of the present invention. Detailed Implementation

[0034] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, and not all of them. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present application. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of the present application can be combined with each other.

[0035] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0036] Example 1 The method embodiment provided in Embodiment 1 of this application can be executed in a vehicle, vehicle controller, processor, computer, or similar processing device. Taking running on a computer as an example, Figure 1 This is a hardware structure block diagram of a computer according to an embodiment of the present invention. For example... Figure 1 As shown, a computer may include one or more ( Figure 1 Only one is shown in the diagram. A processor 102 (which may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.) and a memory 104 for storing data are also shown. Optionally, the computer may further include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the computer described above. For example, the computer may also include components that are larger than... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0037] The memory 104 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the calibration method of a digital key terminal of a computer in this embodiment of the invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, thereby implementing the above-described method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0038] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by a computer's communication provider. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module used for wireless communication with the Internet.

[0039] This embodiment provides a calibration method for a digital key terminal. Figure 2 This is a flowchart of a digital key terminal calibration method according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps: Step S201: Determine the target digital key terminal and target vehicle to be calibrated; Optionally, the target digital key terminal can be a mobile phone or other mobile device with a Bluetooth transmitter module.

[0040] Step S202: Obtain the actual vehicle Bluetooth signal characteristics of the target digital key terminal in the calibration cavity, wherein the sealed shell structure of the calibration cavity is made of electromagnetic shielding material; Optionally, the electromagnetic shielding material can be metal.

[0041] Step S203: Invoke the pre-trained honey badger algorithm backpropagation neural network HBA-BPNN hybrid neural network model, wherein the HBA-BPNN hybrid neural network model is used to output the calibration parameters of the target vehicle through the Bluetooth signal feature output in the calibration cavity by the digital key terminal; The HBA-BPNN hybrid neural network model in this embodiment combines the capabilities of the HBA (Honey Badger Algorithm) neural network model and the BPNN (Back Propagation Neural Network) hybrid neural network model.

[0042] Step S204: Calculate the vehicle calibration parameters of the target digital key terminal for the target vehicle using the actual vehicle Bluetooth signal characteristics and the HBA-BPNN hybrid neural network model.

[0043] The vehicle calibration parameters in this embodiment are used to assist the target vehicle in determining the user's specific area, such as the welcome area, unlocking area, locking area, and interior area, based on the real-time Bluetooth signal of the target digital key terminal.

[0044] Through the above steps, the target digital key terminal and the target vehicle to be calibrated are determined; the actual vehicle Bluetooth signal characteristics of the target digital key terminal in the calibration cavity are obtained, wherein the sealed shell structure of the calibration cavity is made of electromagnetic shielding material; a pre-trained honey badger algorithm backpropagation neural network (HBA-BPNN) hybrid neural network model is invoked, wherein the HBA-BPNN hybrid neural network model is used to output calibration parameters for the target vehicle based on the Bluetooth signal characteristics of the digital key terminal in the calibration cavity; the actual vehicle Bluetooth signal characteristics and the HBA-BPNN hybrid neural network model are used to calculate the actual vehicle calibration parameters of the target digital key terminal for the target vehicle, thereby solving the technical problem of low efficiency in calibrating Bluetooth signals of digital keys in the prior art, realizing fast, accurate, and universal mobile phone Bluetooth signal calibration, and improving the consistency of user experience for digital keys across mobile phone models.

[0045] In one embodiment of this example, obtaining the actual vehicle Bluetooth signal characteristics of the target digital key terminal in the calibration cavity includes: fixing the target digital key terminal at the center position of the calibration cavity; after the Bluetooth module of the target digital key terminal is activated, acquiring the received signal strength (RSSI) collected by multiple acquisition anchor points at multiple directions to obtain multi-directional RSSI data, wherein each acquisition anchor point is arranged at a direction inside the calibration cavity, and each acquisition anchor point collects one channel of RSSI data; preprocessing the multi-directional RSSI data to obtain the actual vehicle Bluetooth signal characteristics of the target digital key terminal in the calibration cavity.

[0046] The calibration cavity in this embodiment is an electromagnetically shielded calibration cavity. It is a sealed cavity structure made of electromagnetic shielding material, used to isolate external electromagnetic interference and provide a standardized electromagnetic environment for signal acquisition.

[0047] Figure 3 This is a schematic diagram of the calibration cavity in an embodiment of the present invention, including a mobile phone carrier platform, six Bluetooth signal acquisition anchor points, a central control unit (not shown), and a data processing unit (not shown). The mobile phone carrier platform is located in the center of the calibration cavity and is used to fix the mobile phone to be calibrated. The six Bluetooth signal acquisition anchor points are respectively arranged in the front, back, left, right, top, and bottom positions of the calibration cavity, with a distance of 3 to 6 meters between the acquisition anchor points and the mobile phone. Each acquisition anchor point includes a Bluetooth signal receiving module for synchronously acquiring the RSSI value of the Bluetooth broadcast signal emitted by the mobile phone to be calibrated. The central control unit is used to control the start and stop of the acquisition process and receive the RSSI data reported by each acquisition anchor point. The data processing unit is used to preprocess the acquired multi-directional RSSI data.

[0048] In this embodiment, before calling the pre-trained HBA-BPNN hybrid neural network model, the method further includes: determining multiple sample digital key terminals; obtaining the sample calibration parameters of the multiple sample digital key terminals for the target vehicle, and obtaining the sample Bluetooth signal characteristics of the multiple sample digital key terminals in the calibration cavity; training an initial network model using the sample calibration parameters and the sample Bluetooth signal characteristics to obtain the HBA-BPNN hybrid neural network model.

[0049] Optionally, the sample digital key terminal is a base mobile phone. The base mobile phone is fixed in the center of the mobile phone carrier platform of the calibration cavity, and RSSI signals are synchronously collected at six anchor points. After data preprocessing, a six-dimensional RSSI feature vector is constructed, and the device calibration parameter P_dev_i (six-dimensional RSSI feature value) of each base mobile phone is recorded as the sample Bluetooth signal feature.

[0050] In one example, obtaining the sample calibration parameters of the multiple sample digital key terminals for the target vehicle includes: setting multiple calibration points for the target vehicle in an ideal real-world environment for each of the multiple sample digital key terminals, wherein each calibration point corresponds to a calibration orientation relative to the target vehicle; obtaining the sample signal strengths emitted by the sample digital key terminals collected by the vehicle anchor points of the target vehicle, and fitting them to obtain the sample calibration parameters, wherein the sample calibration parameters include the vehicle anchor point gain and the reference distance signal strength.

[0051] Taking 20 sample digital key terminals as an example, standardized real-vehicle calibration was performed on each of the 20 selected basic mobile phones in a real vehicle environment. Calibration site requirements: open and flat area, free of large metallic reflective objects, and a clean electromagnetic environment. Calibration point settings: multiple reference distance points (1 meter, 3 meters, 5 meters, 8 meters, and 12 meters) were set in four directions: in front, behind, left, and right of the vehicle. During calibration: the basic mobile phones were placed sequentially at each calibration point, and the mobile phone signal strength collected at each anchor point on the vehicle was recorded. The reference distance signal strength was obtained through parameter fitting. The vehicle anchor point gain was adjusted so that the positioning algorithm output an ideal positioning area during real-vehicle testing, thus obtaining the optimal vehicle anchor point gain coefficient.

[0052] In one example, the HBA-BPNN hybrid neural network model is trained using the sample calibration parameters and the sample Bluetooth signal features. This process includes: generating training data for each sample digital key terminal by pairing the sample calibration parameters and the sample Bluetooth signal features to obtain a training dataset; using the training dataset to perform initial optimization on the initial network model using the honey badger algorithm to obtain the initial parameters of the initial network model; and using the backpropagation algorithm to fine-tune the initial network model to obtain the HBA-BPNN hybrid neural network model.

[0053] The HBA-BPNN hybrid neural network model was trained using pairing data (sample calibration parameters + sample Bluetooth signal features) from multiple sample digital key terminals as the training dataset.

[0054] Optionally, the initial optimization of the initial network model using the honey badger algorithm on the training dataset includes: splitting the training dataset into a training set, a validation set, and a test set; initializing the population by encoding all parameter terms of the initial network model as individual position vectors of HBA, where each individual represents a set of initial weights and initial thresholds, wherein the parameter terms include weights and thresholds; using the training set to perform global optimization of the parameter terms using the honey badger algorithm; using the validation set to calculate the prediction error of the fitness function and using the prediction error as the evaluation criterion to verify the optimization results; searching for the optimal initial value in the parameter space using the honey badger algorithm's mining mode and honey mode; and after reaching a preset condition or termination condition, outputting the optimal individual as the initial parameters of the BP network and using the initial parameters to perform initial optimization of the initial network model.

[0055] First, data preprocessing is performed. The collected six-dimensional RSSI data is normalized to transform the values ​​of each dimension to a uniform scale range, thereby eliminating the impact of magnitude differences between different anchor points on training.

[0056] Then, the honey badger algorithm is used for initial optimization. The honey badger algorithm is used to globally optimize the initial weights and thresholds of the BP neural network: the population is initialized, and each individual represents a set of initial weights and thresholds of the network; the fitness function (prediction error on the validation set) is used as the evaluation criterion; the foraging behavior of honey badgers (digging mode and honey mode) is simulated to search for the optimal initial values ​​in the parameter space; after the preset conditions are met, the optimal individual is output as the initial parameters of the BP network.

[0057] Optionally, the initial network model is fine-tuned using the backpropagation algorithm to obtain the HBA-BPNN hybrid neural network model, including: setting training hyperparameters, using the initial parameters as the starting point of the BP network, and calculating the predicted value using the sample Bluetooth signal features in the training dataset through forward propagation; calculating the loss value between the calibration parameters of paired samples in the training dataset and the predicted value using a loss function; calculating the gradient based on the loss value through backpropagation, and updating the weights of the initial network model using an optimizer to obtain an updated network model; verifying the model accuracy of the updated network model using the test set; if the model accuracy meets the preset accuracy requirement, terminating the training, and outputting the updated network model as the HBA-BPNN hybrid neural network model.

[0058] Then, a backpropagation neural network is used for fine-tuning. The optimal initial parameters obtained by the honey badger algorithm are used as the starting point of the backpropagation network. The backpropagation algorithm is used for fine-tuning: small batches of data are used for step-by-step iteration, and the model performance is monitored through the validation set. When the validation set error no longer decreases for several consecutive rounds, the training is terminated in advance to prevent overfitting.

[0059] In one embodiment of this example, calculating the vehicle calibration parameters of the target digital key terminal for the target vehicle using the actual vehicle Bluetooth signal features and the HBA-BPNN hybrid neural network model includes: normalizing the actual vehicle Bluetooth signal features to obtain model input data; inputting the model input data into the HBA-BPNN hybrid neural network model for forward propagation calculation to output layer normalized intermediate calibration parameters; denormalizing the intermediate calibration parameters to restore them to actual parameter values, and outputting the actual parameter values ​​as the vehicle calibration parameters of the target digital key terminal for the target vehicle.

[0060] Place any mobile phone to be calibrated into the calibration device and perform six-directional synchronous acquisition to obtain the device calibration parameters P_dev_target of the mobile phone; input the device parameters into the trained HBA-BPNN hybrid neural network model, and output the actual vehicle calibration parameters of the mobile phone to be calibrated through model inference calculation.

[0061] After the HBA-BPNN hybrid neural network model is trained, the calibration parameters are generated for a new mobile phone to be calibrated as follows: the mobile phone to be calibrated is placed in the calibration device and a six-dimensional RSSI feature vector is collected; the feature vector is normalized in the same way as the training data; the processed features are fed into the model for forward propagation calculation; the normalized calibration parameters are obtained from the output layer and denormalized to restore the actual parameter values; the final calibration parameters are output.

[0062] In one implementation scenario of this embodiment, after calculating the vehicle calibration parameters of the target digital key terminal for the target vehicle using the real vehicle Bluetooth signal characteristics and the HBA-BPNN hybrid neural network model, the method further includes: associating the vehicle calibration parameters and the terminal identifier of the target digital key terminal and sending them to the target vehicle; collecting the real-time RSSI of the target digital key terminal through the vehicle anchor point of the target vehicle; determining the real-time area of ​​the target vehicle where the target digital key terminal is currently located using the vehicle calibration parameters and the real-time RSSI; and controlling the target vehicle to generate a vehicle control signal matching the real-time area.

[0063] The purpose of this invention is to provide a self-testing device and a calibration parameter generation method based on an HBA-BPNN hybrid neural network model. The core of this invention lies in achieving efficient conversion from device parameters to actual vehicle parameters through a three-stage method: "basic mobile phone group real-vehicle calibration + device acquisition deviation coefficient + HBA-BPNN hybrid neural network model fitting." This aims to solve the problems of high workload and inaccurate calibration parameters due to environmental factors in existing mobile phone Bluetooth calibration technologies, achieving fast, accurate, and universal mobile phone Bluetooth signal calibration, and improving the consistency of user experience across different mobile phone models for digital keys.

[0064] Figure 4 The system architecture diagram of this embodiment includes: a calibration device (data acquisition device), a data processing unit, an HBA-BPNN hybrid neural network model (data fusion model), and a vehicle digital key system.

[0065] The data acquisition device includes: a calibration shielding chamber: a sealed cavity structure made of electromagnetic shielding material, used to isolate external electromagnetic interference and provide a standardized electromagnetic environment for signal acquisition; a mobile phone carrier platform: set in the center of the calibration shielding chamber, used to fix the mobile phone to be calibrated and maintain its consistent posture in space; and acquisition anchor points: arranged in six positions (front, back, left, right, top, and bottom) of the mobile phone to synchronously acquire the RSSI value of the Bluetooth signal emitted by the mobile phone to be calibrated. The distance d0 between the six acquisition points and the mobile phone is a fixed value to ensure that the path loss of the signal transmitted by the mobile phone in different directions is basically symmetrical, and finally the signal strength of the mobile phone in each direction is obtained.

[0066] Data processing unit: Used to control the start and stop of the acquisition process, receive RSSI data reported by each acquisition anchor point, and establish communication with the mobile phone to be calibrated to perform standardized transmission power configuration; at the same time, it preprocesses and extracts features from the acquired multi-directional RSSI data, and uses the HBA-BPNN hybrid neural network model to generate calibration parameters. Data fusion model: An HBA-BPNN hybrid neural network model is employed, with initial parameters globally optimized using the honey badger algorithm. This model is deployed on a cloud or local server. Based on collected multi-directional RSSI feature vectors and combined with historical training data, calibration parameters (including vehicle anchor point gain, reference distance signal strength, etc.) corresponding to the phone to be calibrated are generated. The generated calibration parameters are associated with the phone model identifier and stored in a cloud database. When a car owner with the same phone model uses the digital key, the vehicle obtains the corresponding calibration parameters and applies them to the positioning algorithm.

[0067] The vehicle digital key system uses the vehicle's Bluetooth anchor point to collect the real-time RSSI of the mobile phone, transmit it to the positioning algorithm module, calculate the distance to the mobile phone's real-time location, and then input it into the area determination model to determine the specific area, such as the welcome area, unlocking area, locking area, and interior area.

[0068] In this embodiment, multiple basic mobile phones are selected as samples for each vehicle model for both real-vehicle calibration and device calibration, resulting in paired datasets. A backpropagation neural network optimized using the honey badger algorithm is employed as the HBA-BPNN hybrid neural network model, and the paired data is used for model training. The mobile phone to be calibrated only needs to have its six-dimensional RSSI features collected in the device, and the real-vehicle calibration parameters can be generated through model inference. This invention effectively solves the problems of high workload and significant environmental interference in real-vehicle mobile phone calibration by combining basic mobile phone group calibration with the HBA-BPNN hybrid neural network model, achieving a consistent user experience across different mobile phone models.

[0069] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to 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 the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0070] Example 2 This embodiment also provides a calibration device for a digital key terminal, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0071] Figure 5 This is a structural block diagram of a calibration device for a digital key terminal according to an embodiment of the present invention, such as... Figure 5 As shown, the device includes: The first determining module 51 is used to determine the target digital key terminal to be calibrated and the target vehicle. The first acquisition module 52 is used to acquire the actual vehicle Bluetooth signal characteristics of the target digital key terminal in the calibration cavity, wherein the sealed shell structure of the calibration cavity is made of electromagnetic shielding material. Module 53 is invoked to invoke the pre-trained honey badger algorithm backpropagation neural network HBA-BPNN hybrid neural network model, wherein the HBA-BPNN hybrid neural network model is used to output calibration parameters of the target vehicle through the Bluetooth signal feature output in the calibration cavity by the digital key terminal; The calculation module 54 is used to calculate the vehicle calibration parameters of the target digital key terminal for the target vehicle using the characteristics of the real vehicle Bluetooth signal and the HBA-BPNN hybrid neural network model.

[0072] Optionally, the first acquisition module includes: a fixing unit for fixing the target digital key terminal at the center position of the calibration cavity; an acquisition unit for acquiring received signal strength (RSSI) data from multiple acquisition anchor points at multiple directions after the Bluetooth module of the target digital key terminal is activated, thereby obtaining multi-directional RSSI data, wherein each acquisition anchor point is arranged at a position inside the calibration cavity, and each acquisition anchor point acquires one channel of RSSI data; and a processing unit for preprocessing the multi-directional RSSI data to obtain the actual vehicle Bluetooth signal characteristics of the target digital key terminal in the calibration cavity.

[0073] Optionally, the device further includes: a second determining module, configured to determine multiple sample digital key terminals before calling the pre-trained HBA-BPNN hybrid neural network model; a second acquiring module, configured to acquire the sample calibration parameters of the multiple sample digital key terminals for the target vehicle, and acquire the sample Bluetooth signal features of the multiple sample digital key terminals in the calibration cavity; and a training module, configured to train an initial network model using the sample calibration parameters and the sample Bluetooth signal features to obtain the HBA-BPNN hybrid neural network model.

[0074] Optionally, the second acquisition module includes: a setting unit, configured to set multiple calibration points of the target vehicle in an ideal real-world environment for each of the multiple sample digital key terminals, wherein each calibration point corresponds to a calibration orientation relative to the target vehicle; and a fitting unit, configured to acquire the sample signal strengths emitted by the sample digital key terminals collected by the vehicle anchor points of the target vehicle, and fit them to obtain sample calibration parameters, wherein the sample calibration parameters include vehicle anchor point gain and reference distance signal strength.

[0075] Optionally, the training module includes: a generation unit, used to generate training data for each sample digital key terminal by pairing the sample calibration parameters and the sample Bluetooth signal features to obtain a training dataset; an optimization unit, used to perform initial optimization of the initial network model using the training dataset and the honey badger algorithm to obtain the initial parameters of the initial network model; and a tuning unit, used to perform fine tuning of the initial network model using the backpropagation algorithm to obtain the HBA-BPNN hybrid neural network model.

[0076] Optionally, the optimization unit includes: a splitting subunit for splitting the training dataset into a training set, a validation set, and a test set; an encoding subunit for initializing the population, encoding all parameter terms of the initial network model into individual position vectors of HBA, where each individual represents a set of initial weights and initial thresholds, wherein the parameter terms include weights and thresholds; an optimization subunit for globally optimizing the parameter terms using the training set and the honey badger algorithm, calculating the prediction error of the fitness function using the validation set, and validating the optimization results using the prediction error as an evaluation criterion; a search subunit for searching for optimal initial values ​​in the parameter space using the honey badger algorithm's mining mode and honey mode; and an output subunit for outputting the optimal individual as the initial parameters of the BP network after reaching a preset condition or termination condition, and using the initial parameters to perform initial optimization of the initial network model.

[0077] Optionally, the tuning unit includes: a first calculation subunit, used to set training hyperparameters, using the initial parameters as the starting point of the BP network, and using the sample Bluetooth signal features in the training dataset for forward propagation to calculate the predicted value; a second calculation subunit, used to use a loss function to calculate the loss value between the paired sample calibration parameters in the training dataset and the predicted value; an update subunit, used to backpropagate the gradient based on the loss value, and use an optimizer to update the weights of the initial network model to obtain an updated network model; a verification subunit, used to use the test set to verify the model accuracy of the updated network model; and an output subunit, used to terminate training if the model accuracy meets a preset accuracy requirement, and output the updated network model as the HBA-BPNN hybrid neural network model.

[0078] Optionally, the calculation module includes: a processing unit for normalizing the features of the real vehicle Bluetooth signal to obtain model input data; a first output unit for inputting the model input data into the HBA-BPNN hybrid neural network model for forward propagation calculation and outputting intermediate calibration parameters with normalized layers; and a second output unit for denormalizing the intermediate calibration parameters to restore them to actual parameter values ​​and outputting the actual parameter values ​​as the real vehicle calibration parameters of the target digital key terminal for the target vehicle.

[0079] Optionally, the device further includes: a sending module, configured to send the vehicle calibration parameters and the terminal identifier of the target digital key terminal to the target vehicle after the calculation module calculates the vehicle calibration parameters of the target digital key terminal for the target vehicle using the real vehicle Bluetooth signal features and the HBA-BPNN hybrid neural network model; a collection module, configured to collect the real-time RSSI of the target digital key terminal through the vehicle anchor point of the target vehicle; a determination module, configured to determine the real-time area of ​​the target vehicle where the target digital key terminal is currently located using the vehicle calibration parameters and the real-time RSSI; and a generation module, configured to control the target vehicle to generate a vehicle control signal matching the real-time area.

[0080] It should be noted that the above modules can be implemented by software or hardware. For the latter, they can be implemented in the following ways, but are not limited to: all the above modules are located in the same processor; or, the above modules are located in different processors in any combination.

[0081] Example 3 Embodiments of the present invention also provide a storage medium storing a computer program, wherein the computer program is configured to execute the steps in any of the above method embodiments when running.

[0082] Optionally, in this embodiment, the storage medium may be configured to store a computer program for performing the following steps: S1, determine the target digital key terminal and target vehicle to be calibrated; S2, acquire the actual vehicle Bluetooth signal characteristics of the target digital key terminal in the calibration cavity, wherein the sealed shell structure of the calibration cavity is made of electromagnetic shielding material; S3, invoke the pre-trained honey badger algorithm backpropagation neural network HBA-BPNN hybrid neural network model, wherein the HBA-BPNN hybrid neural network model is used to output the calibration parameters of the target vehicle through the Bluetooth signal feature output in the calibration cavity by the digital key terminal; S4, using the actual vehicle Bluetooth signal characteristics and the HBA-BPNN hybrid neural network model, calculate the actual vehicle calibration parameters of the target digital key terminal for the target vehicle.

[0083] Optionally, in this embodiment, the storage medium may include, but is not limited to, various media capable of storing computer programs, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0084] Embodiments of the present invention also provide an electronic device including a memory and a processor, the memory storing a computer program and the processor being configured to run the computer program to perform the steps in any of the above method embodiments.

[0085] Optionally, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor and the input / output device is connected to the processor.

[0086] Optionally, in this embodiment, the processor can be configured to perform the following steps via a computer program: S1, determine the target digital key terminal and target vehicle to be calibrated; S2, acquire the actual vehicle Bluetooth signal characteristics of the target digital key terminal in the calibration cavity, wherein the sealed shell structure of the calibration cavity is made of electromagnetic shielding material; S3, invoke the pre-trained honey badger algorithm backpropagation neural network HBA-BPNN hybrid neural network model, wherein the HBA-BPNN hybrid neural network model is used to output the calibration parameters of the target vehicle through the Bluetooth signal feature output in the calibration cavity by the digital key terminal; S4, using the actual vehicle Bluetooth signal characteristics and the HBA-BPNN hybrid neural network model, calculate the actual vehicle calibration parameters of the target digital key terminal for the target vehicle.

[0087] Optionally, specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementations, and will not be repeated here.

[0088] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0089] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented using software plus a general-purpose hardware platform, or of course, using hardware. Based on this understanding, the above technical solutions, in essence or the parts that contribute to the related technology, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0090] It should be understood that the terminology used herein is for the purpose of describing particular exemplary embodiments only and is not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms “a,” “an,” and “described” as used herein may also include the plural forms. The terms “comprising,” “including,” “containing,” and “having” are inclusive and therefore indicate the presence of the stated features, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, elements, components, and / or combinations thereof. The method steps, processes, and operations described herein are not construed as requiring them to be performed in a particular order described or illustrated unless the order of performance is explicitly indicated. It should also be understood that additional or alternative steps may be used.

[0091] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A calibration method for a digital key terminal, characterized in that, include: Identify the target digital key terminal and target vehicle to be calibrated; The actual vehicle Bluetooth signal characteristics of the target digital key terminal in the calibration cavity are obtained, wherein the sealed shell structure of the calibration cavity is made of electromagnetic shielding material; The pre-trained honey badger algorithm backpropagation neural network HBA-BPNN hybrid neural network model is invoked, wherein the HBA-BPNN hybrid neural network model is used to output the calibration parameters of the target vehicle through the Bluetooth signal feature output in the calibration cavity by the digital key terminal; The vehicle's Bluetooth signal characteristics and the HBA-BPNN hybrid neural network model are used to calculate the vehicle calibration parameters of the target digital key terminal for the target vehicle.

2. The method according to claim 1, characterized in that, The characteristics of the actual vehicle Bluetooth signal of the target digital key terminal in the calibration cavity include: The target digital key terminal is fixed at the center of the calibration cavity; After the Bluetooth module of the target digital key terminal is started, the received signal strength RSSI collected by multiple acquisition anchor points in multiple directions is obtained to obtain multi-directional RSSI data. Each acquisition anchor point is arranged in one direction inside the calibration cavity, and each acquisition anchor point collects one channel of RSSI data. The multi-directional RSSI data is preprocessed to obtain the actual vehicle Bluetooth signal characteristics of the target digital key terminal in the calibration cavity.

3. The method according to claim 1, characterized in that, The method further includes the following steps before invoking the pre-trained HBA-BPNN hybrid neural network model: Multiple sample digital key terminals were identified; The sample calibration parameters of the multiple sample digital key terminals for the target vehicle are obtained, and the sample Bluetooth signal characteristics of the multiple sample digital key terminals in the calibration cavity are obtained. The initial network model is trained using the sample calibration parameters and the sample Bluetooth signal features to obtain the HBA-BPNN hybrid neural network model.

4. The method according to claim 3, characterized in that, Obtaining the sample calibration parameters of the target vehicle from the multiple sample digital key terminals includes: For each of the multiple sample digital key terminals, multiple calibration points are set for the target vehicle in an ideal real-world environment, wherein each calibration point corresponds to a calibration orientation relative to the target vehicle. The sample signal strengths emitted by the sample digital key terminal are collected from the vehicle anchor points of the target vehicle, and the sample calibration parameters are fitted to obtain the sample calibration parameters, wherein the sample calibration parameters include the vehicle anchor point gain and the reference distance signal strength.

5. The method according to claim 3, characterized in that, The initial network model is trained using the sample calibration parameters and the sample Bluetooth signal features to obtain the HBA-BPNN hybrid neural network model, which includes: For each sample digital key terminal, training data is generated by pairing the sample calibration parameters and the sample Bluetooth signal features to obtain a training dataset. The initial network model is initially optimized using the honey badger algorithm on the training dataset to obtain the initial parameters of the initial network model; The initial network model was finely tuned using the backpropagation algorithm to obtain the HBA-BPNN hybrid neural network model.

6. The method according to claim 5, characterized in that, The initial optimization of the initial network model using the training dataset and the honey badger algorithm includes: The training dataset is split into a training set, a validation set, and a test set; Initialize the population by encoding all parameter terms of the initial network model into individual position vectors of HBA, where each individual represents a set of initial weights and initial thresholds, wherein the parameter terms include weights and thresholds; The training set is used to perform global optimization of the parameters using the honey badger algorithm. The validation set is used to calculate the prediction error of the fitness function, and the prediction error is used as the evaluation criterion to verify the optimization results. The optimal initial value is searched in the parameter space using the mining pattern and honey pattern of the honey badger algorithm; After the preset conditions or termination conditions are met, the optimal individual is output as the initial parameters of the BP network, and the initial parameters are used to perform initial optimization on the initial network model.

7. The method according to claim 6, characterized in that, The initial network model is finely tuned using the backpropagation algorithm to obtain the HBA-BPNN hybrid neural network model, which includes: Set the training hyperparameters, using the initial parameters as the starting point of the BP network, and calculate the predicted values ​​using the Bluetooth signal features of the samples in the training dataset through forward propagation: The loss function is used to calculate the loss value between the paired sample labeling parameters in the training dataset and the predicted value; The gradient is calculated by backpropagation based on the loss value, and the weights of the initial network model are updated using an optimizer to obtain the updated network model. The test set is used to verify the model accuracy of the updated network model; If the model accuracy meets the preset accuracy requirement, training is terminated, and the updated network model is output as the HBA-BPNN hybrid neural network model.

8. The method according to claim 1, characterized in that, The calculation of the target digital key terminal's vehicle calibration parameters for the target vehicle using the actual vehicle Bluetooth signal characteristics and the HBA-BPNN hybrid neural network model includes: The characteristics of the actual vehicle Bluetooth signal are normalized to obtain the model input data; The model input data is input into the HBA-BPNN hybrid neural network model for forward propagation calculation, and the intermediate calibration parameters of the output layer are normalized. The intermediate calibration parameters are denormalized to restore the actual parameter values, and the actual parameter values ​​are output as the actual vehicle calibration parameters of the target digital key terminal for the target vehicle.

9. The method according to claim 1, characterized in that, After calculating the vehicle calibration parameters of the target digital key terminal for the target vehicle using the actual vehicle Bluetooth signal characteristics and the HBA-BPNN hybrid neural network model, the method further includes: The vehicle calibration parameters and the terminal identifier of the target digital key terminal are associated and sent to the target vehicle; The real-time RSSI of the target digital key terminal is collected through the vehicle anchor point of the target vehicle; The real-time area of ​​the target vehicle where the target digital key terminal is currently located is determined using the actual vehicle calibration parameters and the real-time RSSI. The target vehicle is controlled to generate a vehicle control signal that matches the real-time area.

10. A calibration device for a digital key terminal, characterized in that, include: The first determining module is used to determine the target digital key terminal to be calibrated and the target vehicle. The first acquisition module is used to acquire the actual vehicle Bluetooth signal characteristics of the target digital key terminal in the calibration cavity, wherein the sealed shell structure of the calibration cavity is made of electromagnetic shielding material. The calling module is used to call the pre-trained honey badger algorithm backpropagation neural network HBA-BPNN hybrid neural network model, wherein the HBA-BPNN hybrid neural network model is used to output the calibration parameters of the target vehicle through the Bluetooth signal feature output in the calibration cavity by the digital key terminal; The calculation module is used to calculate the vehicle calibration parameters of the target digital key terminal for the target vehicle using the characteristics of the real vehicle Bluetooth signal and the HBA-BPNN hybrid neural network model.

11. A storage medium, characterized in that, The storage medium stores a computer program, wherein the computer program is configured to execute the method described in any one of claims 1 to 9 when it is run.

12. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to run the computer program to perform the method as described in any one of claims 1 to 9.

Images