Vehicle unlocking and locking method, device and equipment, computer readable storage medium and computer program product

Abstract

The invention provides a vehicle unlocking and locking method, device and equipment, a computer readable storage medium and a computer program product. The method comprises the steps that the vehicle unlocking and locking device obtains multiple groups of RSSIs and multiple groups of differential signals corresponding to the multiple groups of RSSIs; each group of RSSIs in the plurality of groups of RSSIs is a signal sent by the user equipment; determining an unlocking identification result or a locking identification result by using an unlocking AI model and an unlocking logic decision rule according to the multiple groups of RSSIs and the multiple groups of differential signals; wherein the unlocking and locking AI model is a model obtained by training a plurality of groups of sample RSSIs and a plurality of groups of sample differential signals corresponding to the plurality of groups of sample RSSIs when historical unlocking and locking are successful; and executing corresponding unlocking operation or locking operation according to the unlocking identification result or the locking identification result. According to the invention, the accuracy of unlocking and locking of the vehicle can be improved.

Description

Technical Field

[0001] This application relates to the field of automotive control technology, and in particular to a vehicle unlocking / locking method, apparatus, device, computer-readable storage medium, and computer program product. Background Technology

[0002] With the continuous advancement of automotive intelligence, more and more vehicles are adopting keyless entry functionality, providing users with the convenience of unlocking or locking their vehicles. Typically, a Bluetooth key is used to achieve this unlocking or locking function.

[0003] In related technologies, Bluetooth keys determine whether to unlock or lock based on logical rules. This involves setting a signal strength threshold and comparing the Received Signal Strength Indication (RSSI) received by the user's device from the vehicle's Bluetooth module with this threshold. The result of this comparison determines whether to unlock or lock. However, this unlocking / locking method is easily affected by environmental factors, device differences, and usage patterns, leading to insufficient accuracy in the determination of whether to unlock or lock. For example, if the vehicle is locked and the user's device is inside a bag or obstructed by the user's body, the RSSI signal strength received by the vehicle's Bluetooth module decreases, causing a deviation from the calibration data and resulting in unlocking failure. Similarly, if the user is near the vehicle and it is unlocked, but the user's device is obstructed by the user's body, the RSSI strength received by the Bluetooth module will decrease, leading to false locking. Ultimately, this reduces the accuracy of vehicle unlocking and locking. Summary of the Invention

[0004] This application provides a vehicle unlocking / locking method, apparatus, device, computer-readable storage medium, and computer program product, which can improve the accuracy of vehicle unlocking / locking.

[0005] The technical solution of this application embodiment is implemented as follows:

[0006] This application provides a method for unlocking and locking a vehicle, the method comprising:

[0007] Acquire multiple sets of RSSIs and multiple sets of differential signals corresponding to the multiple sets of RSSIs; each set of RSSIs is a signal emitted by the user equipment.

[0008] The unlocking or locking identification result is determined by using the unlocking AI model and unlocking logic judgment rules, based on the multiple sets of RSSI and the multiple sets of differential signals; wherein, the unlocking AI model is a model trained using multiple sets of sample RSSI and the multiple sets of sample differential signals corresponding to the multiple sets of sample RSSI when the unlocking was successful in the past.

[0009] Based on the unlocking or locking recognition result, perform the corresponding unlocking or locking operation.

[0010] In the above scheme, determining the locking identification result by utilizing the unlocking AI model and unlocking logic judgment rules, based on the multiple sets of RSSI and the multiple sets of differential signals, includes:

[0011] The first locking identification result is determined by using the locking AI model in the unlocking AI model and based on the multiple sets of RSSI and the multiple sets of differential signals.

[0012] The second locking identification result is determined based on the unlocking logic determination rules and the multiple sets of RSSIs.

[0013] The locking identification result is determined based on the first locking identification result and the second locking identification result.

[0014] In the above scheme, determining the second locking identification result based on the unlocking logic judgment rule and the multiple sets of RSSIs includes:

[0015] Obtain the locking logic determination threshold from the unlocking logic determination rule;

[0016] Determine the comparison results between the multiple sets of RSSI and the latching logic judgment threshold;

[0017] Based on the comparison results, the second locking identification result is determined.

[0018] In the above scheme, determining the locking identification result based on the first locking identification result and the second locking identification result includes:

[0019] If both the first and second locking identification results indicate locking, the locking is determined as the locking identification result.

[0020] In the above scheme, determining the lock as the lock identification result when both the first lock identification result and the second lock identification result identify a lock includes:

[0021] If the second locking identification result is locking, the first locking state machine is adjusted to the set state;

[0022] If the first locking identification result is locking, the second locking state machine is adjusted to the set state;

[0023] When both the first locking state machine and the second locking state machine are in the set state, the locking is taken as the locking identification result.

[0024] In the above scheme, after performing the corresponding locking operation based on the locking identification result, the method further includes:

[0025] When any door of the target vehicle is in the open state, the states of the first locking state machine and the second locking state machine are adjusted to the non-set state;

[0026] Alternatively, when the vehicle controller of the target vehicle switches from a dormant state to a wake-up state, the states of the first locking state machine and the second locking state machine are adjusted to a non-set state.

[0027] Alternatively, if the target vehicle completes the corresponding locking operation, the states of the first locking state machine and the second locking state machine are adjusted to a non-set state.

[0028] In the above scheme, the step of using the unlocking / locking AI model and unlocking / locking logic judgment rules, and determining the unlocking identification result or the locking identification result based on the multiple sets of RSSI and the multiple sets of differential signals, includes:

[0029] The multiple sets of RSSI and the multiple sets of differential signals are input into the unlocking AI model in the unlocking AI model to obtain the first unlocking recognition result;

[0030] The second unlocking identification result is determined based on the unlocking logic determination rules and the multiple sets of RSSIs;

[0031] If the first unlock recognition result is unlocked, the first unlock recognition result is determined as the unlock recognition result; or if the second unlock recognition result is unlocked, the second unlock recognition result is determined as the unlock recognition result.

[0032] In the above scheme, the method further includes:

[0033] The location area of ​​the user equipment is determined based on the multiple sets of RSSIs;

[0034] If the location area is not the rear area of ​​the target vehicle and the first unlock recognition result is unlocked, the first unlock recognition result is determined as the unlock recognition result; or if the location area is not the rear area of ​​the target vehicle and the second unlock recognition result is unlocked, the second unlock recognition result is determined as the unlock recognition result.

[0035] If the location area is the rear area of ​​the target vehicle, the second unlocking recognition result is determined as the unlocking recognition result.

[0036] In the above scheme, before determining the unlocking or locking identification result based on the multiple sets of RSSI and the multiple sets of differential signals using the unlocking / locking AI model and unlocking / locking logic judgment rules, the method further includes:

[0037] Acquire multiple sets of sample RSSIs and corresponding sample differential signals at historical unlocking times, as well as sample unlocking operation results corresponding to the multiple sets of sample RSSIs and sample differential signals;

[0038] The initial unlocking AI model is trained based on the multiple sets of sample RSSI, the multiple sets of sample differential signals, and the sample unlocking and locking operation results to obtain the unlocking AI model; the initial unlocking model in the initial unlocking AI model includes an initial convolutional neural network and an initial recurrent neural network; the initial locking model in the initial unlocking AI model includes the initial convolutional neural network and the initial recurrent neural network.

[0039] In the above scheme, the step of training an initial unlocking AI model based on the multiple sets of sample RSSI, the multiple sets of sample differential signals, and the sample unlocking operation results to obtain the unlocking AI model includes:

[0040] The multiple sets of sample RSSI and the multiple sets of sample difference signals are input into the initial unlocking AI model to obtain the output unlocking identification result;

[0041] Based on the output unlocking identification result and the sample unlocking operation result, determine the binary cross-entropy loss of the initial unlocking AI model;

[0042] If the binary cross-entropy loss is greater than or equal to a preset loss threshold, the initial unlocking AI model is trained again using the multiple sets of sample RSSI, the multiple sets of sample difference signals, and the sample unlocking operation results to obtain the trained model.

[0043] If the corresponding binary training cross-entropy loss of the trained model is less than or equal to the preset loss threshold, the trained model is used as the unlocking AI model.

[0044] In the above scheme, the method further includes:

[0045] If the communication link between the user equipment and the target vehicle is lost, perform the locking operation;

[0046] Alternatively, if the values ​​of the multiple sets of RSSI are less than or equal to a preset threshold, a latching operation is performed.

[0047] This application provides a vehicle unlocking / locking device, including:

[0048] The acquisition unit is used to acquire multiple sets of RSSIs and multiple sets of differential signals corresponding to the multiple sets of RSSIs; each set of RSSIs is a signal emitted by the user equipment.

[0049] The determining unit is used to determine the unlocking identification result or the locking identification result by using the unlocking and locking AI model and the unlocking and locking logic judgment rules, based on the multiple sets of RSSI and the multiple sets of differential signals; wherein, the unlocking and locking AI model is a model trained using multiple sets of sample RSSI and the multiple sets of sample differential signals corresponding to the multiple sets of sample RSSI when the unlocking and locking were successful in the past.

[0050] The execution unit is used to perform the corresponding unlocking operation or locking operation based on the unlocking identification result or the locking identification result.

[0051] In the above scheme, the determining unit is used to determine a first locking identification result by utilizing the locking AI model in the unlocking AI model and based on the multiple sets of RSSI and the multiple sets of differential signals; determine a second locking identification result based on the unlocking logic judgment rule and the multiple sets of RSSI; and determine the locking identification result based on the first locking identification result and the second locking identification result.

[0052] In the above scheme, the acquisition unit is used to acquire the locking logic judgment threshold from the unlocking logic judgment rule;

[0053] The determining unit is used to determine the comparison results of the multiple sets of RSSIs with the locking logic judgment threshold; and to determine the second locking identification result based on the comparison results.

[0054] In the above scheme, the determining unit is used to determine the lock as the lock identification result when both the first lock identification result and the second lock identification result identify the lock.

[0055] In the above scheme, the device further includes an adjustment unit.

[0056] The adjustment unit is used to adjust the first locking state machine to a set state when the second locking identification result is locking; and to adjust the second locking state machine to a set state when the first locking identification result is locking.

[0057] The determining unit is used to take the locking as the locking identification result when both the first locking state machine and the second locking state machine are in the set state.

[0058] In the above scheme, the adjustment unit is used to adjust the state of the first locking state machine and the second locking state machine to a non-set state when any door of the target vehicle is in an open state; or, when the vehicle controller of the target vehicle switches from a sleep state to a wake-up state, adjust the state of the first locking state machine and the second locking state machine to a non-set state; or, when the target vehicle completes the corresponding locking operation, adjust the state of the first locking state machine and the second locking state machine to a non-set state.

[0059] In the above scheme, the device further includes an input unit;

[0060] The input unit is used to input the multiple sets of RSSI and the multiple sets of differential signals into the unlocking AI model in the unlocking AI model to obtain the first unlocking recognition result;

[0061] The determining unit is configured to determine a second unlocking identification result based on the unlocking / locking logic determination rules and the multiple sets of RSSIs; if the first unlocking identification result is unlocked, the first unlocking identification result is determined as the unlocking identification result; or if the second unlocking identification result is unlocked, the second unlocking identification result is determined as the unlocking identification result.

[0062] In the above scheme, the determining unit is further configured to determine the location area of ​​the user equipment based on the multiple sets of RSSIs; if the location area is not the rear area of ​​the target vehicle and the first unlocking recognition result is unlocked, the first unlocking recognition result is determined as the unlocking recognition result; or if the location area is not the rear area of ​​the target vehicle and the second unlocking recognition result is unlocked, the second unlocking recognition result is determined as the unlocking recognition result; if the location area is the rear area of ​​the target vehicle, the second unlocking recognition result is determined as the unlocking recognition result.

[0063] In the above scheme, the device further includes an acquisition unit and a training unit;

[0064] The acquisition unit is used to acquire multiple sets of sample RSSIs and multiple sets of sample differential signals corresponding to the multiple sets of sample RSSIs during historical unlocking, as well as sample unlocking operation results corresponding to the multiple sets of sample RSSIs and multiple sets of sample differential signals.

[0065] The training unit is used to train an initial unlocking AI model based on the multiple sets of sample RSSI, the multiple sets of sample differential signals, and the sample unlocking and locking operation results, to obtain the unlocking AI model; the initial unlocking model in the initial unlocking AI model includes an initial convolutional neural network and an initial recurrent neural network; the initial locking model in the initial unlocking AI model includes the initial convolutional neural network and the initial recurrent neural network.

[0066] In the above scheme, the training unit is used to input the multiple sets of sample RSSI and the multiple sets of sample difference signals into the initial unlocking AI model to obtain the output unlocking recognition result; determine the binary cross-entropy loss of the initial unlocking AI model based on the output unlocking recognition result and the sample unlocking operation result; if the binary cross-entropy loss is greater than or equal to a preset loss threshold, continue to train the initial unlocking AI model using the multiple sets of sample RSSI, the multiple sets of sample difference signals, and the sample unlocking operation result to obtain a training model; if the corresponding binary training cross-entropy loss of the training model is less than or equal to the preset loss threshold, use the training model as the unlocking AI model.

[0067] In the above scheme, the execution unit is used to perform a locking operation when the communication link between the user equipment and the target vehicle is broken; or, when the values ​​of the multiple sets of RSSI are less than or equal to a preset threshold, the locking operation is performed.

[0068] This application embodiment also provides a vehicle unlocking / locking device, the vehicle unlocking / locking device comprising:

[0069] Memory is used to store executable instructions for a computer;

[0070] The processor, when executing computer-executable instructions stored in the memory, implements the vehicle unlocking and locking method provided in the embodiments of this application.

[0071] This application provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the vehicle unlocking / locking method provided in this application.

[0072] This application provides a computer program product, including a computer program or computer-executable instructions. When the computer program or computer-executable instructions are executed by a processor, they implement the vehicle unlocking and locking method provided in this application.

[0073] The embodiments of this application have the following beneficial effects: The vehicle unlocking and locking device acquires multiple sets of RSSIs and multiple sets of differential signals corresponding to these RSSIs. Each set of RSSIs is a signal emitted by the user equipment. Using an unlocking and locking AI model, the first unlocking recognition result or the first locking recognition result can be determined based on the multiple sets of RSSIs and the multiple sets of differential signals. This unlocking and locking AI model is trained using multiple sets of sample RSSIs and the multiple sets of sample differential signals corresponding to these sample RSSIs from historical successful unlocking and locking events. This allows the unlocking and locking AI model to accurately determine the first unlocking result based on the multiple sets of RSSIs and the multiple sets of differential signals corresponding to these RSSIs. The identification result or the first locking identification result is used to assist the second unlocking identification result or the second locking identification result determined by the unlocking and locking logic judgment rules based on multiple sets of RSSI. The final unlocking and locking identification result (i.e., the unlocking identification result or the locking identification result) is determined by comprehensively judging the first unlocking identification result and the second unlocking identification result or by comprehensively judging the first locking identification result and the second locking identification result. This ensures that even if the RSSI strength of the signal emitted by the user equipment is blocked, it will not affect the prediction result of the unlocking and locking identification result, thereby improving the accuracy of vehicle unlocking and locking. Attached Figure Description

[0074] Figure 1 This is a schematic diagram of a vehicle unlocking / locking structure in the prior art provided in an embodiment of this application;

[0075] Figure 2 This is a flowchart of a vehicle unlocking / locking method provided in an embodiment of this application;

[0076] Figure 3 This is a schematic diagram of an exemplary vehicle locking / unlocking structure provided in an embodiment of this application;

[0077] Figure 4 This is an exemplary vehicle locking logic diagram provided in an embodiment of this application;

[0078] Figure 5 This is an exemplary vehicle unlocking logic diagram provided in an embodiment of this application;

[0079] Figure 6 This is an exemplary diagram illustrating the training of an AI model for unlocking / locking, provided in an embodiment of this application.

[0080] Figure 7 This is a schematic flowchart of an exemplary vehicle unlocking / locking method provided in an embodiment of this application;

[0081] Figure 8 This is a schematic diagram of the composition structure of a vehicle unlocking and locking device provided in an embodiment of this application;

[0082] Figure 9 This is a schematic diagram of the composition of a vehicle unlocking and locking device provided in an embodiment of this application.

[0083] It should be noted that the terms "first" and "second" mentioned above are only used to distinguish between different options and do not represent the degree of superiority or inferiority of the options or their priority in the implementation process. Detailed Implementation

[0084] To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0085] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0086] In the following description, the terms "first, second, third" are used merely to distinguish similar objects and do not represent a specific ordering of objects. It is understood that "first, second, third" may be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0087] In this application embodiment, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0088] Unless otherwise defined, all technical and scientific terms used in the embodiments of this application have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in the embodiments of this application is for the purpose of describing the embodiments of this application only and is not intended to limit this application.

[0089] In the implementation of this application, the collection and processing of relevant data should strictly comply with the requirements of relevant national laws and regulations, obtain the informed consent or separate consent of the personal information subject, and carry out subsequent data use and processing within the scope of laws and regulations and the authorization of the personal information subject.

[0090] Before providing a further detailed description of the embodiments of this application, the nouns and terms involved in the embodiments of this application will be explained, and the nouns and terms involved in the embodiments of this application shall be interpreted as follows.

[0091] 1) BLE (Bluetooth Low Energy).

[0092] 2) RSSI (Received Signal Strength Indication).

[0093] 3) XCU (extended domain control unit).

[0094] Currently, Bluetooth key unlocking and locking functions have some shortcomings, primarily because the logic-based judgment cannot cover all scenarios of near-to-unlock and far-to-lock functions. Specifically, this method of unlocking and locking based on signal strength thresholds (i.e., logic-based judgment) is susceptible to environmental factors, device differences, and usage patterns, leading to unstable algorithm performance. For example, when the device sending information to the Bluetooth module is placed in a bag or obstructed by the user's body, the RSSI signal strength received by the vehicle's Bluetooth module decreases significantly, causing a deviation from the calibration data and affecting the unlocking effect (the "standing still" phenomenon). Simultaneously, when the user is near the vehicle, the reduced RSSI strength caused by signal obstruction can also lead to false locking judgments, resulting in incorrect locking.

[0095] For example, such as Figure 1 As shown, the vehicle includes a Bluetooth module, an XCU A / M core, and a vehicle control module. The mobile phone acts as the device. A BLE connection is established between the phone and the Bluetooth module. The Bluetooth module communicates with the XCU A / M core via CAN. If the XCU A / M core determines that the device sending information to the Bluetooth module is inside a bag or obstructed by the user's body, resulting in a significant decrease in the RSSI signal strength received by the vehicle's Bluetooth module, it will prevent the vehicle control module from unlocking. Alternatively, when the user is near the vehicle, a decrease in RSSI strength due to obstruction of the device's signal can also cause the XCU A / M core to misjudge the locking mechanism, thus triggering an erroneous locking of the vehicle control module.

[0096] The problems existing in the related technologies can be solved by means of the methods in the following embodiments.

[0097] This application provides a vehicle unlocking / locking method, which is applied to a vehicle unlocking / locking device. Figure 2 A flowchart of a vehicle unlocking / locking method provided in this application embodiment is shown below. Figure 2 As shown, the vehicle unlocking method applied to the vehicle unlocking device may include:

[0098] S101. Obtain multiple sets of RSSI and multiple sets of differential signals corresponding to the multiple sets of RSSI.

[0099] The vehicle unlocking and locking method provided in this application embodiment is applicable to scenarios where the target vehicle needs to be unlocked or locked.

[0100] In the embodiments of this application, the vehicle unlocking and locking device can be implemented in various forms. For example, the vehicle unlocking and locking device described in this application may include a vehicle, a server, or other devices. The specific vehicle unlocking and locking device can be determined according to the actual situation, and the embodiments of this application do not limit it in this regard.

[0101] In this embodiment, each RSSI in the multiple RSSIs is a signal emitted by the user equipment; each differential signal in the multiple differential signals is information obtained from every two RSSIs in each of the corresponding multiple RSSIs.

[0102] It should be noted that the user equipment refers to the equipment carried by the user to whom the target vehicle belongs. The vehicle unlocking / locking device can be the target vehicle.

[0103] In this embodiment of the application, the user equipment can be a user's mobile phone or other electronic devices that can communicate with the Bluetooth module (or other receiving antenna) in the target vehicle. The specific user equipment can be determined according to the actual situation, and this embodiment of the application does not limit it.

[0104] It should be noted that the user equipment can establish a communication connection with the Bluetooth module via Bluetooth, or the user equipment can establish a connection with the target vehicle through other connection methods. The specific method for establishing a communication connection between the user equipment and the target vehicle can be determined according to the actual situation, and this application embodiment does not limit it.

[0105] In this embodiment of the application, the number of Bluetooth modules in the target vehicle is multiple. The specific number of Bluetooth modules can be determined according to the actual situation, and this embodiment of the application does not limit this.

[0106] For example, the number of Bluetooth modules can be 5, or other values. The specific number can be determined according to the actual situation, and this application does not limit it.

[0107] In this embodiment, the user equipment is equipped with a signal transmitting unit. After the signal transmitting unit sends a signal, multiple Bluetooth modules in the target vehicle receive the signal, obtaining multiple RSSI signals, i.e., a set of RSSI signals. The signal transmitting unit transmits signals according to a transmission cycle. In one signal transmission cycle, the signal transmitting unit sends a signal once, and the multiple Bluetooth modules in the target vehicle each receive the signal once, obtaining a set of RSSI signals. In multiple signal transmission cycles, the signal transmitting unit sends signals multiple times, and the multiple Bluetooth modules in the target vehicle each receive the signals multiple times, obtaining multiple sets of RSSI signals.

[0108] In this embodiment of the application, when the target vehicle receives a set of RSSI signals, it determines the differential information between every two RSSI signals in this set of RSSI signals, thereby obtaining a set of differential signals; when the target vehicle receives multiple sets of RSSI signals, it determines the differential information between every two RSSI signals in each of the multiple sets of RSSI signals, thereby obtaining multiple sets of differential signals.

[0109] For example, the target vehicle is equipped with 5 Bluetooth modules. After the signal transmitting unit on the user equipment sends a signal once, the 5 Bluetooth modules receive the signal respectively and obtain a set of RSSI signals, that is, 5 RSSI signals. Then, the differential signal between every two RSSI signals in the 5 RSSI signals is determined to obtain 20 differential signals, that is, a set of differential signals.

[0110] S102. Using the unlocking / locking AI model and unlocking / locking logic judgment rules, and based on multiple sets of RSSI and multiple sets of differential signals, determine the unlocking recognition result or the locking recognition result.

[0111] In this embodiment, after the vehicle unlocking / locking device obtains multiple sets of RSSIs and multiple sets of differential signals corresponding to the multiple sets of RSSIs from the target vehicle, it can use the unlocking / locking AI model and unlocking / locking logic judgment rules to determine the unlocking recognition result or the locking recognition result based on the multiple sets of RSSIs and multiple sets of differential signals.

[0112] It should be noted that the unlocking AI model is trained using multiple sets of sample RSSIs and the sample difference signals corresponding to the multiple sets of sample RSSIs from historical successful unlocking events.

[0113] In this embodiment, the unlocking / locking logic determination rule can be a rule in the prior art or a rule set by the vehicle unlocking / locking device. The specific rule can be determined according to the actual situation, and this embodiment does not limit it.

[0114] For example, if the unlocking / locking logic determination rule is a rule in the prior art, the unlocking / locking logic determination rule can be to set a signal strength threshold, compare the RSSI of the received signal from the user equipment with the signal strength threshold, such as unlocking if the RSSI is greater than or equal to the first signal strength threshold, and locking if the RSSI is less than the second signal strength threshold.

[0115] It should be noted that the first signal strength threshold and the second signal strength threshold can be thresholds configured in the vehicle unlocking and locking device, thresholds transmitted to the vehicle unlocking and locking device by other devices, or thresholds obtained by the vehicle unlocking and locking device through other means. The specific way in which the vehicle unlocking and locking device obtains the signal strength threshold can be determined according to the actual situation, and this application embodiment does not limit it in this way.

[0116] It should also be noted that the first signal strength threshold and the second signal strength threshold can be the same, or they can be different. The specific threshold can be determined according to the actual situation, and this application does not limit this.

[0117] In this embodiment, the process by which the vehicle unlocking / locking device determines the locking recognition result using an unlocking / locking AI model and unlocking / locking logic judgment rules, based on multiple sets of RSSI and multiple sets of differential signals, includes: determining a first locking recognition result using the locking AI model in the unlocking / locking AI model, based on multiple sets of RSSI and multiple sets of differential signals; determining a second locking recognition result based on the unlocking / locking logic judgment rules and multiple sets of RSSI; and determining the locking recognition result based on the first locking recognition result and the second locking recognition result.

[0118] In this embodiment, the unlocking / locking AI model can be a model configured in the vehicle unlocking / locking device, a model transmitted to the vehicle unlocking / locking device from other devices, or a model obtained by the vehicle unlocking / locking device through other means. The specific way in which the vehicle unlocking / locking device obtains the unlocking / locking AI model can be determined according to the actual situation, and this embodiment does not limit it.

[0119] It should be noted that the locking and unlocking AI model includes both locking and unlocking AI models.

[0120] It should also be noted that the locked AI model includes Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs). The unlocked AI model also uses CNNs and RNNs.

[0121] In this embodiment, the locking AI model further includes a fully connected layer. The CNN in the locking AI model is used to obtain data features of multiple sets of RSSI and multiple sets of differential signals. The RNN in the locking AI model is used to obtain the temporal dependencies of multiple sets of RSSI and multiple sets of differential signals. Then, the fully connected layer is used to map the confidence of the unlocking event based on the data features and dependencies, thus obtaining the first locking identification result.

[0122] In this embodiment, the unlocking / locking logic determination rule can be a rule configured in the vehicle unlocking / locking device, a rule transmitted to the vehicle unlocking / locking device from other devices, or a rule obtained by the vehicle unlocking / locking device through other means. The specific way in which the vehicle unlocking / locking device obtains the unlocking / locking logic determination rule can be determined according to the actual situation, and this embodiment does not limit it.

[0123] It should be noted that the unlocking logic determination rules include unlocking logic determination rules and locking logic determination rules.

[0124] In this embodiment, determining the second locking identification result based on the unlocking logic determination rule and multiple sets of RSSIs includes: determining the second locking identification result based on the locking logic determination rule in the unlocking logic determination rule and multiple sets of RSSIs. If the locking logic determination rule is a rule in the prior art, the multiple sets of RSSIs can be compared with a second signal strength threshold to determine the second locking identification result.

[0125] Understandably, by using the unlocking logic judgment rules and the locking AI model in the unlocking AI model to process multiple sets of RSSIs and multiple differential signals between multiple sets of RSSIs, it is not necessary to confuse with the unlocking process, thus improving the accuracy of processing multiple sets of RSSIs and multiple differential signals between multiple sets of RSSIs.

[0126] In this embodiment of the application, the process by which the vehicle unlocking / locking device determines the second locking identification result based on the unlocking / locking logic determination rules and multiple sets of RSSIs includes: obtaining the locking logic determination threshold from the unlocking / locking logic determination rules; determining the comparison results of multiple sets of RSSIs with the locking logic determination threshold; and determining the second locking identification result based on the comparison results.

[0127] In this embodiment, the latching logic determination threshold is the second signal strength threshold.

[0128] In this embodiment, multiple comparison results are obtained by comparing multiple sets of RSSI with the latching logic decision threshold. Each comparison result includes either an RSSI less than the latching logic decision threshold, or an RSSI greater than or equal to the latching logic decision threshold.

[0129] It should be noted that when the second locking identification result is determined to be locked based on multiple sets of RSSI using the unlocking logic judgment rules, the state of the first locking state machine is adjusted to the set state.

[0130] In this embodiment, if the RSSI of each comparison result identifier in the plurality of comparison results is less than the locking logic judgment threshold, the second locking identification result is determined to be locking; or if the RSSI of 2 / 3 of the comparison result identifiers in the plurality of comparison results is less than the locking logic judgment threshold, the second locking identification result is determined to be locking. The specific method of determining the second locking identification result based on the comparison results can be determined according to the actual situation, and this embodiment does not limit it.

[0131] For example, if it is determined that 2 / 3 of the comparison results are less than the lockout logic judgment threshold, then the second lockout identification result is determined to be a lockout, and then the state of the first lockout state machine is adjusted to the set state (indicating that the second lockout identification result is a lockout), thus obtaining the second lockout identification result.

[0132] It should be noted that the first latching state machine is the M-core latching state machine.

[0133] Understandably, the unlocking and locking logic determination rule performs locking determination processing on multiple sets of RSSI. The corresponding computing power requirement of the unlocking and locking logic determination rule is low, which reduces the computing power consumption in the vehicle unlocking and locking device. Moreover, when the locking AI model is not working, the unlocking and locking logic determination rule can be used directly to realize the locking recognition process, ensuring the reliable operation of locking recognition in the vehicle unlocking and locking device.

[0134] In this embodiment of the application, the process of determining the first locking recognition result by the vehicle unlocking device using the locking AI model in the unlocking AI model and based on multiple sets of RSSI and multiple sets of differential signals includes: inputting multiple sets of RSSI and multiple sets of differential signals into the locking AI model to obtain the first locking recognition result.

[0135] In this embodiment, the CNN in the locking AI model can be used to obtain data features of multiple RSSIs and multiple differential signals, and the RNN can be used to obtain the temporal dependencies of multiple RSSIs and multiple differential signals. Then, the fully connected layer can be used to map the confidence of the locking event (i.e., determine the probability of locking) based on the data features and dependencies. When the confidence is within the first data range, the state of the second locking state machine is adjusted to the set state to obtain the first locking recognition result.

[0136] It should be noted that the second latching state machine is the A-core latching state machine.

[0137] It should be noted that the first data range can be the data range configured in the vehicle unlocking and locking device, the data range transmitted from other devices to the vehicle unlocking and locking device, or the data range obtained by the vehicle unlocking and locking device through other means. The specific way in which the vehicle unlocking and locking device obtains the first data range can be determined according to the actual situation, and this application embodiment does not limit it.

[0138] For example, the first data range can be a range of 0.5 to 1. The confidence level is a value within the range of 0-1. For instance, if the confidence level of a locking event is determined to be 0.9 using a locking AI model, and the first data range is 0.5-1, then 0.9 falls within the first data range. The state of the second locking state machine is then adjusted to the set state to obtain the first locking identification result.

[0139] For example, such as Figure 3 As shown, the vehicle unlocking and locking device includes a Bluetooth module, an M-core of the XCU, an A-core of the XCU, and a vehicle control module. The M-core of the XCU contains unlocking and locking logic rules, and the A-core of the XCU contains an unlocking and locking AI model. The mobile phone (i.e., user equipment) establishes a connection with the Bluetooth module. The Bluetooth module listens for and receives multiple sets of RSSI signals sent by the mobile phone and transmits these RSSIs to the M core of the XCU via CAN messages. The XCU determines the multiple differential signals corresponding to the multiple RSSIs. The M core of the XCU uses the logic positioning module to determine the second unlock recognition result or the second lock recognition result based on the multiple RSSIs. The A core of the XCU uses the AI ​​model positioning module (unlocking AI model) to determine the first unlock recognition result or the first lock recognition result based on the multiple RSSIs and the multiple differential signals. The M core of the XCU obtains the final unlock recognition result or lock recognition result based on the second unlock recognition result and the first unlock recognition result or the second lock recognition result and the first lock recognition result. Then, it sends an unlocking or locking command to the vehicle control module, which then executes the corresponding unlocking or locking operation.

[0140] Understandably, multiple sets of RSSI and multiple sets of differential signals are input into the locking AI model, and the locking AI model outputs the first locking recognition result. Since the locking AI model is trained using multiple sets of sample RSSI and multiple sets of sample differential signals corresponding to the sample RSSI from historical successful locking, the locking AI model can determine the accurate first locking recognition result based on multiple sets of RSSI and multiple sets of differential signals.

[0141] In this embodiment of the application, the process of determining the lock identification result based on the first lock identification result and the second lock identification result by the vehicle unlocking and locking device includes: when both the first lock identification result and the second lock identification result indicate lock, the lock is determined as the lock identification result.

[0142] In the embodiments of this application, if either the first locking identification result or the second locking identification result is not locked, then non-locking is determined as the locking identification result.

[0143] In this embodiment, the vehicle unlocking / locking device can determine the locking recognition result by using an unlocking / locking AI model and unlocking / locking logic judgment rules, based on multiple sets of RSSI and multiple sets of differential signals; alternatively, it can determine the locking recognition result by using an unlocking / locking AI model, based on multiple sets of RSSI and multiple sets of differential signals. The specific execution method can be determined according to the actual situation, and this embodiment does not limit it.

[0144] In this embodiment of the application, the process by which the vehicle unlocking and locking device determines the locking as the locking identification result when both the first locking identification result and the second locking identification result indicate locking includes: when the second locking identification result is locking, adjusting the first locking state machine to a set state; when the first locking identification result is locking, adjusting the second locking state machine to a set state; and when both the first locking state machine and the second locking state machine are in a set state, taking the locking as the locking identification result.

[0145] In this embodiment, after the vehicle locking / unlocking device performs the corresponding locking operation based on the locking identification result, if any door of the target vehicle is in the open state, the state of the first locking state machine and the second locking state machine are adjusted to the non-set state; or, if the vehicle controller of the target vehicle switches from the sleep state to the wake-up state, the state of the first locking state machine and the second locking state machine are adjusted to the non-set state; or, if the target vehicle completes the corresponding locking operation, the state of the first locking state machine and the second locking state machine are adjusted to the non-set state.

[0146] In the embodiments of this application, both the first locking state machine and the second locking state machine include two states: a set state and a non-set state.

[0147] It should be noted that the non-set state indicator is an indicator of non-locking.

[0148] In this embodiment, any door in the target vehicle includes the four doors of the target vehicle and the trunk of the target vehicle. That is, when any door in the target vehicle is in the open state, the states of the first locking state machine and the second locking state machine are adjusted to the non-set state, at which time both the first state machine and the second state machine are marked as not locked.

[0149] It should be noted that the vehicle controller of the target vehicle can be an Extended Domain Control Unit (XCU).

[0150] In this embodiment of the application, when the vehicle controller of the target vehicle switches from a dormant state to a wake-up state, the states of the first locking state machine and the second locking state machine are adjusted to a non-set state, at which time both the first state machine and the second state machine are marked as not locked.

[0151] In this embodiment of the application, after the corresponding locking operation is performed based on the locking identification result, the states of the first locking state machine and the second locking state machine are adjusted to the non-set state. At this time, both the first state machine and the second state machine are marked as not locked.

[0152] In this embodiment of the application, when the communication link between the user equipment and the target vehicle is disconnected, the state of the first locking state machine and the second locking state machine is not determined, and the locking operation is performed directly; or, when the values ​​of multiple RSSIs are less than or equal to a preset threshold, the state of the first locking state machine and the second locking state machine is not determined, and the locking operation is performed.

[0153] It should be noted that the preset threshold can be a threshold configured in the vehicle unlocking and locking device, a threshold transmitted to the vehicle unlocking and locking device from other devices, or a threshold obtained by the vehicle unlocking and locking device in other ways. The specific way in which the vehicle unlocking and locking device obtains the preset threshold can be determined according to the actual situation, and this application embodiment does not limit it in this regard.

[0154] In this embodiment of the application, the process for determining the locking recognition result is as follows: Figure 4 As shown, the system includes: a vehicle unlocking / locking device that can determine a first lockout recognition result using an unlocking / locking AI model and based on multiple sets of RSSIs and differential signals; a vehicle unlocking / locking device that can also determine a second lockout recognition result based on the lockout logic judgment rules in the unlocking / locking logic judgment rules and multiple sets of RSSIs; if the second lockout recognition result is locked, an M-core locking command is generated, and the state of the M-core locking state machine is adjusted to the set state; the first lockout recognition result is determined using the locking AI model in the unlocking / locking AI model and based on multiple sets of RSSIs and differential signals; if the first lockout recognition result is locked, the state of the A-core locking state machine is adjusted to the set state; and then, the fusion judgment module determines the lockout recognition result based on the first and second lockout recognition results and outputs it (outputs vehicle control commands).

[0155] In this embodiment of the application, the process for determining the locking recognition result includes:

[0156] a. If the latching AI model of core A is not working, core M will execute the latch normally.

[0157] b. If the A-core locking AI model is working, it is necessary to perform a fusion judgment based on the first locking recognition result of the A-core and the second locking recognition result of the M-core to determine the locking recognition result. If the locking recognition result is locking, then send the vehicle exit locking command.

[0158] c. The process of judging the locking recognition result specifically includes:

[0159] i. If the M core meets the locking condition (the second locking identification result is determined to be locked using multiple sets of RSSI), then the state of the M core locking state machine is adjusted to the set state.

[0160] ii. After receiving the locking command from the A core, the vehicle locking / unlocking device adjusts the state of the A core locking state machine to the set state according to the VeINP_BtkeyModelCtrlCmd(1-3)_enum:0x3 Phone1&&Lock command.

[0161] iii. The fusion judgment module (&&) judges whether the two unlocking state machines of the M core and A core simultaneously meet the set state (i.e., determine whether the first lock identification result and the second lock identification result are both locked). If they are simultaneously met (i.e., determine that the first lock identification result and the second lock identification result are both locked), then a lock command is sent to execute the corresponding lock function.

[0162] It should be noted that opening any one of the five doors (four passenger doors and one trunk door) of the target vehicle simultaneously clears the state machines of the M-core and A-core (i.e., when any door of the target vehicle is open, the states of the first and second locking state machines are adjusted to the non-set state). Upon XCU sleep / wake-up, the states of the M-core and A-core state machines are simultaneously cleared (when the vehicle controller of the target vehicle switches from sleep to wake-up, the states of the first and second locking state machines are adjusted to the non-set state). After locking, the states of the M-core and A-core state machines are simultaneously cleared (when the target vehicle completes the corresponding locking operation, the states of the first and second locking state machines are adjusted to the non-set state). The M-core maintains the secondary locking loop function: When the main module threshold > 90 (i.e., multiple sets of RSSI signal strengths are determined to be weak), it directly locks without judging the state machine state (locking operation is performed when the values ​​of multiple sets of RSSI are less than or equal to the preset threshold). For example, if it is determined that the signal strength of RSSI within 1 second is weak, locking is performed directly without judging the state machine state. It should be noted that the main module is the module in the target vehicle that communicates with the user equipment. When the Bluetooth connection between the user equipment and the target vehicle is lost, locking is performed directly without judging the state machine state (i.e., locking operation is performed when the communication link between the user equipment and the target vehicle is broken).

[0163] d. After receiving the locking command from the A core, the vehicle unlocking / locking device, according to the VeINP_BtkeyModelCtrlCmd(1-3)_en um:0x2Phone1 Lock instruction, directly sends the locking command and executes the locking function without checking the states of the M core state machine and the A core state machine. Locking simultaneously clears the states of the M core and the A core state machine.

[0164] e. During the Bluetooth connection process between the user equipment and the target vehicle, the M core does not respond to the unlock and lock signals of the A core and does not cache the A core commands. When the Bluetooth connection is successful (Bkeyvalid = valid), it then responds to the unlock and lock commands of the A core, and this does not conflict with the above logic.

[0165] Understandably, by fusing the results with those of traditional algorithms (i.e., the first locking recognition result), a faster and more stable unlocking effect can be achieved, reducing the occurrence of "standing still" and erroneous locking phenomena, and improving the user experience.

[0166] In this embodiment, the process by which the vehicle unlocking / locking device determines the unlocking identification result using an unlocking / locking AI model and unlocking / locking logic judgment rules, based on multiple sets of RSSIs and multiple sets of differential signals, includes: inputting multiple sets of RSSIs and multiple sets of differential signals into the unlocking AI model in the unlocking / locking AI model to obtain a first unlocking identification result; determining a second unlocking identification result based on the unlocking / locking logic judgment rules and multiple sets of RSSIs; if the first unlocking identification result is unlocked, determining the first unlocking identification result as the unlocking identification result; or if the second unlocking identification result is unlocked, determining the second unlocking identification result as the unlocking identification result.

[0167] In this embodiment of the application, the process of determining the second unlocking identification result based on the unlocking logic determination rule and multiple sets of RSSI includes: determining the second unlocking identification result based on the unlocking logic determination rule and multiple sets of RSSI in the unlocking logic determination rule.

[0168] In this embodiment, the vehicle unlocking / locking device can determine the first time corresponding to the first unlocking recognition result while simultaneously determining the first unlocking recognition result, and determine the second time corresponding to the second unlocking recognition result while simultaneously determining the second unlocking recognition result. Alternatively, the vehicle unlocking / locking device can determine the first unlocking recognition result as the unlocking recognition result if the second time has not been generated but the first time already exists; or determine the second unlocking recognition result as the unlocking recognition result if the first time has not been generated but the second time already exists.

[0169] In this embodiment, the CNN in the unlocking AI model can be used to obtain data features of multiple sets of RSSI and multiple sets of differential signals, the RNN in the unlocking AI model can be used to obtain the temporal dependencies of multiple sets of RSSI and multiple sets of differential signals, and then the fully connected layer in the unlocking AI model can be used to map the confidence of the unlocking event (i.e., determine the probability that the result is unlocking) based on the data features and dependencies. When the confidence is within the second data range, the first unlocking recognition result is obtained.

[0170] It should be noted that the second data range can be the data range configured in the vehicle unlocking and locking device, the data range transmitted from other devices to the vehicle unlocking and locking device, or the data range obtained by the vehicle unlocking and locking device through other means. The specific way in which the vehicle unlocking and locking device obtains the second data range can be determined according to the actual situation, and this application embodiment does not limit it.

[0171] For example, the second data range can be a range of 0.5 to 1. The confidence level is a value within the range of 0-1. For instance, if the unlocking AI model determines the confidence level of the unlocking event to be 0.9, and the second data range is 0.5-1, then 0.9 falls within the second data range. Therefore, the model output is determined to be unlocked, thus obtaining the first unlocking recognition result.

[0172] In the embodiments of this application, the first data range and the second data range may be the same or different. The specific range can be determined according to the actual situation, and the embodiments of this application do not limit this.

[0173] In this embodiment, if the vehicle unlocking / locking device first determines a first unlocking identification result and the first unlocking identification result is unlocked, then it determines the first unlocking identification result as the unlocking identification result. If the vehicle unlocking / locking device first determines a second unlocking identification result and the second unlocking identification result is unlocked, then it determines the second unlocking identification result as the unlocking identification result.

[0174] Understandably, by using the unlocking logic judgment rules and the unlocking AI model in the unlocking AI model to process multiple sets of RS SI and the multiple sets of differential signals corresponding to multiple sets of RSSI, and by using the result that determines the unlocking conclusion the fastest as the final unlocking recognition result, the speed of unlocking the target vehicle is improved.

[0175] In this embodiment, the vehicle unlocking / locking device can also determine the location area of ​​the user equipment based on multiple sets of RSSIs; if the location area is not the rear area of ​​the target vehicle and the first unlocking identification result is unlocked, the first unlocking identification result is determined as the unlocking identification result; or if the location area is not the rear area of ​​the target vehicle and the second unlocking identification result is unlocked, the second unlocking identification result is determined as the unlocking identification result; if the location area is the rear area of ​​the target vehicle, the second unlocking identification result is determined as the unlocking identification result.

[0176] In this embodiment, the rear area of ​​the target vehicle can be the area configured by the vehicle unlocking device, information transmitted to the vehicle unlocking device from other devices, or area information obtained by the vehicle unlocking device in other ways. The specific way in which the vehicle unlocking device obtains the rear area of ​​the target vehicle can be determined according to the actual situation, and this embodiment does not limit it.

[0177] In this embodiment of the application, if the location area is not the rear area of ​​the target vehicle, and the first unlock recognition result is obtained first and the first unlock recognition result is unlocked, the first unlock recognition result can be determined as the unlock recognition result; or if the location area is not the rear area of ​​the target vehicle, and the second unlock recognition result is obtained first and the second unlock recognition result is unlocked, the second unlock recognition result can be determined as the unlock recognition result.

[0178] For example, the rear area of ​​the target vehicle can be the rear bumper PE area of ​​the target vehicle.

[0179] In this embodiment of the application, the unlocking fusion logic is as follows: Figure 5 As shown, it includes:

[0180] S1. The vehicle unlocking and locking device determines the second unlocking identification result based on the unlocking and locking logic judgment rules and multiple sets of RSSI.

[0181] It should be noted that if the vehicle unlocking / locking device determines that the unlocking AI model of core A is not working, it directly uses the unlocking logic judgment rules in core M to determine the second unlocking recognition result and directly uses the second unlocking recognition result as the unlocking recognition result. If the vehicle unlocking / locking device determines that the unlocking AI model of core A is working, it uses the unlocking logic judgment rules in core M to determine the second unlocking recognition result. Using the unlocking AI model within the unlocking / locking AI model, based on multiple sets of RSSI and differential signals, it determines the first unlocking recognition result. If the second unlocking recognition result is obtained first and is indeed unlocked, then the second unlocking recognition result is used as the unlocking recognition result.

[0182] S2. The vehicle unlocking and locking device inputs multiple sets of RSSI and multiple sets of differential signals into the unlocking AI model in the unlocking and locking AI model to obtain the first unlocking recognition result.

[0183] It should be noted that when the vehicle unlocking device determines that the unlocking AI model of core A is working, multiple sets of RSSI and multiple sets of differential signals are input into the unlocking AI model in the unlocking AI model to obtain the first unlocking recognition result.

[0184] S3. The vehicle unlocking and locking device determines the location area of ​​the user equipment based on multiple sets of RSSI.

[0185] In the embodiments of this application, the method of determining the location area of ​​the user equipment based on multiple sets of RSSI is the prior art, and this embodiment of the application does not limit it.

[0186] S4. If the vehicle unlocking device determines that the location area is not the rear area of ​​the target vehicle and the first unlocking identification result is unlocked, the first unlocking identification result shall be determined as the unlocking identification result; or if the location area is determined to be not the rear area of ​​the target vehicle and the second unlocking identification result is unlocked, the second unlocking identification result shall be determined as the unlocking identification result.

[0187] S5. When the vehicle unlocking and locking device determines that the location area is the rear area of ​​the target vehicle, it determines the second unlocking identification result as the unlocking identification result.

[0188] Understandably, the location of the user equipment is determined based on multiple sets of RSSI, and then it is determined whether the user has an unlocking need based on the location of the user equipment. When it is determined that the user has an unlocking need for the vehicle based on the location of the user equipment, the target vehicle is unlocked, thus improving the accuracy of vehicle unlocking.

[0189] In this embodiment, before the vehicle unlocking / locking device determines the unlocking or locking recognition result based on multiple sets of RSSI and multiple sets of differential signals using the unlocking / locking AI model and unlocking / locking logic judgment rules, it also acquires multiple sets of sample RSSI and multiple sets of sample differential signals corresponding to the historical unlocking / locking, as well as sample unlocking / locking operation results corresponding to the multiple sets of sample RSSI and multiple sets of sample differential signals. Based on the multiple sets of sample RSSI, multiple sets of sample differential signals, and sample unlocking / locking operation results, an initial unlocking / locking AI model is trained to obtain the unlocking / locking AI model.

[0190] It should be noted that the initial unlocking model in the initial unlocking AI model includes an initial convolutional neural network and an initial recurrent neural network; the initial locking model in the initial unlocking AI model includes an initial convolutional neural network and an initial recurrent neural network.

[0191] It should be noted that the multiple sample RSSIs refer to the RSSIs of successful historical unlocking and de-locking events. The multiple sample differential signals are signals determined based on the multiple sample RSSIs. The method for determining the multiple sample differential signals based on the multiple sample RSSIs is the same as the method for determining the multiple differential signals based on the multiple RSSIs.

[0192] In this embodiment, the process of training an initial unlocking AI model based on multiple sets of sample RSSI, multiple sets of sample difference signals, and sample unlocking operation results to obtain an unlocking AI model includes: inputting multiple sets of sample RSSI and multiple sets of sample difference signals into the initial unlocking AI model to obtain an output unlocking recognition result; determining the binary cross-entropy loss of the initial unlocking AI model based on the output unlocking recognition result and the sample unlocking operation results; if the binary cross-entropy loss is greater than or equal to a preset loss threshold, continuing to train the initial unlocking AI model using multiple sets of sample RSSI, multiple sets of sample difference signals, and sample unlocking operation results to obtain a trained model; and if the corresponding binary training cross-entropy loss of the trained model is less than or equal to a preset loss threshold, using the trained model as the unlocking AI model.

[0193] In this embodiment, the preset loss threshold can be a threshold configured in the vehicle unlocking / locking device, a threshold transmitted to the vehicle unlocking / locking device from other devices, or a threshold obtained by the vehicle unlocking / locking device through other means. The specific way in which the vehicle unlocking / locking device obtains the preset loss threshold can be determined according to the actual situation, and this embodiment does not limit it.

[0194] For example, online data, i.e., historical unlocking and blocking data, can be acquired and filtered for useful data (data from successful historical unlocking and blocking). This yields multiple sets of sample RSSIs and corresponding differential signals (e.g., 5-channel RSSI and 20-channel inter-channel RSSI differential signals) from historical unlocking and blocking times. These are used as input to the initial unlocking and blocking AI model. The initial unlocking and blocking AI model outputs a confidence score (i.e., the probability that the initial unlocking model in the initial unlocking and blocking AI model determines the result as unlocked, and the probability that the initial blocking model in the initial blocking and blocking AI model determines the result as blocked). The confidence score threshold is 0.5 (which can be adjusted according to actual conditions). Based on the confidence score, the unlocking model outputs either unlock or not (unlocking is determined if the probability of unlocking determined by the initial unlocking model is within the second data range; otherwise, it is not unlocked). Similarly, the blocking model outputs either blocking or not (blocking is determined if the probability of blocking determined by the initial blocking model is within the first data range; otherwise, it is not blocked). The initial unlocking AI model consists of a CNN module and an RNN module. The CNN is used to acquire feature information from the data, and the RNN network is used to acquire the temporal dependencies of the data. Then, a fully connected layer maps the features to the confidence of unlocking events to determine whether to output an unlocking signal. The binary cross-entropy loss (BCE loss) function can be used as the loss during the training of the initial unlocking AI model. Specifically, compared with the unlocking model, each second in the 20 seconds before the "opening" action in the online data can be labeled as the unlocking behavior label, and these 20 sets of losses can be calculated with a regularization penalty term (the earlier the unlocking time, the smaller the penalty). Finally, the result of the set with the smallest loss value is selected as the final loss of the model for training. Compared with the locking model, each second in the 10 seconds after the last "closing" action in the online data and before the Bluetooth connection is disconnected can be labeled as the locking behavior label, and these 10 sets of losses can be calculated. Finally, the result of the set with the smallest loss value is selected as the final loss of the model for training.

[0195] For example, such as Figure 6As shown: Obtain multiple sets of sample RSSIs and corresponding sample difference signals during historical unlocking operations; determine the labels of these signals to obtain the sample unlocking operation results; input the multiple sets of sample RSSIs and sample difference signals into the initial unlocking AI model (including CNN, BLOC, KS, and LSTM) to obtain the output unlocking recognition result; based on the output unlocking recognition result and the sample unlocking operation results, determine the binary cross-entropy loss of the initial unlocking AI model; if the binary cross-entropy loss is greater than or equal to a preset loss threshold, continue training the initial unlocking AI model using the multiple sets of sample RSSIs, sample difference signals, and sample unlocking operation results until the binary training cross-entropy loss of the trained initial unlocking AI model (i.e., the training model) is less than or equal to the preset loss threshold, thus obtaining the unlocking AI model.

[0196] Understandably, by fully utilizing big data for training, the model learns signal characteristics related to signals surrounding the vehicle and lingering around it, thereby achieving stable locking and significantly reducing the probability of false locking. Furthermore, by fully utilizing big data for training, the model learns signal characteristics robust to signal occlusion, enabling stable unlocking in signal occlusion scenarios. On average, it can unlock 3-4 seconds earlier than traditional methods. Specific data comparisons are shown in Table 1.

[0197] Table 1

[0198]

[0199] S103. Based on the unlocking or locking recognition result, perform the corresponding unlocking or locking operation.

[0200] In this embodiment, the vehicle unlocking / locking device uses an unlocking / locking AI model and unlocking / locking logic judgment rules, based on multiple sets of RSSI and multiple sets of differential signals, to determine the unlocking or locking recognition result, and then performs the corresponding unlocking or locking operation based on the unlocking or locking recognition result.

[0201] In this embodiment, if the unlocking identification result is unlocked, an unlocking command is sent to the controller of the target vehicle to perform an unlocking operation; if the locking identification result is locked, a locking command is sent to the controller of the target vehicle to perform a locking operation.

[0202] For example, such as Figure 7As shown: The mobile phone (i.e., user equipment) establishes a connection with the Bluetooth module in the target vehicle. The Bluetooth module listens for and receives multiple sets of RSSIs sent by the mobile phone, representing the positioning mode. These RSSIs are then transmitted via CAN messages to the M core of the XCU. The M core of the XCU transmits the RSSIs to the A core of the XCU through the RTE layer. Based on the multiple sets of RSSIs, multiple differential signals are determined. The A core of the XCU uses the AI ​​model positioning module (unlocking / locking AI model) to perform model positioning, obtaining the first unlocking recognition result or the first locking recognition result (i.e., the model algorithm locking request). The A core of the XCU then closes the model algorithm. The lock request is transmitted to the M core of the XCU. The M core of the XCU uses the logic positioning module to determine the second unlock recognition result or the second lock recognition result based on multiple sets of RSSI. The compatibility processing module in the M core of the XCU makes a comprehensive judgment based on the first unlock recognition result and the second unlock recognition result or the first lock recognition result and the second lock recognition result to determine whether the lock condition is met, and obtains the unlock recognition result or the lock recognition result. Then, it sends an exit lock command or an approach unlock command to the vehicle control module, and uses the vehicle control module to execute the corresponding unlock operation or lock operation.

[0203] Understandably, the vehicle unlocking / locking device acquires multiple sets of RSSIs and their corresponding differential signals. Each RSSI is a signal emitted by the user equipment. Using an unlocking / locking AI model, the first unlocking or locking result can be determined based on these multiple RSSIs and differential signals. This unlocking / locking AI model is trained using multiple sample RSSIs and their corresponding differential signals from historical successful unlocking / locking events. This allows the unlocking / locking AI model to accurately determine the first unlocking result based on the multiple RSSIs and their corresponding differential signals. The first unlocking or locking result is used to assist the second unlocking or locking result determined by multiple RSSIs using the unlocking / locking logic judgment rules. A comprehensive judgment is made based on the first unlocking and second unlocking results or the first locking and second locking results to determine the final unlocking / locking result (i.e., the unlocking or locking result). This ensures that even if the RSSI strength of the signal emitted by the user equipment is blocked, it will not affect the prediction result of the unlocking / locking result, thereby improving the accuracy of vehicle unlocking / locking.

[0204] Based on the same inventive concept as the vehicle unlocking method applied in the vehicle unlocking device described above, this application provides a vehicle unlocking device 1, corresponding to a vehicle unlocking method applied in the vehicle unlocking device. Figure 8A schematic diagram of the composition structure of a vehicle unlocking / locking device provided in this application embodiment. Figure 1 The vehicle unlocking / locking device 1 may include:

[0205] Acquisition unit 11 is used to acquire multiple sets of RSSIs and multiple sets of differential signals corresponding to the multiple sets of RSSIs; each set of RSSIs is a signal emitted by the user equipment.

[0206] The determining unit 12 is used to determine the unlocking identification result or the locking identification result by using the unlocking and locking AI model and the unlocking and locking logic judgment rules, based on the multiple sets of RSSI and the multiple sets of differential signals; wherein, the unlocking and locking AI model is a model trained by using multiple sets of sample RSSI and the multiple sets of sample differential signals corresponding to the multiple sets of sample RSSI when the unlocking and locking were successful in the past.

[0207] The execution unit 13 is used to perform the corresponding unlocking operation or locking operation based on the unlocking identification result or the locking identification result.

[0208] In some embodiments of this application, the determining unit 12 is used to determine a first locking identification result by utilizing the locking AI model in the unlocking AI model and based on the multiple sets of RSSIs and the multiple sets of differential signals; determine a second locking identification result based on the unlocking logic judgment rule and the multiple sets of RSSIs; and determine the locking identification result based on the first locking identification result and the second locking identification result.

[0209] In some embodiments of this application, the acquisition unit 11 is used to acquire the locking logic determination threshold from the unlocking logic determination rule;

[0210] The determining unit 12 is used to determine the comparison results of the multiple sets of RSSI with the locking logic judgment threshold respectively; and to determine the second locking identification result based on the comparison results.

[0211] In some embodiments of this application, the determining unit 12 is used to determine the lock as the lock identification result when both the first lock identification result and the second lock identification result identify the lock.

[0212] In some embodiments of this application, the device further includes an adjustment unit;

[0213] The adjustment unit is used to adjust the first locking state machine to a set state when the second locking identification result is locking; and to adjust the second locking state machine to a set state when the first locking identification result is locking.

[0214] The determining unit 12 is used to take the locking as the locking identification result when both the first locking state machine and the second locking state machine are in the set state.

[0215] In some embodiments of this application, the adjustment unit is used to adjust the states of the first locking state machine and the second locking state machine to a non-set state when any door of the target vehicle is in an open state; or, when the vehicle controller of the target vehicle switches from a sleep state to a wake-up state, adjust the states of the first locking state machine and the second locking state machine to a non-set state; or, when the target vehicle completes the corresponding locking operation, adjust the states of the first locking state machine and the second locking state machine to a non-set state.

[0216] In some embodiments of this application, the device further includes an input unit;

[0217] The input unit is used to input the multiple sets of RSSI and the multiple sets of differential signals into the unlocking AI model in the unlocking AI model to obtain the first unlocking recognition result;

[0218] The determining unit 12 is used to determine a second unlocking identification result based on the unlocking logic determination rule and the multiple sets of RSSIs; if the first unlocking identification result is unlocked, the first unlocking identification result is determined as the unlocking identification result; or if the second unlocking identification result is unlocked, the second unlocking identification result is determined as the unlocking identification result.

[0219] In some embodiments of this application, the determining unit 12 is further configured to determine the location area of ​​the user equipment based on the multiple sets of RSSIs; if the location area is not the rear area of ​​the target vehicle and the first unlocking identification result is unlocked, determine the first unlocking identification result as the unlocking identification result; or if the location area is not the rear area of ​​the target vehicle and the second unlocking identification result is unlocked, determine the second unlocking identification result as the unlocking identification result; if the location area is the rear area of ​​the target vehicle, determine the second unlocking identification result as the unlocking identification result.

[0220] In some embodiments of this application, the apparatus further includes a training unit;

[0221] The acquisition unit 11 is used to acquire multiple sets of sample RSSIs and multiple sets of sample differential signals corresponding to the multiple sets of sample RSSIs during historical unlocking, as well as sample unlocking operation results corresponding to the multiple sets of sample RSSIs and multiple sets of sample differential signals.

[0222] The training unit is used to train an initial unlocking AI model based on the multiple sets of sample RSSI, the multiple sets of sample differential signals, and the sample unlocking and locking operation results, to obtain the unlocking AI model; the initial unlocking model in the initial unlocking AI model includes an initial convolutional neural network and an initial recurrent neural network; the initial locking model in the initial unlocking AI model includes the initial convolutional neural network and the initial recurrent neural network.

[0223] In some embodiments of this application, the training unit is configured to input the multiple sets of sample RSSIs and the multiple sets of sample difference signals into an initial unlocking AI model to obtain an output unlocking identification result; determine the binary cross-entropy loss of the initial unlocking AI model based on the output unlocking identification result and the sample unlocking operation result; if the binary cross-entropy loss is greater than or equal to a preset loss threshold, continue training the initial unlocking AI model using the multiple sets of sample RSSIs, the multiple sets of sample difference signals, and the sample unlocking operation result to obtain a training model; if the corresponding binary training cross-entropy loss of the training model is less than or equal to the preset loss threshold, use the training model as the unlocking AI model.

[0224] In some embodiments of this application, the execution unit 13 is configured to perform a locking operation when the communication link between the user equipment and the target vehicle is disconnected; or, when the values ​​of the multiple sets of RSSIs are less than or equal to a preset threshold, perform a locking operation.

[0225] It should be noted that, in practical applications, the aforementioned acquisition unit 11, determination unit 12, and execution unit 13 can be implemented by the processor 14 on the vehicle unlocking and locking device, specifically by a CPU (Central Processing Unit), MPU (Microprocessor Unit), DSP (Digital Signal Processor), or Field Programmable Gate Array (FPGA), etc.; the aforementioned data storage can be implemented by the memory 15 on the vehicle unlocking and locking device.

[0226] This application also provides a vehicle unlocking / locking device, such as... Figure 9As shown, the vehicle unlocking and locking device includes a processor 14, a memory 15, and a communication bus 16. The memory 15 communicates with the processor 14 through the communication bus 16. The memory 15 stores programs executable by the processor 14. When the program is executed, the processor 14 executes the vehicle unlocking and locking method applied to the vehicle unlocking and locking device as described above.

[0227] In practical applications, the aforementioned memory 15 can be volatile memory, such as random-access memory (RAM); or non-volatile memory, such as read-only memory (ROM), flash memory, hard disk drive (HDD), or solid-state drive (SSD); or a combination of the above types of memory, and provide instructions and data to the processor 14.

[0228] This application provides a computer program product, which includes a computer program or computer-executable instructions stored in a computer-readable storage medium. A processor of a vehicle unlocking / locking device reads the computer-executable instructions from the computer-readable storage medium and executes the computer-executable instructions, causing the vehicle unlocking / locking device to perform the vehicle unlocking / locking method described above in this application.

[0229] This application provides a computer-readable storage medium storing computer-executable instructions. When these computer-executable instructions are executed by a processor, they cause the processor to execute the vehicle unlocking / locking method provided in this application. For example, ... Figure 1 The method for unlocking and locking the vehicle is shown.

[0230] In some embodiments, the computer-readable storage medium may be a memory such as FRAM, ROM, PROM, EPROM, EEP ROM, flash memory, magnetic surface memory, optical disk, or CD-ROM; or it may be a variety of devices including one or any combination of the above-mentioned memories.

[0231] In some embodiments, computer-executable instructions may take the form of programs, software, software modules, scripts, or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as stand-alone programs or as modules, components, subroutines, or other units suitable for use in a computing environment.

[0232] As an example, computer-executable instructions may, but do not necessarily, correspond to files in a file system. They may be stored as part of a file that holds other programs or data, for example, in one or more scripts in a Hyper Text Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple co-located files (e.g., files that store one or more modules, subroutines, or code sections).

[0233] Understandably, the vehicle unlocking / locking device acquires multiple sets of RSSIs and their corresponding differential signals. Each RSSI is a signal emitted by the user equipment. Using an unlocking / locking AI model, the first unlocking or locking result can be determined based on these multiple RSSIs and differential signals. This unlocking / locking AI model is trained using multiple sample RSSIs and their corresponding differential signals from historical successful unlocking / locking events. This allows the unlocking / locking AI model to accurately determine the first unlocking result based on the multiple RSSIs and their corresponding differential signals. The first unlocking or locking result is used to assist the second unlocking or locking result determined by multiple RSSIs using the unlocking / locking logic judgment rules. A comprehensive judgment is made based on the first unlocking and second unlocking results or the first locking and second locking results to determine the final unlocking / locking result (i.e., the unlocking or locking result). This ensures that even if the RSSI strength of the signal emitted by the user equipment is blocked, it will not affect the prediction result of the unlocking / locking result, thereby improving the accuracy of vehicle unlocking / locking.

Claims

1. A method for unlocking and locking a vehicle, characterized in that, The method includes: Acquire multiple sets of RSSIs and multiple sets of differential signals corresponding to the multiple sets of RSSIs; each set of RSSIs is a signal emitted by the user equipment. The unlocking or locking identification result is determined by using the unlocking AI model and unlocking logic judgment rules, based on the multiple sets of RSSI and the multiple sets of differential signals; wherein, the unlocking AI model is a model trained using multiple sets of sample RSSI and the multiple sets of sample differential signals corresponding to the multiple sets of sample RSSI when the unlocking was successful in the past. Based on the unlocking or locking recognition result, perform the corresponding unlocking or locking operation.

2. The method according to claim 1, characterized in that, The process of determining the locking identification result using the unlocking AI model and unlocking logic judgment rules, based on the multiple sets of RSSI and the multiple sets of differential signals, includes: The first locking identification result is determined by using the locking AI model in the unlocking AI model and based on the multiple sets of RSSI and the multiple sets of differential signals. The second locking identification result is determined based on the unlocking logic determination rules and the multiple sets of RSSIs. The locking identification result is determined based on the first locking identification result and the second locking identification result.

3. The method according to claim 2, characterized in that, The step of determining the second locking identification result based on the unlocking logic determination rule and the multiple sets of RSSIs includes: Obtain the locking logic determination threshold from the unlocking logic determination rule; Determine the comparison results between the multiple sets of RSSI and the latching logic judgment threshold; Based on the comparison results, the second locking identification result is determined.

4. The method according to claim 2, characterized in that, The step of determining the locking identification result based on the first locking identification result and the second locking identification result includes: If both the first and second locking identification results indicate locking, the locking is determined as the locking identification result.

5. The method according to claim 4, characterized in that, When both the first and second locking identification results indicate locking, determining the locking as the locking identification result includes: If the second locking identification result is locking, the first locking state machine is adjusted to the set state; If the first locking identification result is locking, the second locking state machine is adjusted to the set state; When both the first locking state machine and the second locking state machine are in the set state, the locking is taken as the locking identification result.

6. The method according to claim 5, characterized in that, After performing the corresponding locking operation based on the locking identification result, the method further includes: When any door of the target vehicle is in the open state, the states of the first locking state machine and the second locking state machine are adjusted to the non-set state; Alternatively, when the vehicle controller of the target vehicle switches from a dormant state to a wake-up state, the states of the first locking state machine and the second locking state machine are adjusted to a non-set state. Alternatively, if the target vehicle completes the corresponding locking operation, the states of the first locking state machine and the second locking state machine are adjusted to a non-set state.

7. The method according to claim 1, characterized in that, The process of determining the unlocking identification result using the unlocking AI model and unlocking logic judgment rules, based on the multiple sets of RSSI and the multiple sets of differential signals, includes: The multiple sets of RSSI and the multiple sets of differential signals are input into the unlocking AI model in the unlocking AI model to obtain the first unlocking recognition result; The second unlocking identification result is determined based on the unlocking logic determination rules and the multiple sets of RSSIs; If the first unlock recognition result is unlocked, the first unlock recognition result is determined as the unlock recognition result; or if the second unlock recognition result is unlocked, the second unlock recognition result is determined as the unlock recognition result.

8. The method according to claim 7, characterized in that, The method further includes: The location area of ​​the user equipment is determined based on the multiple sets of RSSIs; If the location area is not the rear area of ​​the target vehicle and the first unlock recognition result is unlocked, the first unlock recognition result is determined as the unlock recognition result; or if the location area is not the rear area of ​​the target vehicle and the second unlock recognition result is unlocked, the second unlock recognition result is determined as the unlock recognition result. If the location area is the rear area of ​​the target vehicle, the second unlocking recognition result is determined as the unlocking recognition result.

9. The method according to claim 1, characterized in that, Before determining the unlocking or locking identification result based on the multiple sets of RSSI and the multiple sets of differential signals using the unlocking / locking AI model and unlocking / locking logic judgment rules, the method further includes: Acquire multiple sets of sample RSSIs and corresponding sample differential signals at historical unlocking times, as well as sample unlocking operation results corresponding to the multiple sets of sample RSSIs and sample differential signals; The initial unlocking AI model is trained based on the multiple sets of sample RSSI, the multiple sets of sample differential signals, and the sample unlocking and locking operation results to obtain the unlocking AI model; the initial unlocking model in the initial unlocking AI model includes an initial convolutional neural network and an initial recurrent neural network; the initial locking model in the initial unlocking AI model includes the initial convolutional neural network and the initial recurrent neural network.

10. The method according to claim 9, characterized in that, The step of training an initial unlocking AI model based on the multiple sets of sample RSSI, the multiple sets of sample differential signals, and the sample unlocking operation results to obtain the unlocking AI model includes: The multiple sets of sample RSSI and the multiple sets of sample difference signals are input into the initial unlocking AI model to obtain the output unlocking identification result; Based on the output unlocking identification result and the sample unlocking operation result, determine the binary cross-entropy loss of the initial unlocking AI model; If the binary cross-entropy loss is greater than or equal to a preset loss threshold, the initial unlocking AI model is trained again using the multiple sets of sample RSSI, the multiple sets of sample difference signals, and the sample unlocking operation results to obtain the trained model. If the corresponding binary training cross-entropy loss of the trained model is less than or equal to the preset loss threshold, the trained model is used as the unlocking AI model.

11. The method according to claim 1, characterized in that, The method further includes: If the communication link between the user equipment and the target vehicle is lost, perform the locking operation; Alternatively, if the values ​​of the multiple sets of RSSI are less than or equal to a preset threshold, a latching operation is performed.

12. A vehicle unlocking / locking device, characterized in that, The device includes: The acquisition unit is used to acquire multiple sets of RSSIs and multiple sets of differential signals corresponding to the multiple sets of RSSIs; each set of RSSIs is a signal emitted by the user equipment. The determining unit is used to determine the unlocking identification result or the locking identification result by using the unlocking and locking AI model and the unlocking and locking logic judgment rules, based on the multiple sets of RSSI and the multiple sets of differential signals; wherein, the unlocking and locking AI model is a model trained using multiple sets of sample RSSI and the multiple sets of sample differential signals corresponding to the multiple sets of sample RSSI when the unlocking and locking were successful in the past. The execution unit is used to perform the corresponding unlocking operation or locking operation based on the unlocking identification result or the locking identification result.

13. A vehicle unlocking / locking device, characterized in that, The device includes: Memory is used to store executable instructions for a computer; A processor, when executing computer-executable instructions stored in the memory, implements the method according to any one of claims 1 to 9.

14. A computer-readable storage medium storing computer-executable instructions, characterized in that, When the computer-executable instructions are executed by a processor, they implement the method according to any one of claims 1 to 11.

15. A computer program product comprising computer-executable instructions or a computer program, characterized in that, When the computer-executable instructions or computer program are executed by a processor, they implement the method according to any one of claims 1 to 11.

Images