Vehicle sentry monitoring method and related devices

Abstract

The application discloses a vehicle sentry monitoring method and related equipment, and relates to the technical field of intelligent vehicles. The method comprises the following steps: information interaction is carried out among a plurality of vehicles through short-range communication to construct a temporary vehicle group network; vehicle state information of each vehicle in the temporary vehicle group network is used to allocate monitoring task roles to each vehicle in the temporary vehicle group network; the working parameters of the monitoring hardware of each vehicle in the temporary vehicle group network are determined based on the monitoring task roles of each vehicle in the temporary vehicle group network and the comprehensive risks of the scenes in which the vehicles are located; and in the case that a risk event is detected through the monitoring hardware, early warning information is broadcasted through the temporary vehicle group network to trigger at least part of the vehicles in the temporary vehicle group network to perform cooperative monitoring operations. The application can change the vehicle monitoring from a single independent mode to a group cooperative mode, thereby avoiding repeated configuration of resources and improving the overall resource utilization efficiency and monitoring response capability.

Description

Technical Field

[0001] This application relates to the field of intelligent vehicle technology, and more specifically, to a vehicle sentry monitoring method and related equipment. Background Technology

[0002] With the continuous development of intelligent connected vehicle technology and the ongoing improvement of in-vehicle sensing and computing capabilities, vehicle sentry monitoring has gradually become an important means of ensuring vehicle parking safety and is widely used in urban parking lots, public roads, and complex environments. By continuously sensing and recording the environment around the vehicle, sentry monitoring systems can provide effective evidence in the event of abnormal events, thereby enhancing the vehicle's security capabilities. However, with the increasing complexity of application scenarios and the growing number of vehicles, how to ensure monitoring effectiveness while also considering resource utilization efficiency has become a pressing issue that needs to be addressed in the development of vehicle sentry monitoring technology.

[0003] In related technologies, vehicle sentry monitoring typically operates in a single-vehicle independent mode, with each vehicle relying on its own sensors to continuously perform environmental monitoring and data recording. While this approach is simple to implement, it suffers from problems such as repetitive monitoring tasks and low resource utilization in scenarios with multiple vehicles parked densely. Furthermore, existing solutions often employ fixed monitoring strategies, making it difficult to dynamically adjust based on different vehicle states and environmental risks, easily leading to uneven distribution of monitoring resources. In other words, related technologies suffer from insufficient vehicle monitoring collaboration capabilities, low resource utilization efficiency, and poor risk response capabilities. Summary of the Invention

[0004] In the summary section of this application, the relevant technical solutions are described in general terms, and a series of simplified concepts are introduced. These concepts will be further elaborated in the detailed embodiments section. This summary section should not be construed as limiting the key or essential technical features of the claimed solutions, nor is it intended to limit the scope of protection of the claimed solutions.

[0005] The vehicle sentry monitoring method and related equipment provided in this application can realize information sharing among vehicles by building a vehicle group network, and perform differentiated task allocation and dynamic adjustment of hardware parameters based on vehicle status and scenario risks. At the same time, it can trigger multi-vehicle collaborative response when risk events occur, which can transform vehicle monitoring from a single independent mode to a group collaborative mode, thereby improving the overall resource utilization efficiency and monitoring response capability while avoiding redundant resource configuration.

[0006] In a first aspect, this application provides a vehicle sentinel monitoring method, comprising: exchanging information among multiple vehicles via short-range communication to construct a temporary vehicle group network; assigning monitoring task roles to each vehicle in the temporary vehicle group network based on the vehicle status information of each vehicle in the temporary vehicle group network, wherein the monitoring task roles include a primary monitoring role, an auxiliary sensing role, and a dormant role; determining the operating parameters of the monitoring hardware of each vehicle in the temporary vehicle group network based on the monitoring task roles of each vehicle in the temporary vehicle group network and the comprehensive risk of the scene in which they are located; and broadcasting early warning information through the temporary vehicle group network when a risk event is detected by the monitoring hardware, thereby triggering at least some vehicles in the temporary vehicle group network to perform collaborative monitoring operations.

[0007] In some implementations, the vehicle status information includes battery status, location information, and sensor capability information; the step of assigning monitoring task roles to each vehicle in the temporary vehicle group network based on the vehicle status information of each vehicle in the temporary vehicle group network includes: generating functional quantification parameters for each vehicle in the temporary vehicle group network based on the battery status, location information, and sensor capability information of each vehicle in the temporary vehicle group network; assigning vehicles in the temporary vehicle group network whose functional quantification parameters are greater than or equal to a first parameter threshold as the primary monitoring role, assigning vehicles whose functional quantification parameters are greater than or equal to a second parameter threshold and less than the first parameter threshold as the auxiliary perception role, and assigning vehicles whose functional quantification parameters are less than the second parameter threshold as the dormant role.

[0008] In some implementations, determining the operating parameters of the monitoring hardware for each vehicle in the temporary vehicle group network based on the monitoring task role of each vehicle and the comprehensive risk of the scenario includes: querying a monitoring strategy mapping table to obtain a set of control parameters for each vehicle in the temporary vehicle group network based on the monitoring task role and the comprehensive risk of each vehicle in the temporary vehicle group network, wherein the monitoring strategy mapping table is used to characterize the correspondence between the comprehensive risk, the monitoring task role, and the set of control parameters, and the set of control parameters includes at least one of camera parameters, radar sensor parameters, and main control chip parameters, wherein the camera parameters include the camera's resolution, frame rate, sensitivity, and wake-up duty cycle, the radar sensor parameters include the radar sensor's detection frequency and transmission power, and the main control chip parameters include the main control chip's operating voltage and operating frequency; and adjusting the operating parameters of the monitoring hardware for each vehicle in the temporary vehicle group network based on the set of control parameters of each vehicle in the temporary vehicle group network.

[0009] In some implementations, the warning information is issued by a first vehicle that detects the risk event, and the warning information includes the location of the risk event. Broadcasting the warning information through the temporary vehicle network to trigger at least some vehicles in the temporary vehicle network to perform collaborative monitoring operations includes: a second vehicle assigned to the dormant role switching from a dormant state to an awakened state in response to receiving the warning information; the first vehicle, the awakened second vehicle, and the third vehicle assigned to the auxiliary perception role adjusting the acquisition angle of the onboard sensors based on the event location to collect sensor data of the risk event from different angles; and fusing the sensor data collected by the first vehicle, the second vehicle, and the third vehicle to obtain a multi-view joint evidence package for the risk event.

[0010] In some implementations, the step of determining the comprehensive risk for any fourth vehicle in the temporary vehicle network includes: determining a first risk coefficient based on geofence information corresponding to the geographical location of the fourth vehicle; identifying abnormal acoustic events in the environmental sound signals of the fourth vehicle, and determining a second risk coefficient based on the type and intensity of the abnormal acoustic events; determining the population density and movement speed around the fourth vehicle based on the visual perception data of the fourth vehicle, and determining a third risk coefficient based on the population density and movement speed; and weighting and fusing the first risk coefficient, the second risk coefficient, and the third risk coefficient to obtain the comprehensive risk of the scene in which the fourth vehicle is located.

[0011] In some embodiments, the vehicle sentry monitoring method further includes: acquiring a feedback label of the risk event and a risk feature vector that triggered the risk event, wherein the feedback label includes a false alarm label and a real threat label; associating and storing the feedback label and the risk feature vector in a local sample library of a first vehicle; when the number of samples in the local sample library reaches a preset sample threshold and the first vehicle is in a charging or dormant state, incrementally training the local risk perception model of the first vehicle based on the samples in the local sample library to obtain a first model parameter update amount; updating the local risk perception model according to the first model parameter update amount, wherein the local risk perception model is used to determine the comprehensive risk of the first vehicle.

[0012] In some embodiments, the vehicle sentry monitoring method further includes: for any fifth vehicle in the temporary vehicle group network, training a local collaborative strategy model of the fifth vehicle based on security event data stored locally by the fifth vehicle to obtain a second model parameter update amount, wherein the security event data is risk event record data generated by the fifth vehicle during historical monitoring, and the local collaborative strategy model is used to implement the allocation of monitoring task roles for the fifth vehicle; encrypting the second model parameter update amount to obtain an encryption gradient; transmitting the encryption gradient to a coordination node in the temporary vehicle group network, so that the coordination node can aggregate the encryption gradients from multiple vehicles in the temporary vehicle group network to generate model parameters of a global collaborative strategy model, wherein the global collaborative strategy model is used to implement the allocation of monitoring task roles for each vehicle in the temporary vehicle group network; and updating the local collaborative strategy model of the fifth vehicle based on the model parameters returned by the coordination node.

[0013] Secondly, this application also provides a vehicle sentinel monitoring device, comprising: a network establishment unit for exchanging information among multiple vehicles via short-range communication to construct a temporary vehicle group network; a role allocation unit for assigning monitoring task roles to each vehicle in the temporary vehicle group network based on the vehicle status information of each vehicle in the temporary vehicle group network, wherein the monitoring task roles include a primary monitoring role, an auxiliary sensing role, and a dormant role; a parameter determination unit for determining the operating parameters of the monitoring hardware of each vehicle in the temporary vehicle group network based on the monitoring task roles of each vehicle in the temporary vehicle group network and the comprehensive risk of the scene in which they are located; and a monitoring execution unit for broadcasting early warning information through the temporary vehicle group network when a risk event is detected by the monitoring hardware, thereby triggering at least some vehicles in the temporary vehicle group network to perform collaborative monitoring operations.

[0014] Thirdly, this application also provides a vehicle, including: a memory and a processor, the processor being configured to execute a computer program stored in the memory to implement the steps of the vehicle sentry monitoring method described in the first aspect.

[0015] Fourthly, this application also provides a computer-readable storage medium storing computer-executable instructions or a computer program, which, when executed by a processor, implement the steps of the vehicle sentry monitoring method described in the first aspect.

[0016] Fifthly, this application also provides a computer program product, including a computer program or computer-executable instructions, which, when executed by a processor, implement the steps of the vehicle sentry monitoring method provided in the embodiments of this application.

[0017] In summary, this application establishes information exchange among multiple vehicles through short-range communication, thereby constructing a temporary vehicle group network. This transforms the vehicles from independent monitoring entities into a collaborative whole capable of sharing information, providing the foundation for unified scheduling and collaborative operation of previously dispersed monitoring resources. By assigning primary monitoring, auxiliary sensing, and dormant roles to each vehicle based on its status information, different vehicles undertake differentiated monitoring tasks. This shifts the approach from "all vehicles simultaneously performing the same type of monitoring task" to "division of labor based on capability and status," avoiding redundant configuration of monitoring tasks and making resource allocation within the vehicle group more rational and orderly. Furthermore, considering the specific tasks undertaken by each vehicle... Based on the comprehensive risks of the monitoring tasks undertaken by each vehicle and the surrounding scenarios, the operating parameters of the monitoring hardware are determined to match the hardware's operating status with actual monitoring needs. This allows high-risk or critical roles to receive higher monitoring intensity, while low-risk or non-critical roles receive lower workload, thereby reducing unnecessary resource consumption. When any vehicle detects a risk event through its monitoring hardware, a warning message is broadcast via the vehicle group network, triggering at least some vehicles to participate in collaborative monitoring. This expands single-vehicle response to multi-vehicle coordinated response, improving the response range and collaborative capability for risk events. The monitoring process transforms from local perception to group collaborative processing, enhancing the overall monitoring effect. In summary, the vehicle sentinel monitoring method provided in this application achieves information sharing among vehicles by constructing a vehicle group network, and performs differentiated task allocation and dynamic hardware parameter adjustment based on vehicle status and scenario risks. Simultaneously, it triggers multi-vehicle collaborative response when a risk event occurs, transforming vehicle monitoring from a single independent mode to a group collaborative mode. This avoids redundant resource configuration while improving overall resource utilization efficiency and monitoring response capability. Attached Figure Description

[0018] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit this specification. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart illustrating a vehicle sentry monitoring method provided in an embodiment of this application; Figure 2 This is a schematic diagram of the composition structure of a vehicle sentry monitoring device provided in an embodiment of this application; Figure 3 This is a schematic diagram of the composition structure of a vehicle provided in an embodiment of this application. Detailed Implementation

[0019] The terms used in the specification, claims, and drawings of this application, such as "first," "second," "third," "fourth," etc. (if any), are used to distinguish similar objects and not to describe a specific order or sequence. Therefore, it is to be understood that these terms can be used interchangeably where appropriate, allowing the described embodiments to be used in different orders, unless specifically required by the illustrations or description. Furthermore, the terms "is" and "has," and any variations thereof, are intended to cover, non-exclusively, all possible constituent elements. For example, a process, method, system, product, or apparatus comprising several steps or units is not necessarily limited to the steps or units explicitly listed, but may also include other steps or units not explicitly listed, or steps or units inherent to the process, method, product, or apparatus.

[0020] In this application, a "module" or "unit" refers to a computer program or part of a computer program that has a specific function and works in conjunction with other related parts to achieve a predetermined goal. These modules or units can be implemented by software, hardware (e.g., processing circuitry or memory), or a combination of both. One or more processors or memories can implement one or more modules or units. Furthermore, each module or unit can also be part of a larger module or unit.

[0021] The technical solutions of this application will be described in detail below with reference to the accompanying drawings of the embodiments. It should be noted that the described embodiments are only a part of this application, and not all embodiments. In the following description, the "some embodiments" mentioned are only a subset of all possible embodiments, which may be the same or different subsets, and different embodiments can be combined with each other without conflict.

[0022] Figure 1 This is a flowchart illustrating a vehicle sentry monitoring method provided in an embodiment of this application. For example, see [link to example]. Figure 1 The vehicle sentry monitoring method provided in this application embodiment may include the following steps 101 to 104: Step 101: Information is exchanged between multiple vehicles through short-range communication to build a temporary vehicle group network.

[0023] In some examples, short-range communication (SMR) is a communication technology that does not rely on fixed infrastructure such as cloud servers and ground base stations, and can only realize data interaction between devices within a short distance. Its role is to provide low-latency and high-reliability communication support for real-time information interaction between multiple vehicles, without the need for additional communication infrastructure. Each vehicle participating in the monitoring can be equipped with a corresponding SMR module, which is integrated into the vehicle's onboard control system and can automatically start or go into sleep mode according to the actual application scenario. The SMR technology in this application embodiment can use Bluetooth (BT), NearLink (NL), Ultra Wide Band (UWB), and Cellular Vehicle-to-Everything (C-V2X) short-range communication, etc. Among them, C-V2X short-range communication is the preferred short-range communication method in this application embodiment due to its low latency and high reliability. For example, each vehicle is equipped with a C-V2X short-range communication module, which can realize real-time information interaction between vehicles within a range of 100 meters. Multiple vehicles are located in the same sentinel monitoring scenario (such as a parking lot, open-air parking area, etc.), and are all equipped with the control system corresponding to the vehicle sentinel monitoring method of this application embodiment. Two or more intelligent vehicles with short-range communication capabilities have the basic conditions to participate in the construction of a temporary vehicle group network, complete information interaction, and collaborative monitoring. After each vehicle enters the same monitoring scenario, it can broadcast its own presence signal and basic equipment information through its own short-range communication module. Other vehicles in the scenario with corresponding capabilities can receive the signal and identify surrounding vehicles that meet the conditions, forming a collection of multiple vehicles. For example, in an open-air parking lot of an office building, three intelligent electric vehicles drive in and park one after another. All three vehicles are equipped with the monitoring and control system and vehicle network short-range communication module corresponding to the embodiment of this application. These three vehicles constitute multiple vehicles in the embodiment of this application.

[0024] A temporary vehicle swarm network is a decentralized collaborative network formed autonomously by multiple vehicles after completing information exchange, identity verification, and link establishment through short-range communication. This network only exists temporarily while the vehicles are in the same monitoring scenario and need to collaboratively perform sentinel monitoring tasks. When a vehicle leaves the monitoring scenario, completes its monitoring task, or the short-range communication link is interrupted, the network automatically disbands. Multiple vehicles can complete mutual identification, communication capability matching, and basic information exchange through short-range communication modules. Based on a preset self-organizing network protocol, they can autonomously negotiate to establish communication links and allocate temporary network identifiers, thereby forming a temporary collaborative network that enables information sharing and command transmission. For example, the three intelligent electric vehicles in the parking lot mentioned above autonomously form a temporary network for collaboratively performing sentinel monitoring tasks after completing information exchange through vehicle-to-everything (V2X) short-range communication. When one of the vehicles leaves the parking lot, the temporary vehicle swarm network automatically adjusts its network structure, retaining only the communication links of the remaining two vehicles. If all the remaining vehicles leave, the network automatically disbands.

[0025] For example, when the first intelligent vehicle enters the parking lot and parks itself, its onboard control system automatically activates its vehicle-to-everything (V2X) short-range communication module, broadcasting its vehicle identification, V2X module model, and monitoring system readiness status to a radius of 50 to 100 meters. Subsequently, the second and third intelligent vehicles enter the parking lot and park themselves. Their onboard control systems also automatically activate their V2X modules, receiving the information broadcast by the first vehicle and immediately providing feedback on their vehicle identification and communication capabilities. Upon receiving this feedback, the first vehicle verifies the identities of both vehicles, confirming their readiness for collaborative monitoring. Afterward, the three vehicles continuously interact through their V2X modules, negotiating and determining their respective temporary network node identifiers, establishing interconnected communication links, and defining the protocol specifications for information exchange. The construction of a temporary vehicle network can be completed in just 5 to 10 seconds, laying the foundation for subsequent monitoring task assignments and collaborative monitoring operations.

[0026] By implementing step 101, short-range communication is used to establish information interaction among multiple vehicles, thereby constructing a temporary vehicle group network. This enables vehicles that were originally operating independently to form a collaborative whole that can share information, providing the basic conditions for collaborative decision-making and joint monitoring among vehicles, and enabling dispersed monitoring resources to have the ability to be organized and utilized in a unified manner.

[0027] Step 102: Based on the vehicle status information of each vehicle in the temporary vehicle group network, assign monitoring task roles to each vehicle in the temporary vehicle group network.

[0028] Among them, the monitoring task roles include the main monitoring role, the auxiliary sensing role, and the dormant role.

[0029] In some examples, the vehicle status information of each vehicle in a temporary vehicle group network is the basic status data that is acquired and synchronized in real time by the on-board acquisition unit of each intelligent vehicle in the network. This type of information is the basis for assigning monitoring task roles and does not involve additional quantitative calculation parameters. For example, networked vehicles in the same open-air parking lot collect and synchronize their own remaining battery status, actual parking location information, and basic capability information of on-board monitoring sensors to form a complete set of vehicle status information.

[0030] Monitoring roles are differentiated monitoring assignments given to vehicles within a temporary vehicle network to achieve collaborative sentinel monitoring and rationally allocate vehicle hardware and energy resources. All roles are defined around the goals of collaborative monitoring and energy-saving management. The temporary vehicle network can match and assign roles based on the status information uploaded by each vehicle through decentralized autonomous negotiation logic. The assignment results are synchronized to each networked vehicle via short-range communication. For example, multiple intelligent vehicles within the same temporary vehicle network will be assigned to corresponding monitoring roles, and each vehicle will perform its specific monitoring duties according to its assigned role. The primary monitoring role is the task role within the temporary vehicle network that bears the main monitoring responsibility, conducting real-time monitoring of the entire core area. The temporary vehicle network can select vehicles suitable for the core monitoring needs based on vehicle status information and assign them to this role. Vehicles in this role must maintain the normal operation of their basic monitoring hardware. For example, an intelligent vehicle parked in the core area of ​​a parking lot with complete monitoring sensor configuration will be assigned the primary monitoring role, undertaking the main sentinel monitoring tasks of the vehicle group. The auxiliary sensing role is a collaborative task role that works with the main monitoring role to fill monitoring blind spots and provide auxiliary monitoring of the surrounding area. It is used to improve the overall monitoring coverage of the vehicle group and avoid monitoring loopholes. The temporary vehicle group network can select vehicles suitable for auxiliary monitoring needs based on vehicle status information and assign them to this role. These vehicles maintain low-load operation of their basic sensing hardware. For example, intelligent vehicles parked at the edge of a parking lot that can cover the blind spots of the main monitoring vehicles will be assigned to the auxiliary sensing role to assist in monitoring the surrounding environment. The dormant role is an energy-saving task role that only retains short-range communication listening functions and disables unnecessary monitoring hardware. Its function is to reduce vehicle energy consumption while maintaining the connectivity of the vehicle group network. The temporary vehicle group network can assign vehicles suitable for energy-saving management needs to this role based on vehicle status information. In the dormant state, vehicles do not actively perform monitoring operations. For example, vehicles with weak hardware load capacity or that do not need to undertake core and auxiliary monitoring tasks will be assigned to the dormant role, only maintaining a low-power listening state for their communication modules.

[0031] For example, each intelligent vehicle in a temporary vehicle group network first collects its own status information in real time through its onboard acquisition module, and then synchronizes the status information to all vehicles in the network through cellular vehicle-to-everything (V2X) communication, achieving full-domain sharing of status information. Subsequently, the vehicle group initiates an autonomous negotiation process, determining the appropriate role based on the status information of each vehicle, and sequentially assigning the main monitoring role, auxiliary sensing role, and dormant role. After the assignment is completed, the vehicle group sends the final role instructions to the corresponding vehicles through short-range communication. After receiving the instructions, each vehicle confirms its own monitoring task role and prepares for subsequent monitoring execution.

[0032] By implementing step 102, each vehicle is assigned a primary monitoring role, an auxiliary sensing role, and a dormant role based on its vehicle status information. This allows different vehicles to undertake differentiated monitoring tasks, transforming the original homogeneous monitoring method into a collaborative approach. This avoids multiple vehicles performing repetitive monitoring tasks, making the allocation of resources within the vehicle group more rational and improving overall resource utilization efficiency.

[0033] Step 103: Based on the monitoring task role of each vehicle in the temporary vehicle group network and the comprehensive risks of the scenario, determine the working parameters of the monitoring hardware of each vehicle in the temporary vehicle group network.

[0034] In some examples, comprehensive risk refers to the overall security risk situation of the monitoring scenario (such as a parking lot or open-air parking area) where each vehicle in the temporary vehicle group network is located. It is used to characterize the probability and severity of security incidents that may occur in the scenario. For example, in an open-air parking lot of an office building, the networked vehicles collect basic data such as the flow of people in the scenario and abnormal environmental signals. Each vehicle determines its current comprehensive risk status in the parking lot.

[0035] The operating parameters of the monitoring hardware refer to the operational configuration standards that the monitoring-related hardware devices on each vehicle in the temporary vehicle network must follow to complete their respective monitoring tasks. These parameters directly determine the operating status, monitoring capabilities, and energy consumption levels of the monitoring hardware. The operating parameters of the monitoring hardware can be determined autonomously through the temporary vehicle network based on the comprehensive risks of each vehicle's monitoring task role and the surrounding scenario. Once determined, the operating parameters are sent to the monitoring hardware control unit of the corresponding vehicle to regulate the hardware operation. For example, the monitoring hardware includes vehicle-mounted cameras, millimeter-wave radar, and main control chips. The corresponding operating parameters include the operating status of the cameras, the detection intensity of the radar, and the operating power of the main control chip. The operating parameters of each piece of hardware are set differently under different monitoring task roles and different comprehensive risks.

[0036] For example, the temporary vehicle network first summarizes and determines the comprehensive risks of the scenario, clarifies the monitoring role of each vehicle, and then, based on the responsibilities of different monitoring roles and the level of comprehensive risk in the current scenario, formulates an appropriate operating configuration standard for the monitoring hardware of each vehicle. This ensures that the operating status of the monitoring hardware can meet the monitoring needs of the corresponding role while reasonably controlling energy consumption. The entire determination process is completed autonomously by the temporary vehicle network without relying on external equipment intervention. For example, in a medium-risk scenario, the monitoring hardware operating parameters of vehicles assigned to the primary monitoring role will be set to a higher operating standard to ensure monitoring effectiveness, while the monitoring hardware operating parameters of vehicles assigned to the dormant role will be set to a low-power operating standard to save energy.

[0037] By implementing step 103, the working parameters of the monitoring hardware are determined by combining the vehicle's monitoring task role and the comprehensive risk of the scene. This allows the hardware's operating status to be dynamically adjusted according to actual monitoring needs, enabling higher monitoring intensity for high-risk or critical roles and lower monitoring intensity for low-risk or non-critical roles. This reduces unnecessary resource consumption while ensuring that monitoring needs are met.

[0038] Step 104: If a risk event is detected by the monitoring hardware, a warning message is broadcast through the temporary vehicle group network to trigger at least some vehicles in the temporary vehicle group network to perform collaborative monitoring operations.

[0039] In some examples, risk events are various abnormal events detected by vehicle monitoring hardware that may threaten the safety of vehicles within the temporary vehicle ensemble network. These events pose potential security risks and require triggering a coordinated response from the vehicle ensemble to handle the risks and collect evidence. Each vehicle in the temporary vehicle ensemble network can use its own onboard monitoring hardware to collect environmental data of its surroundings in real time. The collected data can be analyzed in real time, and when the detected data meets the preset risk judgment criteria, a risk event is determined to have been detected. For example, if the monitoring hardware detects abnormal approach of people around the vehicle, a collision with the vehicle body, or abnormal noises, these are all risk events in this step. Among these, a collision with the vehicle body is one of the most common types of risk events.

[0040] Warning information is a notification generated by a vehicle that detects a risk event, used to inform other vehicles in the temporary vehicle group network of the existence of a risk. Its function is to transmit risk-related information and trigger a coordinated response from the vehicle group. Warning information can be automatically generated by a vehicle that detects a risk event through its own onboard control unit. This information contains the basic information of the risk event and, after generation, does not require external equipment review. It is directly sent to the temporary vehicle group network through a short-range communication module. For example, a vehicle that detects a vehicle collision automatically generates warning information that includes the time of the risk and the approximate location of the risk. This information serves as the carrier for triggering coordinated monitoring operations.

[0041] At least some vehicles in a temporary vehicle convoy network refer to the networked vehicles that, excluding those that detected a risk event, need to participate in collaborative monitoring operations based on the actual situation of the risk event. This could also include all vehicles in the network, which possess the ability to respond to early warning information and perform collaborative monitoring. After the information is broadcast in the temporary vehicle convoy network, vehicles suitable for collaborative monitoring can be autonomously selected based on their monitoring task roles and current operating status. The selection process is completed autonomously by the vehicle convoy without external intervention. For example, if a temporary vehicle convoy network contains three networked vehicles, after detecting a risk event, the convoy selects the primary monitoring vehicle and the auxiliary sensing vehicle to participate in collaborative monitoring. These two vehicles constitute at least some of the vehicles in the temporary vehicle convoy network.

[0042] The collaborative monitoring operation is a series of actions performed by vehicles that detect a risk event and at least some of the triggered vehicles in a temporary vehicle network to monitor, collect evidence, and handle the risk. Its purpose is to comprehensively cover the risk event scenario and improve the effectiveness and comprehensiveness of risk handling. After receiving the warning information, the triggered vehicles automatically initiate corresponding monitoring operations based on the warning information. The operations of each vehicle cooperate and coordinate to form a complete collaborative monitoring closed loop. For example, the vehicle that detects the risk event continuously collects data from the risk site, while the triggered vehicles adjust their monitoring angles and supplement the data collected, jointly completing comprehensive monitoring and evidence collection of the risk event. This series of coordinated operations constitutes the collaborative monitoring operation.

[0043] For example, a vehicle with an auxiliary sensing role in a temporary vehicle group network has its monitoring hardware collecting real-time data on the surrounding environment. When it detects a person holding a foreign object abnormally approaching an adjacent vehicle, it immediately determines that a risk event has been detected. The vehicle quickly generates a warning message containing the location of the risk and the type of risk event, and broadcasts the warning message to the entire temporary vehicle group network via cellular vehicle-to-everything (V2X) communication. The temporary vehicle group network autonomously selects a primary monitoring vehicle and another auxiliary sensing vehicle as vehicles to participate in collaborative monitoring. After receiving the warning message, these two vehicles immediately adjust the operating status of their monitoring hardware. The primary monitoring vehicle focuses on collecting high-definition data in the area where the risk occurred, while the other auxiliary sensing vehicle supplements the data collection from the side. Working together with the vehicle that detected the risk event, they complete comprehensive monitoring and evidence collection of the risk event, ensuring that the risk event is traceable.

[0044] By implementing step 104, when any vehicle detects a risk event, a warning message is broadcast through a temporary vehicle group network, triggering other vehicles to participate in collaborative monitoring. This expands the monitoring response from single-vehicle processing to multi-vehicle linkage processing, thereby increasing the monitoring coverage and improving the response capability and coordination level to risk events, and enhancing the overall monitoring effect.

[0045] In summary, this application embodiment establishes information interaction among multiple vehicles through short-range communication, thereby constructing a temporary vehicle group network. This transforms the vehicles from being in an independent monitoring state into a collaborative whole capable of sharing information, providing the foundation for unified scheduling and collaborative work of previously dispersed monitoring resources. Based on the vehicle status information, each vehicle is assigned a primary monitoring role, an auxiliary sensing role, and a dormant role, allowing different vehicles to undertake differentiated monitoring tasks. This shifts the approach from "all vehicles simultaneously performing the same type of monitoring task" to "division of labor based on capability and status," avoiding redundant configuration of monitoring tasks and making resource allocation within the vehicle group more rational and orderly. Combined with the vehicle's... The monitoring roles undertaken and the comprehensive risks of the scenarios in which they occur are considered to determine the operating parameters of the monitoring hardware for each vehicle. This ensures that the hardware's operating status matches the actual monitoring needs, allowing high-risk or critical roles to receive higher monitoring intensity, while low-risk or non-critical roles receive lower workload, thereby reducing unnecessary resource consumption. When any vehicle detects a risk event through its monitoring hardware, a warning message is broadcast through the vehicle group network, triggering at least some vehicles to participate in collaborative monitoring. This expands single-vehicle response to multi-vehicle coordinated response, improving the response range and collaborative capabilities for risk events. The monitoring process transforms from local perception to group collaborative processing, enhancing the overall monitoring effect. In summary, the vehicle sentinel monitoring method provided in this application achieves information sharing among vehicles by constructing a vehicle group network, and performs differentiated task allocation and dynamic hardware parameter adjustment based on vehicle status and scenario risks. Simultaneously, it triggers multi-vehicle collaborative responses when risk events occur, transforming vehicle monitoring from a single independent mode to a group collaborative mode. This avoids redundant resource configuration while improving overall resource utilization efficiency and monitoring response capabilities.

[0046] In some embodiments, the aforementioned vehicle status information may include battery status, location information, and sensor capability information; the aforementioned step 102 may include: generating functional quantification parameters for each vehicle in the temporary vehicle group network based on the battery status, location information, and sensor capability information of each vehicle in the temporary vehicle group network; assigning vehicles in the temporary vehicle group network whose functional quantification parameters are greater than or equal to a first parameter threshold to the primary monitoring role, assigning vehicles whose functional quantification parameters are greater than or equal to a second parameter threshold and less than the first parameter threshold to the auxiliary perception role, and assigning vehicles whose functional quantification parameters are less than the second parameter threshold to the dormant role.

[0047] In some examples, the battery status refers to the remaining battery power of each vehicle in a temporary vehicle network. This directly determines the duration for which a vehicle can continuously perform monitoring tasks and the hardware load it can handle. It is one of the bases for generating functional quantification parameters and assigning monitoring task roles. Each vehicle can collect the remaining battery power data in real time through its Battery Management System (BMS) as its battery status. For example, if a smart vehicle in a temporary vehicle network collects data from its BMS showing a current remaining battery power of 85%, this data represents the vehicle's battery status, which can support it in performing high-load monitoring tasks for extended periods. Location information refers to the specific parking location and spatial coordinates of each vehicle in a temporary vehicle network within the current monitoring scenario. It is used to determine whether a vehicle can cover blind spots or is located in the core monitoring area of ​​the scenario, and serves as an important reference for role allocation. Each vehicle can collect its real-time location data through the Global Positioning System (GPS) or the BeiDou Navigation Satellite System (BDS), and combine it with the preset geographical coordinate range of the scenario to determine its specific location within the monitoring scenario, which serves as its location information. For example, in an open-air parking lot of an office building, an intelligent vehicle collects its own location coordinates through the GPS, and combined with the preset geographical range of the parking lot, determines that it is parked in the core monitoring area of ​​the parking lot. This coordinate information and area affiliation constitute the vehicle's location information. Sensor capability information comprises the performance parameters, operating status, and achievable monitoring functions of the monitoring-related sensors carried by each vehicle in a temporary vehicle network. It is used to determine the intensity of monitoring tasks that a vehicle can undertake and is a key basis for generating functional quantification parameters. Each vehicle can monitor the operating status of the monitoring sensors in real time through its onboard control system, collect the sensor performance parameters, and integrate them to form sensor capability information. For example, if a vehicle is equipped with a high-definition onboard camera and millimeter-wave radar, and both types of sensors are in normal operating condition, enabling high-definition image acquisition and close-range obstacle detection, the type, performance, and operating status of these sensors together constitute the sensor capability information of that vehicle.

[0048] Functional quantification parameters are single values ​​calculated using a pre-defined quantification algorithm based on the battery status, location information, and sensor capability information of each vehicle. These parameters serve to uniformly measure the comprehensive monitoring capabilities of each vehicle, providing a quantitative basis for the precise allocation of monitoring roles. A temporary vehicle network can aggregate synchronized battery status, location information, and sensor capability information from each vehicle. Through decentralized quantification logic, corresponding weights are assigned to each of the three types of information, and a comprehensive calculation is performed to obtain the functional quantification parameters for each vehicle. The calculation process is completed autonomously by the vehicle network without external intervention. The pre-defined quantification algorithm is pre-written and embedded in the vehicle's onboard control system program module. It also adapts to the standardized calculation rules of the decentralized negotiation logic of the temporary vehicle network. This algorithm transforms the three types of multi-dimensional and different-level vehicle status information—battery status, location information, and sensor capability information—into a unified-dimensional standard score and performs weighted calculations, ultimately outputting a single numerical functional quantification parameter. For example, the battery status, location information, and sensor capability information can first be normalized, converting each type of information into a standard score within the range of 0 to 100. Then, the three standard scores are weighted according to a pre-defined weight ratio. A weighted summation is performed, with weight allocation following the principle of matching monitoring task requirements with vehicle capabilities, and the sum of all weight values ​​equals 1. The final result obtained by weighted summation is the functional quantification parameter. For example, according to this algorithm, the weight of battery status is set to 0.4, the weight of location information is set to 0.3, and the weight of sensor capability information is set to 0.3. The standard score of battery status (90), the standard score of location information (85), and the standard score of sensor capability information (80) of a certain vehicle are calculated to obtain 90×0.4+85×0.3+80×0.3=85.5. This value is the functional quantification parameter calculated by the preset quantification algorithm.

[0049] The first parameter threshold is a pre-set critical value of the temporary vehicle group network used to distinguish between the main monitoring role and the auxiliary perception role. Its value is higher than the second parameter threshold and is used to define the range of vehicles capable of undertaking core monitoring tasks. The temporary vehicle group network can autonomously negotiate and set the first parameter threshold according to the scale of the current monitoring scenario and the intensity of monitoring needs. This threshold can be dynamically adjusted according to changes in the scenario, but remains fixed within the same monitoring cycle. For example, in a monitoring scenario of an open-air parking lot, the temporary vehicle group network can autonomously set the first parameter threshold to 80 based on the monitoring needs of the scenario. Vehicles whose functional quantification parameters reach or exceed this value have the comprehensive ability to undertake the main monitoring role. The second parameter threshold is a pre-set critical value of the temporary vehicle group network used to divide auxiliary perception roles and dormant roles. Its value is lower than the first parameter threshold and is used to define the range of vehicles capable of undertaking auxiliary monitoring tasks. The temporary vehicle group network can autonomously negotiate and set the second parameter threshold in combination with scene monitoring needs and vehicle energy consumption control targets, and adjust it synchronously with the first parameter threshold to ensure the rationality of role division. For example, corresponding to the first parameter threshold of 80, the temporary vehicle group network sets the second parameter threshold to 60. Vehicles whose functional quantification parameters reach or exceed this value but are lower than the first parameter threshold have the comprehensive ability to undertake auxiliary perception roles, while vehicles with values ​​lower than this value are suitable for dormant roles.

[0050] After generating the functional quantification parameters for all vehicles, the functional quantification parameters of each vehicle can be compared with the first parameter threshold and the second parameter threshold respectively. Based on the comparison results, the monitoring task roles are accurately assigned. Vehicles whose functional quantification parameters reach or exceed the first parameter threshold are assigned to the primary monitoring role and undertake core monitoring tasks. Vehicles whose functional quantification parameters are between the second parameter threshold and the first parameter threshold are assigned to the auxiliary perception role and cooperate with the primary monitoring role to fill monitoring blind spots. Vehicles whose functional quantification parameters are below the second parameter threshold are assigned to the dormant role and retain only the communication listening function to save energy. The entire allocation process is completed autonomously by the vehicle group, and the allocation results are synchronized to each vehicle. For example, if the functional quantification parameters of three vehicles in a temporary vehicle group network are 85, 70, and 55 respectively, combined with the first parameter threshold of 80 and the second parameter threshold of 60, the vehicle with parameter 85 is assigned to the primary monitoring role, the vehicle with parameter 70 is assigned to the auxiliary perception role, and the vehicle with parameter 55 is assigned to the dormant role.

[0051] For example, after the temporary vehicle group network is constructed, each vehicle collects its own battery status, location information, and sensor capability information through the vehicle battery management system, global positioning system, and vehicle control system, respectively, and synchronizes them to the entire vehicle group via cellular vehicle-to-everything (V2X) communication. After the temporary vehicle group network aggregates the three types of information from all vehicles, it initiates a preset quantitative calculation logic, setting the weights of battery status, location information, and sensor capability information to 40%, 30%, and 30%, respectively. The information of each vehicle is standardized and weighted and summed to generate functional quantitative parameters for each vehicle. Subsequently, the vehicle group, based on the current parking lot monitoring requirements, autonomously negotiates and sets a first parameter threshold of 80 and a second parameter threshold of 60, comparing the functional quantitative parameters of each vehicle with the two thresholds. Finally, the two vehicles with functional quantitative parameters of 83 and 81 are assigned to the primary monitoring role, the three vehicles with functional quantitative parameters of 75, 68, and 62 are assigned to the auxiliary perception role, and the two vehicles with functional quantitative parameters of 58 and 52 are assigned to the dormant role. After the assignment is completed, the vehicle group sends role instructions to each vehicle via short-range communication. After receiving the instructions, each vehicle confirms its own responsibilities and prepares for monitoring.

[0052] By implementing the above embodiments, vehicle capabilities are quantified by introducing power status, location information, and sensor capability information. Based on functional quantification parameters and preset thresholds, roles are automatically assigned, transforming role allocation from subjective or fixed rules to quantitative decisions based on objective indicators. This enables more accurate matching of vehicle capabilities with monitoring task requirements, avoiding low-capability vehicles from undertaking high-load tasks or high-capability vehicles from being idle. This further improves the efficiency of vehicle fleet resource utilization and enhances the rationality and stability of overall monitoring collaboration.

[0053] In some embodiments, step 103 may include: querying a monitoring strategy mapping table to obtain a set of control parameters for each vehicle in the temporary vehicle group network based on the monitoring task role and comprehensive risk of each vehicle in the temporary vehicle group network. The monitoring strategy mapping table is used to characterize the correspondence between comprehensive risk, monitoring task role and control parameter set. The control parameter set may include at least one of camera parameters, radar sensor parameters and main control chip parameters. The camera parameters may include the camera's resolution, frame rate, sensitivity and wake-up duty cycle. The radar sensor parameters may include the radar sensor's detection frequency and transmission power. The main control chip parameters may include the main control chip's operating voltage and operating frequency. Based on the control parameter set of each vehicle in the temporary vehicle group network, the operating parameters of the monitoring hardware of each vehicle in the temporary vehicle group network are adjusted.

[0054] In some examples, the monitoring strategy mapping table is a standardized table used to clearly represent the correspondence between comprehensive risk, monitoring task roles, and control parameter sets. Its function is to provide a basis for quick query of control parameter sets and ensure that the adjustment of monitoring hardware operating parameters is uniform and standardized. It can be preset and generated by R&D personnel according to common monitoring scenarios, different role responsibilities, and energy consumption control requirements, and stored in the on-board storage unit of each vehicle. After the temporary vehicle group network is built, the mapping table of the whole network can be synchronized through short-range cellular vehicle-to-everything communication to ensure that the query standards of each vehicle are consistent.

[0055] Based on the monitoring role and overall risk of each vehicle in the temporary vehicle group network, the process of querying the monitoring policy mapping table to obtain the control parameter set for each vehicle in the temporary vehicle group network is as follows: After the temporary vehicle group network determines the overall risk of the current scenario and the monitoring role of each vehicle, each vehicle synchronously retrieves the monitoring policy mapping table in its own on-board storage unit, using its own monitoring role and the current overall risk as dual query conditions, and matches the corresponding control parameter set in the mapping table. The query process is completed autonomously by each vehicle, without the need for centralized control of the vehicle group, ensuring the efficiency of parameter query. For example, if a vehicle is assigned the primary monitoring role and the current scenario's overall risk is high, this vehicle retrieves the monitoring policy mapping table, using the primary monitoring role and high overall risk as query conditions, and quickly matches the corresponding control parameter set, providing a basis for subsequent hardware parameter adjustments.

[0056] The control parameter set is a set of parameters obtained from the monitoring strategy mapping table based on the monitoring task role and comprehensive risk. It is used to regulate the operating status of vehicle monitoring hardware. Its function is to clarify the specific operating standards of each monitoring hardware, ensuring that the hardware operation meets the monitoring requirements while reasonably controlling energy consumption. For example, for an auxiliary perception vehicle in a low comprehensive risk scenario, the control parameter set obtained includes parameters such as camera resolution, frame rate, radar detection frequency, and main control chip operating voltage. This set is the basis for regulating the vehicle monitoring hardware. Camera parameters are specific parameters used in the control parameter set to regulate the operating status of vehicle-mounted cameras. They directly determine the camera's acquisition effect, operating load, and energy consumption level, and are a component of the control parameter set. For example, in high-risk scenarios, the camera parameters retrieved by the primary monitoring vehicle include high-definition resolution and high-frequency frame rate, ensuring that the camera captures clear monitoring images and meets core monitoring requirements. The camera's resolution, frame rate, sensitivity, and wake-up duty cycle are all specific components of the camera parameters, collectively determining the camera's monitoring performance and energy consumption. Resolution refers to the clarity of the image captured by the camera; for example, high-definition resolution corresponds to 1920×1080 pixels, and medium-definition resolution corresponds to... The resolution should be 1280×720 pixels, which can be flexibly matched according to monitoring needs; the frame rate is the number of frames of images captured by the camera per second. For example, the high-frequency frame rate is 30 frames / second, the medium-frequency frame rate is 15 frames / second, and the low-frequency frame rate is 1 frame / second. The higher the frame rate, the smoother the monitoring picture; the sensitivity is the sensitivity of the camera sensor to light. For example, high sensitivity is used in night or low-light scenes to improve the brightness of the picture, while low sensitivity is used in well-lit scenes to reduce the noise of the picture; the wake-up duty cycle is the percentage of time that the camera is in a wake-up working state per unit time. For example, a high wake-up duty cycle is 80%, which means that the camera is in a working state for 80% of the time per unit time, while a low wake-up duty cycle is 20%, which can effectively save energy consumption.Radar sensor parameters are specific parameters used in the control parameter set to regulate the operating state of vehicle-mounted radar sensors. They directly determine the detection range, sensitivity, and energy consumption of the radar sensor and are an important component of the control parameter set. For example, the radar sensor parameters for the primary monitoring vehicle are high-frequency detection frequency and high transmission power, enabling wide-range and high-precision detection. The radar sensor parameters for the auxiliary sensing vehicle are medium-frequency detection frequency and medium transmission power, which can save energy while meeting auxiliary detection requirements. The detection frequency and transmission power of the radar sensor are both specific components of the radar sensor parameters and together determine the detection performance of the radar sensor. The detection frequency is the frequency at which the radar sensor transmits the detection signal. For example, the high-frequency detection frequency is 24GHz, and the medium-frequency detection frequency is 12GHz. High-frequency detection can improve detection accuracy, while medium-frequency detection can balance accuracy and energy consumption. The transmission power is the power of the radar sensor transmitting the detection signal. For example, high transmission power is 10W, and medium transmission power is 5W. High transmission power can expand the detection range, while medium transmission power can reduce energy consumption. The main control chip parameters are specific parameters used in the control parameter set to regulate the operating status of the vehicle's main control chip. They directly determine the chip's computing speed, processing power, and energy consumption, and are a component of the control parameter set. For example, the main control chip parameters for vehicles in the primary monitoring role are high operating voltage and high operating frequency, which can improve the chip's computing speed and meet the real-time processing requirements of high-definition monitoring data. The main control chip parameters for vehicles in the dormant role are low operating voltage and low operating frequency, which can minimize energy consumption. The operating voltage and operating frequency of the main control chip are both specific components of the main control chip parameters, jointly determining the chip's operating performance and energy consumption level. The operating voltage is the voltage required for the main control chip to operate normally. For example, the high operating voltage is 3.3V, and the low operating voltage is 1.8V. The higher the voltage, the stronger the chip's computing power, but the higher the energy consumption. The operating frequency is the main control chip's operating frequency. For example, the high operating frequency is 1GHz, and the low operating frequency is 500MHz. The higher the frequency, the faster the chip processes data, but the energy consumption also increases accordingly.

[0057] For example, after the temporary vehicle group network determines that the overall risk of the current parking lot scenario is medium, each vehicle synchronously clarifies its own monitoring task role, with two vehicles as primary monitoring roles, three vehicles as auxiliary sensing roles, and two vehicles as dormant roles. Each vehicle retrieves the monitoring strategy mapping table in its own on-board storage unit, using its own monitoring task role and medium overall risk as query conditions to obtain its corresponding set of control parameters. The control parameter set obtained by the primary monitoring vehicle includes a camera resolution of 1920×1080 pixels, a frame rate of 30 frames per second, high sensitivity, 80% wake-up duty cycle, a radar sensor detection frequency of 24GHz, a transmit power of 10W, and a main control chip operating voltage of 3.3V and an operating frequency of 1GHz. The control parameter set obtained by the auxiliary perception vehicle includes a camera with a resolution of 1280×720 pixels, a frame rate of 15 frames per second, medium sensitivity, and a 50% wake-up duty cycle; a radar sensor with a detection frequency of 12GHz and a transmit power of 5W; and a main control chip with an operating voltage of 2.5V and an operating frequency of 800MHz. The control parameter set obtained by the dormant vehicle includes a camera with a low resolution, a frame rate of 1 frame per second, low sensitivity, and a 20% wake-up duty cycle; a radar sensor with a low detection frequency and a transmit power of 2W; and a main control chip with an operating voltage of 1.8V and an operating frequency of 500MHz. Each vehicle's onboard control system sends the parameters from the control parameter set to the corresponding monitoring hardware control unit. Each hardware unit adjusts its operating status according to the parameters. After adjustment, each vehicle synchronizes its hardware operating status to the temporary vehicle group network to ensure that the vehicle group monitoring hardware operation is adapted to the current comprehensive risk and role requirements, achieving a balance between monitoring effectiveness and energy consumption.

[0058] Through the implementation of the above embodiments, a mapping relationship between comprehensive risk, monitoring task roles and control parameter sets is constructed, and hardware parameters such as cameras, radar sensors and main control chips are dynamically adjusted accordingly. This enables the working state of the monitoring hardware to be precisely matched with different risk levels and role divisions, ensuring monitoring performance in high-risk or critical task scenarios, reducing hardware operating intensity in low-risk or non-critical task scenarios, achieving a balance between monitoring capabilities and resource consumption, improving overall energy efficiency and avoiding unnecessary resource consumption.

[0059] In some embodiments, the aforementioned warning information is issued by a first vehicle that detects a risk event, and the warning information may include the location of the risk event. Broadcasting the warning information through a temporary vehicle network to trigger at least some vehicles in the temporary vehicle network to perform collaborative monitoring operations may include: a second vehicle assigned to a dormant role switching from a dormant state to an awake state in response to receiving the warning information; the first vehicle, the awakened second vehicle, and a third vehicle assigned to an auxiliary perception role adjusting the acquisition angle of the onboard sensors based on the event location to acquire sensor data of the risk event from different angles; and fusing the sensor data acquired by the first vehicle, the second vehicle, and the third vehicle to obtain a multi-view joint evidence package for the risk event.

[0060] In some examples, the first vehicle in a temporary vehicle group network is the vehicle that first detects a risk event through its own monitoring hardware, actively generates a warning message, and broadcasts it to the group. The first vehicle in the network to trigger risk detection and warning broadcast is the initiating entity for collaborative monitoring operations. Each vehicle in the temporary vehicle group network can detect scene data in real time through its own monitoring hardware. When the monitoring hardware of a vehicle first detects a risk event that meets preset standards, that vehicle becomes the first vehicle without the need for vehicle group negotiation or external instructions. For example, an auxiliary perception vehicle in the temporary vehicle group network whose monitoring hardware first detects abnormal scratching behavior by a person is the first vehicle and is responsible for initiating the warning broadcast process.

[0061] The location of a risk event is the specific spatial position and direction of the risk event detected by the first vehicle. It is used to clarify the area where the risk event is located and to provide accurate positioning information for other vehicles to adjust the collection angle of their onboard sensors and carry out collaborative monitoring. For example, if the first vehicle determines that it is located in area A of the parking lot through the global positioning system, and the risk event is captured by the camera 3 meters in front of it, the location of the event is calculated to be 3 meters in front of area A of the parking lot. This information is the location of the risk event.

[0062] The second vehicle is a network vehicle in the temporary vehicle group network that has been pre-assigned to a dormant role. It is normally in a dormant state, retaining only short-range communication monitoring functions. It can quickly respond and switch states after receiving early warning information. When the temporary vehicle group network completes the monitoring task role assignment, it assigns vehicles with functional quantification parameters lower than the second parameter threshold to the dormant role. These vehicles are the second vehicles, and their identity is determined after the role assignment is completed until the vehicle group network is disbanded or the role is adjusted. For example, two vehicles in the temporary vehicle group network with functional quantification parameters of 55 and 52 are assigned to the dormant role. These two vehicles are the second vehicles and are normally in a low-power dormant state. When the second vehicle is in sleep mode, it can retain only the low-power monitoring function of the cellular vehicle-to-everything (V2X) short-range communication module, while all other monitoring hardware is turned off to save energy. When the second vehicle receives a warning message broadcast by the first vehicle through the short-range communication module, the vehicle control system immediately triggers a wake-up command, activates all its own monitoring hardware, and switches from sleep mode to normal operation, preparing to participate in collaborative monitoring operations. The entire switching process does not require manual intervention and is completed autonomously by the vehicle control system. For example, if the second vehicle in sleep mode receives an abnormal scratch warning message broadcast by the first vehicle, it triggers a wake-up command within 1 second, activates monitoring hardware such as cameras and radar sensors, completes the switch from sleep mode to wake-up mode, and simultaneously prepares for collaborative monitoring.

[0063] The third vehicle is a network vehicle in a temporary vehicle group network that is pre-assigned as an auxiliary sensing role. It is normally in a low-load monitoring state and has the ability to quickly adjust the monitoring angle and supplement the collection of sensor data. It is the main body of the collaborative monitoring operation. The third vehicle can be assigned as an auxiliary sensing role by the temporary vehicle group network when completing the monitoring task role assignment. Vehicles with functional quantification parameters greater than or equal to the second parameter threshold and less than the first parameter threshold are assigned as auxiliary sensing roles. These vehicles are the third vehicles. Their identity remains fixed after the role assignment is completed until the vehicle group network is disbanded or the role is adjusted. For example, three vehicles in the temporary vehicle group network with functional quantification parameters of 75, 68 and 62 are assigned as auxiliary sensing roles. These three vehicles are the third vehicles and are normally in a low-load auxiliary monitoring state.

[0064] After receiving the warning information and the location of the risk event, the first, second, and third vehicles, through their respective vehicle control systems, autonomously adjust the acquisition angle of their onboard sensors based on the relative relationship between the event location and their own parking positions. The first vehicle, as the triggering entity for risk detection, keeps its sensors pointed at the location of the risk event, continuously collecting sensor data in the core area. The awakened second vehicle adjusts its sensor angle to cover the side or rear area of ​​the risk event. The third vehicle adjusts its sensor angle to fill the monitoring blind spots of the first and second vehicles. The three vehicles collect data from different spatial angles to ensure comprehensive coverage of the risk event. The collected data is stored in real time and synchronized to the vehicle group. For example, if the first vehicle is directly in front of the risk event, it keeps its camera pointed at the location of the event; the awakened second vehicle is behind the risk event, adjusting its camera angle to the front to collect data from the rear view; and the third vehicle is to the side of the risk event, adjusting its camera angle to the center of the event to collect data from the side view. The three vehicles complete the sensor data collection from three different angles. A multi-view joint evidence package is a complete data set formed by summarizing and integrating risk event sensor data collected from different angles by the first vehicle, the activated second vehicle, and the third vehicle. Its purpose is to provide comprehensive and multi-dimensional evidence support for tracing and collecting evidence of risk events, ensuring the integrity and validity of the evidence. For example, a multi-view joint evidence package includes high-definition frontal video of the risk event collected by the first vehicle, rear video of the event collected by the second vehicle, side video of the event collected by the third vehicle, and radar detection data, forming a complete multi-view evidence set.

[0065] The process of fusing sensor data collected by the first, second, and third vehicles to obtain a multi-perspective joint evidence package for a risk event involves the first vehicle, the activated second vehicle, and the third vehicle collecting sensor data for the risk event. These vehicles then synchronize their collected video data, radar detection data, and other sensor data to a temporary vehicle ensemble network via short-range cellular vehicle-to-everything (V2X) communication. A pre-defined data fusion algorithm is used to deduplicate, sort, and stitch the data collected by the three types of vehicles, eliminating invalid data and retaining valid and comprehensive multi-perspective data. This data is then integrated to form a multi-perspective joint evidence package for the risk event. This evidence package contains complete process data of the risk event, clearly presenting its occurrence and development. The entire fusion process is completed autonomously by the vehicle ensemble without external intervention. For example, the vehicle ensemble stitches together the frontal video of the first vehicle, the rear video of the second vehicle, and the side video of the third vehicle along a timeline, fusing radar-detected personnel movement trajectory data, and eliminating blurry or repetitive segments to form a complete multi-perspective joint evidence package for the evidence collection and tracing of risk events.

[0066] For example, in a temporary vehicle convoy network, a vehicle acting as an auxiliary sensing role is the first to detect a risk event where a person with a handheld tool approaches the vehicle. This vehicle becomes the first vehicle. The first vehicle collects its own location via GPS and, combined with the relative direction of the event captured by the camera, calculates the location of the risk event as 5 meters to the left of area B in the parking lot. It then generates a warning message containing this location and broadcasts it to the entire convoy via cellular vehicle-to-everything (V2X) communication. Upon receiving the warning message, two second vehicles in a dormant state immediately trigger a wake-up command, quickly switching from dormant to awake and activating all their monitoring hardware. The three third vehicles, assigned as auxiliary sensing roles... Together with the first vehicle and the awakened second vehicle, the vehicle adjusts the acquisition angles of its onboard cameras and radar sensors based on the location of the event in the warning information and its own parking position. The first vehicle focuses on the center of the event to collect high-definition video from the front, while the two second vehicles adjust their angles to collect data from the rear and side rear of the event, respectively. The three third vehicles fill in the monitoring blind spots and collect video and radar data from different angles. All vehicles synchronize the collected sensor data to the vehicle group via short-range communication. The vehicle group uses a preset algorithm to deduplicate, stitch, and integrate the data, ultimately forming a multi-view joint evidence package containing multi-view video and radar trajectory data, which fully records the entire process of the risk event and provides reliable support for subsequent evidence collection.

[0067] Through the implementation of the above embodiments, when a risk event occurs, the detection vehicle broadcasts a warning message containing the location of the event, triggers the waking up of dormant vehicles and the coordinated adjustment of sensor acquisition angles by multiple vehicles, enabling multiple vehicles to jointly perceive the same event from different spatial locations, and fuse multi-source data to generate a multi-view joint evidence package. This can effectively expand the monitoring coverage, improve the completeness and accuracy of risk event capture, and enhance the system's response capability and evidence collection capability for risk events in complex scenarios.

[0068] In some embodiments, the step of determining the comprehensive risk for any fourth vehicle in a temporary vehicle network may include: determining a first risk coefficient based on geofence information corresponding to the geographical location of the fourth vehicle; identifying abnormal acoustic events in the environmental sound signals based on the environmental sound signals of the fourth vehicle, and determining a second risk coefficient based on the type and intensity of the abnormal acoustic events; determining the density and movement speed of people around the fourth vehicle based on the visual perception data of the fourth vehicle, and determining a third risk coefficient based on the density and movement speed of people; and weightedly fusing the first risk coefficient, the second risk coefficient, and the third risk coefficient to obtain the comprehensive risk of the scene in which the fourth vehicle is located.

[0069] In some examples, the fourth vehicle is any one of the networked vehicles in the temporary vehicle group network that needs to determine the comprehensive risk of its current scenario. As an independent entity for comprehensive risk calculation, it covers all vehicles participating in collaborative monitoring within the temporary vehicle group network, without any specific role restrictions. For example, the temporary vehicle group network contains five networked vehicles, of which two are primary monitoring roles, two are auxiliary perception roles, and one is a dormant role. All five vehicles belong to the fourth vehicle, and each independently calculates the comprehensive risk of its own scenario.

[0070] The geofence information corresponding to the geographical location of the fourth vehicle is the preset geographical area range of the current parking location of the fourth vehicle and the safety risk attribute information of that area. It is used as the basis for determining the first risk coefficient and its function is to provide a basic risk reference at the geographical location level. The fourth vehicle can collect its real-time geographical location coordinates through its own onboard GPS or Beidou navigation satellite system, and combine them with the preset geofence database in the vehicle's storage unit to obtain the geofence information corresponding to the geographical location. At the same time, it can be synchronized to the temporary vehicle group network through cellular vehicle-to-everything communication. For example, the fourth vehicle collects its location in the underground parking lot of a shopping mall through the GPS. After matching it with the preset geofence database on the vehicle, it obtains geofence information such as that the parking lot is a medium-risk area with a clear monitoring coverage area. This information is the geofence information corresponding to the geographical location of the fourth vehicle. The first risk coefficient is a numerical value representing the degree of safety risk at the geographical location level, calculated based on the geofence information corresponding to the geographical location of the fourth vehicle through preset quantification rules. Its function is to reflect the basic risk level of the geographical area where the fourth vehicle is located and is an important component of comprehensive risk calculation. For example, if the geofence information shows that the fourth vehicle is in a high-risk area, the first risk coefficient calculated by the preset quantification algorithm is 0.8; if it is in a medium-risk area, the first risk coefficient is 0.5; and if it is in a low-risk area, the first risk coefficient is 0.2.

[0071] Ambient sound signals are various sound waveform data existing in the environment surrounding the fourth vehicle, covering normal and abnormal sounds in the environment. They are the data source for identifying abnormal acoustic events and determining the second risk factor. The fourth vehicle can collect ambient sounds in real time through its own low-power microphone, convert the collected sound signals into processable electrical signals, and after preliminary processing by the vehicle control system, store them locally for subsequent identification of abnormal acoustic events. For example, the low-power microphone of the fourth vehicle continuously collects sounds in the surrounding environment, including normal sounds such as pedestrian footsteps, vehicle horns, and wind, as well as abnormal sounds such as the sound of breaking glass and metal collisions. All of these collected sound data belong to ambient sound signals. An abnormal acoustic event is a specific sound event identified from the environmental sound signals collected by the fourth vehicle that does not conform to the normal sound characteristics of the surrounding environment and may be associated with safety risks. It has the characteristics of abnormality and potential safety hazards and is the key basis for determining the second risk coefficient. The fourth vehicle can use the lightweight audio recognition model built into the vehicle control system to analyze the collected environmental sound signals in real time and compare the analysis results with the preset abnormal sound feature library. When the matching degree reaches the preset threshold, it is determined that an abnormal acoustic event has been identified. For example, the audio recognition model of the fourth vehicle can identify the sound of breaking glass, the sound of metal collision, and the sound of abnormal alarm from the environmental sound signals. These sound events are all abnormal acoustic events. Among them, the sound of breaking glass is one of the abnormal acoustic events that is most likely to trigger risk warnings. The second risk coefficient is a numerical value representing the degree of safety risk at the sound level, calculated based on the type and intensity of the identified abnormal acoustic events using preset quantification rules. Its function is to reflect the potential safety risk level in environmental sounds and is an important component of comprehensive risk calculation. After the fourth vehicle identifies an abnormal acoustic event, the specific type of the event is first determined, and then the sound intensity of the event is analyzed using an audio recognition model. Combining preset type weights and intensity quantification standards, the second risk coefficient is calculated. Different types and intensities of abnormal acoustic events correspond to different second risk coefficients. For example, identifying a high-intensity glass-breaking sound corresponds to a second risk coefficient of 0.9; identifying a medium-intensity metallic impact sound corresponds to a second risk coefficient of 0.6; and when no abnormal acoustic event is identified, the second risk coefficient is 0.1.

[0072] Visual perception data refers to the images or videos of the surrounding environment collected by the fourth vehicle through its onboard cameras. This data is used to capture dynamic information about people and objects in the surrounding environment and serves as a source for determining population density, movement speed, and the third risk factor. The fourth vehicle can activate its onboard cameras based on its monitoring role and the overall risk of the scene, collecting images or videos of the surrounding environment at a preset frame rate and resolution. The collected visual perception data is then processed by the onboard control system for subsequent analysis of population density and movement speed. For example, the fourth vehicle's onboard camera might collect video of the surrounding environment at a frame rate of 15 frames per second, capturing images of pedestrians walking and vehicles moving in a parking lot; this video data constitutes visual perception data. The density and speed of people around the fourth vehicle are determined by analyzing the visual perception data collected by the fourth vehicle. This analysis shows the distribution of people and their speed of movement within a certain range around the fourth vehicle, reflecting the dynamic safety risks of the surrounding environment and serving as the basis for determining the third risk coefficient. For example, if the visual perception data analysis of the fourth vehicle shows that there are 5 pedestrians within a 10-meter radius, the population density is 0.5 people / square meter, and the average pedestrian speed is 1.2 meters / second, this data represents the population density and speed around the fourth vehicle. The third risk coefficient is a numerical value representing the level of safety risk at the dynamic level of surrounding personnel, calculated based on the density and movement speed of people around the fourth vehicle using preset quantification rules. Its function is to reflect the potential safety risk level brought about by the activities of surrounding personnel and is an important component of comprehensive risk calculation. After obtaining personnel density and movement speed data for the fourth vehicle, it combines preset density and speed thresholds and uses a quantification algorithm to convert personnel density and movement speed into specific values, thus obtaining the third risk coefficient. The higher the personnel density and the faster the movement speed, the higher the third risk coefficient. For example, when the personnel density is 0.8 people / square meter and the average movement speed is 1.5 meters / second, the third risk coefficient is 0.7; when the personnel density is 0.3 people / square meter and the average movement speed is 0.8 meters / second, the third risk coefficient is 0.3; and when there is no personnel activity, the third risk coefficient is 0.1.

[0073] The process of weightedly fusing the first, second, and third risk coefficients to obtain the comprehensive risk of the scenario in which the fourth vehicle is located can be as follows: First, the fourth vehicle determines the specific values ​​of the first, second, and third risk coefficients. Then, according to a preset weighting rule, each of the three risk coefficients is assigned a corresponding weight. The coefficients are then fused using a weighted summation method, and the final single value obtained is the comprehensive risk of the scenario in which the fourth vehicle is located. This weighting rule is preset in the fourth vehicle's onboard control system and can be fine-tuned according to the actual needs of the monitoring scenario. The overall risk calculation is completed autonomously by the fourth vehicle. The calculated overall risk is synchronized to the temporary vehicle group network for subsequent adjustment of the monitoring hardware operating parameters. For example, if the first risk coefficient is preset to have a weight of 30%, the second risk coefficient to have a weight of 40%, and the third risk coefficient to have a weight of 30%, and if the first risk coefficient of a certain fourth vehicle is 0.5, the second risk coefficient is 0.6, and the third risk coefficient is 0.3, the overall risk is calculated by weighted summation as 0.5×30%+0.6×40%+0.3×30%=0.48. This value is the overall risk of the scenario in which the fourth vehicle is located.

[0074] By implementing the above embodiments, geofence information, environmental sound signals, and visual perception data are integrated to comprehensively analyze the vehicle's environment from multiple dimensions. A comprehensive risk is generated through a weighted approach, transforming risk assessment from a single-factor judgment to a multi-source information fusion judgment. This can more comprehensively and accurately reflect the risks in the actual scenario, reduce misjudgments or omissions, provide a reliable basis for subsequent monitoring task allocation and hardware parameter adjustments, and thus improve the effectiveness of overall monitoring decisions.

[0075] In some embodiments, the aforementioned vehicle sentry monitoring method may further include: acquiring feedback tags of risk events and risk feature vectors that trigger risk events, wherein the feedback tags may include false alarm tags and real threat tags; associating and storing the feedback tags and risk feature vectors in a local sample library of a first vehicle; when the number of samples in the local sample library reaches a preset sample threshold and the first vehicle is in a charging or dormant state, incrementally training the local risk perception model of the first vehicle based on the samples in the local sample library to obtain a first model parameter update amount; updating the local risk perception model according to the first model parameter update amount, wherein the local risk perception model is used to determine the comprehensive risk of the first vehicle.

[0076] In some examples, feedback tags are used to determine the nature of risk events detected by the first vehicle. Their function is to indicate the authenticity of the risk event, providing labeled data for incremental training of the local risk perception model and ensuring the relevance and accuracy of the model training. After the first vehicle detects a risk event and completes collaborative monitoring and generates a multi-view joint evidence package, the vehicle user can manually label it through the in-vehicle interface, or the in-vehicle control system can automatically label it according to preset risk event judgment rules. After labeling, the data is synchronized to the first vehicle's in-vehicle storage unit. For example, if the abnormal sound detected by the first vehicle is verified as wind blowing debris and hitting the vehicle body, it is labeled with the corresponding feedback tag; if it is verified as someone scratching the vehicle, it is labeled with another type of feedback tag. False alarm tags and real threat tags are two specific types of feedback tags. They are independent and incompatible, but together they cover all scenarios for determining the nature of risk events. False alarm tags are used to indicate that the risk event detected by the first vehicle is not a real security threat and is a false trigger, while real threat tags are used to indicate that the risk event detected by the first vehicle is a real security threat that requires further action.

[0077] The risk feature vector that triggers a risk event is a multi-dimensional data set used to characterize various features that trigger a risk event. Its function is to record key feature information when a risk event occurs. After being associated with feedback labels, it serves as input samples for incremental training of the local risk perception model. When the first vehicle detects a risk event, it can simultaneously collect multi-dimensional data such as geofence information, environmental sound signal features, and visual perception data features corresponding to the risk event. The vehicle control system standardizes these data and integrates them to form a one-dimensional or multi-dimensional risk feature vector, which is then synchronously associated with the feedback labels. For example, the risk feature vector of a certain risk event includes multi-dimensional data such as the regional risk features corresponding to the geofence, the frequency features of the glass breaking sound, and the movement features of people in the monitoring screen, which fully characterize the triggering features of the risk event.

[0078] The local sample library is a database pre-installed in the vehicle's onboard storage unit, used to store samples associated with risk feature vectors and feedback labels. Its function is to provide data support for incremental training of the local risk perception model, ensuring the local storage and security of training data. The local sample library can be stored on the vehicle's onboard solid-state drive (SSD) or embedded storage unit, and has sample storage, classification, query, and retrieval functions. Access is only granted to the vehicle's onboard control system. For example, the local sample library of the first vehicle stores multiple associated samples, each containing a risk feature vector and a corresponding feedback label, and is categorized and stored according to risk event type for easy retrieval during subsequent incremental training. The preset sample threshold is a critical value set within the vehicle's onboard control system to trigger incremental training of the local risk perception model. Its function is to control the timing of incremental training, ensuring that the number of training samples meets the model optimization requirements while avoiding frequent training that consumes vehicle resources. For example, based on the model training accuracy requirements, the researchers preset the sample threshold to 20, meaning that when the number of associated samples in the local sample library reaches 20, the sample quantity condition for incremental training is met.

[0079] The local risk perception model is an algorithmic model deployed within the vehicle's onboard control system to calculate the comprehensive risk of the scenario in which the vehicle is located. Its function is to output the comprehensive risk of the vehicle based on multimodal perception data, providing a basis for adjusting the operating parameters of the monitoring hardware and detecting risk events. The local risk perception model can be pre-trained and deployed within the vehicle's onboard control system. The initial model parameters are adapted to common monitoring scenarios, and subsequent parameter optimization can be achieved through incremental training. It is always deployed locally within the vehicle and does not rely on cloud servers. For example, the local risk perception model of the vehicle is a lightweight deep learning model that can receive inputs such as geofence information, environmental sound signals, and visual perception data. It calculates and outputs a comprehensive risk value through a built-in algorithm, supporting the risk detection and judgment of the vehicle. The first model parameter update amount is the adjusted value of the model parameters obtained after the first vehicle incrementally trains the local risk perception model based on associated samples in the local sample library. Its purpose is to optimize the parameters of the local risk perception model and improve the model's accuracy in identifying risk events and the accuracy of comprehensive risk calculation. After meeting the incremental training conditions, the first vehicle starts the preset incremental training algorithm, uses associated samples in the local sample library as training data, and fine-tunes the initial parameters of the local risk perception model. After training is completed, the parameter adjustment amount is calculated, which is the first model parameter update amount. For example, through incremental training, the weight parameter of environmental sound feature recognition in the local risk perception model is adjusted from 0.4 to 0.45. This adjustment amount of 0.05 is the first model parameter update amount, which is used to optimize the model's recognition accuracy for sound-related risks.

[0080] The first vehicle continuously counts the number of related samples in its local sample library. When the number of samples reaches a preset threshold, it simultaneously detects its own operating status to determine whether it is in charging or dormant mode. If both conditions are met—the number of samples reaching the threshold and the vehicle being in charging or dormant mode—the vehicle's onboard control system automatically initiates an incremental training process. It uses related samples from the local sample library as training data and employs a preset incremental training algorithm to specifically train the local risk perception model. During training, the model parameters are continuously fine-tuned, and the adjusted values ​​of the model parameters are output after training, thus obtaining the first model parameter update. Choosing to train in charging or dormant mode avoids consuming excessive power and normal vehicle monitoring resources, ensuring the normal operation of the vehicle monitoring function. For example, if the preset sample threshold for the first vehicle is 20, when the number of related samples in the local sample library reaches 20 and the first vehicle is charging, the onboard control system initiates incremental training, uses the 20 related samples to train the local risk perception model, and finally obtains the first model parameter update, completing the training process.

[0081] After the first vehicle completes incremental training and obtains the first model parameter update, the vehicle control system automatically calls upon this parameter update to replace or adjust the initial parameters of the local risk perception model. The updated local risk perception model retains the original algorithm framework, only optimizing the model parameters to ensure that the model's accuracy in identifying risk events and the accuracy of comprehensive risk calculation are improved. After the update is completed, the vehicle control system stores the updated model parameters locally, overwriting the original parameters, and records the update log for subsequent model maintenance and traceability. The entire update process is completed autonomously by the first vehicle without relying on cloud or external device intervention, and does not affect the vehicle's normal monitoring functions. For example, the first model parameter update obtained by the first vehicle includes multiple parameter adjustment values ​​such as sound recognition weight and visual feature recognition threshold. The vehicle control system updates the corresponding parameters of the local risk perception model one by one according to these adjustment values. After the update is completed, the model's recognition accuracy for glass breaking sounds and abnormal human movements is significantly improved.

[0082] By implementing the above embodiments, a local sample library is constructed by introducing risk event feedback labels and risk feature vectors, and the local risk perception model is incrementally trained when preset conditions are met. This enables vehicles to continuously optimize their risk identification capabilities based on their own historical monitoring data, thereby improving their adaptability and identification accuracy to different environments and event types. The model can continuously evolve without relying on external data, which helps to improve the intelligence level and long-term stability of the monitoring system.

[0083] In some embodiments, the aforementioned vehicle sentry monitoring method may further include: for any fifth vehicle in a temporary vehicle group network, training a local collaborative strategy model of the fifth vehicle based on security event data stored locally by the fifth vehicle to obtain a second model parameter update, wherein the security event data is risk event record data generated by the fifth vehicle during historical monitoring, and the local collaborative strategy model is used to implement the allocation of monitoring task roles for the fifth vehicle; encrypting the second model parameter update to obtain an encrypted gradient; transmitting the encrypted gradient to a coordination node in the temporary vehicle group network, so that the coordination node can aggregate the encrypted gradients from multiple vehicles in the temporary vehicle group network to generate model parameters of a global collaborative strategy model, wherein the global collaborative strategy model is used to implement the allocation of monitoring task roles for each vehicle in the temporary vehicle group network; and updating the local collaborative strategy model of the fifth vehicle based on the model parameters returned by the coordination node.

[0084] In some examples, the fifth vehicle is any vehicle in the temporary vehicle group network that participates in local collaborative policy model training, encrypted gradient transmission, and model updates. As the local training subject in the federated learning process, it covers all vehicles participating in collaborative monitoring within the temporary vehicle group network, without specific role restrictions. After the temporary vehicle group network is built, each vehicle in the network automatically becomes the fifth vehicle without additional screening or designation. Each fifth vehicle independently completes the training of the local collaborative policy model, the generation and transmission of encrypted gradients, and the updating of the local model. For example, the temporary vehicle group network contains five vehicles, two of which are primary monitoring roles, two are auxiliary perception roles, and one is a dormant role. All five vehicles belong to the fifth vehicle and independently carry out local collaborative policy model training and related operations.

[0085] Security event data consists of complete data related to various risk events recorded and stored in real time by the fifth vehicle during historical monitoring. Its purpose is to serve as input data for training the fifth vehicle's local collaborative strategy model, providing real historical scenarios to support model optimization. Essentially, it is the risk event record data generated by the fifth vehicle in its past monitoring work. Each time the fifth vehicle detects a risk event and completes a collaborative monitoring operation, it automatically summarizes the risk feature vector, feedback tag, event location, collaborative monitoring process data, and other information corresponding to that risk event, storing it in its own onboard storage unit to form historical security event data. This process requires no manual intervention; the entire process is automatically collected, organized, and stored by the onboard control system. For example, during the fifth vehicle's monitoring over the past month, it recorded three instances of abnormal personnel approaching and two instances of minor vehicle collisions. The risk characteristics, feedback tags, and collaborative handling processes for each event were summarized and stored, forming the fifth vehicle's security event data.

[0086] The local collaborative strategy model is an algorithmic model deployed within the fifth vehicle's onboard control system to assign monitoring task roles to the fifth vehicle itself. Its function is to output suitable monitoring task roles based on the overall needs of the temporary vehicle group network, the fifth vehicle's status information, and the comprehensive risks of the scenario, providing local algorithmic support for role allocation in collaborative vehicle group monitoring. For example, the fifth vehicle's local collaborative strategy model is a lightweight decision tree model that can receive inputs such as the fifth vehicle's battery status, location information, sensor capability information, and comprehensive risks, and output the appropriate primary monitoring, auxiliary sensing, or dormant role for the vehicle through a built-in algorithm. The second model parameter update is the adjusted model parameter value obtained after the fifth vehicle trains its local collaborative strategy model based on its locally stored security event data. Its purpose is to optimize the parameters of the local collaborative strategy model, improving the model's rationality and adaptability in allocating monitoring tasks to the fifth vehicle. The difference between the second and first model parameter updates lies in their corresponding model types: the former targets the local collaborative strategy model, while the latter targets the local risk perception model. The fifth vehicle can periodically access its locally stored security event data to initiate a preset model training algorithm. Using the security event data as training data, the initial parameters of the local collaborative strategy model are fine-tuned. The parameter adjustment value calculated after training is the second model parameter update. For example, through training based on security event data, the weight parameter of the battery status in the local collaborative strategy model is adjusted from 0.3 to 0.35, and the weight parameter of the sensor capability information is adjusted from 0.4 to 0.38. These two sets of parameter adjustments together constitute the second model parameter update.

[0087] The encrypted gradient is the encrypted data obtained by the fifth vehicle after encrypting the parameter updates of the second model. Its purpose is to protect the privacy and security of the parameter updates of the second model, prevent the parameter data from being leaked or tampered with during transmission, and ensure data privacy in the federated learning process. After obtaining the parameter updates of the second model, the fifth vehicle can encrypt the parameter updates using an encryption algorithm built into the vehicle control system. After encryption, an encrypted gradient is generated. The encryption algorithm adopts a preset lightweight symmetric encryption algorithm (SEA), which does not rely on external encryption devices and is completed autonomously by the fifth vehicle. For example, the fifth vehicle uses a lightweight symmetric encryption algorithm to encrypt the parameter updates of the second model, transforming the originally directly readable parameter adjustment values ​​into encrypted data that cannot be directly parsed. This encrypted data is the encrypted gradient.

[0088] The coordinating node is an elected node in the temporary vehicle ensemble network, responsible for receiving and aggregating encryption gradients transmitted by multiple fifth vehicles and generating parameters for the global collaborative strategy model. Its role is to summarize and process encryption gradients within the vehicle ensemble, and to drive the generation and updating of the global collaborative strategy model. It is the coordinating entity in the federated learning process. After the temporary vehicle ensemble network is constructed, a decentralized negotiation mechanism can be used to elect one or more vehicles with high battery power, high performance, and high stability as coordinating nodes from all the fifth vehicles. The selection criteria mainly include the vehicle's battery status, sensor capabilities, and network communication stability. After the election, the results are synchronized to the entire temporary vehicle ensemble network. For example, in the temporary vehicle ensemble network, a main monitoring vehicle with 90% battery power, equipped with a high-performance main control chip, and stable communication is elected as the coordinating node, responsible for receiving and aggregating encryption gradients transmitted by other fifth vehicles.

[0089] The global collaborative strategy model is a global algorithm model generated by the aggregation and processing of encrypted gradients from multiple fifth vehicles in a temporary vehicle group network. It is used to assign monitoring roles to each vehicle in the temporary vehicle group network. Its function is to provide a unified and optimized strategy basis for role allocation across the entire vehicle group, improving the overall efficiency of collaborative monitoring. Unlike the locality of local collaborative strategy models, it possesses global adaptability. The coordination node receives encrypted gradients transmitted from multiple fifth vehicles in the temporary vehicle group network, decrypts, summarizes, and calculates the encrypted gradients using a preset aggregation algorithm, generating model parameters for the global collaborative strategy model. This allows for the construction of a complete global collaborative strategy model, whose parameters can be synchronized to all fifth vehicles for local model updates. For example, if the coordination node aggregates the encrypted gradients of five fifth vehicles and processes them using the aggregation algorithm, it generates a global collaborative strategy model adapted to the temporary vehicle group network. This model can achieve a more reasonable role allocation based on the status of each vehicle within the vehicle group. The model parameters of the global collaborative strategy model are the set of parameters that constitute the global collaborative strategy model. Their function is to define the decision logic of the global collaborative strategy model, determining the rules and accuracy of the model's allocation of monitoring task roles to each vehicle in the vehicle group. As a component of the global collaborative strategy model, these model parameters can be synchronously returned to each of the fifth vehicles in the temporary vehicle group network via short-range communication. After receiving the model parameters returned by the coordination node, the onboard control system of the fifth vehicle automatically calls upon these model parameters to replace or adjust the original parameters of its own local collaborative strategy model. The updated local collaborative strategy model integrates the training experience of multiple vehicles in the vehicle group, possessing a better role allocation decision. The system retains the original model's algorithm framework while optimizing only the model parameters. After the update, the vehicle control system stores the updated local collaborative strategy model parameters on the vehicle's solid-state drive, overwriting the original parameters. It also records update logs for subsequent model maintenance and traceability. The entire update process is completed autonomously by the fifth vehicle and does not affect the vehicle's normal monitoring functions. For example, after receiving the global collaborative strategy model parameters returned by the coordination node, the fifth vehicle adjusts the weight parameter of the position information in its local collaborative strategy model from 0.3 to the 0.32 specified by the global parameters, completing the update of the local collaborative strategy model. The updated model can more accurately adapt to the needs of vehicle group collaborative role allocation.

[0090] Through the implementation of the above embodiments, each vehicle trains a local collaborative strategy model based on local security event data, and the model update is encrypted and uploaded to the coordination node for aggregation to generate a global collaborative strategy model and update the local model in reverse. This enables the vehicle group to continuously optimize the collaborative strategy without sharing the original data, thereby balancing data security and model performance improvement. At the same time, by integrating the experience of multiple vehicles, the allocation of monitoring task roles becomes more reasonable and efficient, which can further improve the overall collaborative capability of the vehicle group and the level of system intelligence.

[0091] In a specific implementation scenario, Mr. Chen, a business professional who frequently works late into the night, drives an electric vehicle equipped with this solution. He parks it in an open-air public parking lot next to his office building at 11 PM on a weekday and leaves to rest. After the vehicle is turned off, the perception layer immediately activates, confirming the parking area as the company's nighttime parking lot via GPS. Combining this with historical monitoring records, the basic risk factor of the geofencing is determined to be medium. The vehicle broadcasts a beacon via short-range communication, quickly detecting three other vehicles equipped with the same system nearby. The four vehicles self-organize and form a temporary vehicle convoy security network within 10 seconds. The vehicle convoy then uses the battery status, sensor performance, and parking location of each vehicle to... The decentralized negotiation process completes the assignment of monitoring roles. Mr. Chen's vehicle, due to its sufficient battery power and the presence of a high-definition gimbal sensor, is assigned the primary monitoring role. The SUV on the left, also with sufficient battery power, is assigned the secondary sensing role. The other two vehicles with lower battery power are assigned the dormant role. When the parking lot enters the low-traffic period late at night, the system optimizes the risk identification logic based on an incremental learning model. It has the ability to adaptively filter conventional environmental noise. The vehicle group executes a low-power round-robin monitoring strategy. Only the low-power millimeter-wave radar of the primary monitoring vehicle runs periodically. When a pedestrian passes by, only the secondary sensing vehicle is awakened to complete cross-verification. The overall group power consumption is only 35% of the traditional single-vehicle sentry mode.

[0092] At 1:15 AM, a suspicious person approached the main monitoring vehicle. The main monitoring vehicle's audio acquisition module was the first to detect an abnormal metallic friction sound. The audio recognition model's confidence level exceeded 95%, and the overall risk value instantly rose to an extremely high level. The main monitoring vehicle immediately activated the highest-level monitoring plan. The pan-tilt camera locked onto the source of the risk and activated infrared illumination for full HD recording. At the same time, a red alert was broadcast via the temporary vehicle network. Upon receiving the alert, the dormant vehicles immediately switched from dormant to awakened mode and activated their corresponding cameras to block the suspicious person's movement path according to the alert instructions. The auxiliary sensing vehicles simultaneously adjusted their sensor acquisition angles to capture a panoramic view from the side and record the suspicious person's facial features. Within 3 seconds, the vehicle group completed multi-view collaborative evidence collection and automatically generated a joint evidence report containing a complete risk event timeline, a seamless multi-view video summary, a suspicious person's movement trajectory map, and annotations of key evidence segments.

[0093] The following day, when Mr. Chen was driving, the vehicle system pushed a nighttime safety report. After the user confirmed that the incident was an attempted vandalism, the system stored the risk event sample in the local sample library. During the subsequent charging period of the vehicle, incremental learning was initiated to fine-tune the parameters of the local risk perception model and improve the recognition accuracy of similar abnormal sounds. Mr. Chen's vehicle encrypted the model parameter update into an encrypted gradient, which was uploaded to the vehicle group coordination node while charging. The coordination node aggregated the encrypted gradients of similar events from more than 100 vehicles in the city and generated a new generation of collaborative defense strategy for open-air parking lots at night through federated learning, and silently pushed it to all vehicles in the network. A week later, when Mr. Chen parked his vehicle in an unfamiliar shopping mall parking lot, his vehicle had been loaded with the globally optimized collaborative strategy, which could quickly complete risk identification and efficiently build a collaborative monitoring network with surrounding vehicles, realizing the continuous iteration and capability upgrade of the system.

[0094] Furthermore, as an implementation of the aforementioned method embodiments, this application also provides a vehicle sentry monitoring device for implementing the aforementioned method embodiments. This device embodiment corresponds to the aforementioned method embodiments. For ease of reading, this vehicle sentry monitoring device embodiment will not repeat the details of the aforementioned method embodiments one by one, but it should be understood that the device in this application embodiment can correspondingly implement all the contents of the aforementioned method embodiments. For example... Figure 2 As shown, the vehicle sentry monitoring device 20 includes: a network establishment unit 201, a role allocation unit 202, a parameter determination unit 203, and a monitoring execution unit 204. The network establishment unit 201 is used to exchange information among multiple vehicles via short-range communication to construct a temporary vehicle group network. The role allocation unit 202 is used to assign monitoring task roles to each vehicle in the temporary vehicle group network based on the vehicle status information of each vehicle in the aforementioned temporary vehicle group network. These monitoring task roles may include a primary monitoring role, an auxiliary sensing role, and a dormant role. The parameter determination unit 203 is used to determine the operating parameters of the monitoring hardware of each vehicle in the temporary vehicle group network based on the monitoring task roles of each vehicle in the temporary vehicle group network and the comprehensive risk of the scenario. The monitoring execution unit 204 is used to broadcast early warning information through the temporary vehicle group network when a risk event is detected by the monitoring hardware, thereby triggering at least some vehicles in the temporary vehicle group network to perform collaborative monitoring operations.

[0095] In some embodiments, vehicle status information includes battery status, location information, and sensor capability information; the role allocation unit 202 is further configured to generate functional quantification parameters for each vehicle in the temporary vehicle group network based on the battery status, location information, and sensor capability information of each vehicle in the temporary vehicle group network; and to assign vehicles in the temporary vehicle group network whose functional quantification parameters are greater than or equal to a first parameter threshold to the main monitoring role, vehicles whose functional quantification parameters are greater than or equal to a second parameter threshold and less than the first parameter threshold to the auxiliary perception role, and vehicles whose functional quantification parameters are less than the second parameter threshold to the dormant role.

[0096] In some embodiments, the parameter determination unit 203 is further configured to query a monitoring strategy mapping table to obtain a set of control parameters for each vehicle in the temporary vehicle group network based on the monitoring task role and comprehensive risk of each vehicle in the temporary vehicle group network. The monitoring strategy mapping table is used to characterize the correspondence between comprehensive risk, monitoring task role and control parameter set. The control parameter set includes at least one of camera parameters, radar sensor parameters and main control chip parameters. The camera parameters include the camera resolution, frame rate, sensitivity and wake-up duty cycle. The radar sensor parameters include the radar sensor detection frequency and transmission power. The main control chip parameters include the main control chip operating voltage and operating frequency. Based on the control parameter set of each vehicle in the temporary vehicle group network, the operating parameters of the monitoring hardware of each vehicle in the temporary vehicle group network are adjusted.

[0097] In some embodiments, the warning information is issued by a first vehicle that detects a risk event, and the warning information includes the location of the risk event; the monitoring execution unit 204 is further configured to allow a second vehicle, which is assigned to a dormant role, to switch from a dormant state to an awake state in response to receiving the warning information; the first vehicle, the awakened second vehicle, and the third vehicle, which is assigned to an auxiliary perception role, adjust the acquisition angle of the on-board sensors based on the location of the event to acquire sensing data of the risk event from different angles; and the sensing data acquired by the first vehicle, the second vehicle, and the third vehicle are fused to obtain a multi-view joint evidence package for the risk event.

[0098] In some embodiments, the vehicle sentry monitoring device 20 further includes a risk determination unit, configured to determine a first risk coefficient based on geofence information corresponding to the geographical location of the fourth vehicle; identify abnormal acoustic events in the environmental sound signals based on the environmental sound signals of the fourth vehicle, and determine a second risk coefficient based on the type and intensity of the abnormal acoustic events; determine the density and movement speed of people around the fourth vehicle based on the visual perception data of the fourth vehicle, and determine a third risk coefficient based on the density and movement speed; and perform weighted fusion of the first risk coefficient, the second risk coefficient, and the third risk coefficient to obtain the comprehensive risk of the scene in which the fourth vehicle is located.

[0099] In some embodiments, the vehicle sentry monitoring device 20 further includes a first model update unit, configured to acquire feedback labels of risk events and risk feature vectors that trigger risk events, wherein the feedback labels include false alarm labels and real threat labels; associate and store the feedback labels and risk feature vectors in a local sample library of the first vehicle; when the number of samples in the local sample library reaches a preset sample threshold and the first vehicle is in a charging or dormant state, incrementally train the local risk perception model of the first vehicle based on the samples in the local sample library to obtain a first model parameter update amount; update the local risk perception model according to the first model parameter update amount, wherein the local risk perception model is used to determine the comprehensive risk of the first vehicle.

[0100] In some embodiments, the vehicle sentry monitoring device 20 further includes a second model update unit, configured to train a local collaborative strategy model for any fifth vehicle in the temporary vehicle group network based on security event data stored locally by the fifth vehicle, to obtain a second model parameter update amount, wherein the security event data is risk event record data generated by the fifth vehicle during historical monitoring, and the local collaborative strategy model is used to implement the allocation of monitoring task roles for the fifth vehicle; encrypt the second model parameter update amount to obtain an encryption gradient; transmit the encryption gradient to a coordination node in the temporary vehicle group network, so that the coordination node can aggregate the encryption gradients from multiple vehicles in the temporary vehicle group network to generate model parameters of a global collaborative strategy model, wherein the global collaborative strategy model is used to implement the allocation of monitoring task roles for each vehicle in the temporary vehicle group network; and update the local collaborative strategy model of the fifth vehicle based on the model parameters returned by the coordination node.

[0101] This application also provides a computer-readable storage medium storing computer-executable instructions or a computer program that, when executed by a processor, will cause the processor to perform any step of the vehicle sentry monitoring method provided in this application.

[0102] In some embodiments, the computer-readable storage medium may be a random access memory (RAM), a read-only memory (ROM), flash memory, a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM); or it may be a variety of devices that include one or any combination of the above-mentioned memories.

[0103] 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.

[0104] In some embodiments, computer-executable instructions may, but do not necessarily, correspond to files in a file system, and may be stored as part of a file that holds other programs or data; for example, stored in one or more scripts in a HyperText Markup Language (HTML) document, stored in a single file dedicated to the program in question, or stored in multiple co-located files (e.g., files that store one or more modules, subroutines, or code sections).

[0105] like Figure 3 As shown, this application also provides a vehicle 30, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor. When the processor 320 executes the computer program 311, it implements any step of the above-described vehicle sentry monitoring method.

[0106] This application also provides a computer program product comprising a computer program or computer-executable instructions stored in a computer-readable storage medium. A vehicle's processor reads the computer program or computer-executable instructions from the computer-readable storage medium and executes the computer program or computer-executable instructions, causing the vehicle to perform any step of the vehicle sentry monitoring method described above.

[0107] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A vehicle sentry monitoring method, characterized in that, include: Short-range communication enables information exchange between multiple vehicles to build temporary vehicle groups; Based on the vehicle status information of each vehicle in the temporary vehicle group network, a monitoring task role is assigned to each vehicle in the temporary vehicle group network. The monitoring task role includes a main monitoring role, an auxiliary sensing role, and a dormant role. Based on the monitoring task role of each vehicle in the temporary vehicle group network and the comprehensive risk of the scenario, the working parameters of the monitoring hardware of each vehicle in the temporary vehicle group network are determined. If a risk event is detected by the monitoring hardware, a warning message is broadcast through the temporary vehicle group network to trigger at least some of the vehicles in the temporary vehicle group network to perform collaborative monitoring operations.

2. The vehicle sentry monitoring method according to claim 1, characterized in that, The vehicle status information includes battery status, location information, and sensor capability information; the process of assigning monitoring task roles to each vehicle in the temporary vehicle group network based on the vehicle status information includes: Based on the battery status, location information, and sensor capability information of each vehicle in the temporary vehicle group network, functional quantification parameters of each vehicle in the temporary vehicle group network are generated. In the temporary vehicle group network, vehicles with functional quantization parameters greater than or equal to the first parameter threshold are assigned to the main monitoring role, vehicles with functional quantization parameters greater than or equal to the second parameter threshold and less than the first parameter threshold are assigned to the auxiliary perception role, and vehicles with functional quantization parameters less than the second parameter threshold are assigned to the dormant role.

3. The vehicle sentry monitoring method according to claim 1, characterized in that, The determination of the operating parameters of the monitoring hardware for each vehicle in the temporary vehicle group network, based on the monitoring task role of each vehicle and the comprehensive risk of the scenario, includes: Based on the monitoring task roles and comprehensive risks of each vehicle in the temporary vehicle group network, a set of control parameters for each vehicle in the temporary vehicle group network is obtained by querying the monitoring strategy mapping table. The monitoring strategy mapping table is used to characterize the correspondence between the comprehensive risk, the monitoring task roles, and the control parameter set. The control parameter set includes at least one of camera parameters, radar sensor parameters, and main control chip parameters. The camera parameters include the camera's resolution, frame rate, sensitivity, and wake-up duty cycle. The radar sensor parameters include the radar sensor's detection frequency and transmission power. The main control chip parameters include the main control chip's operating voltage and operating frequency. Based on the control parameter set of each vehicle in the temporary vehicle group network, the operating parameters of the monitoring hardware of each vehicle in the temporary vehicle group network are adjusted.

4. The vehicle sentry monitoring method according to claim 1, characterized in that, The warning information is issued by the first vehicle that detects the risk event, and the warning information includes the location of the risk event; The step of broadcasting early warning information through the temporary vehicle group network to trigger at least some vehicles in the temporary vehicle group network to perform collaborative monitoring operations includes: The second vehicle, assigned to the hibernation role, switches from hibernation to wake-up state in response to receiving the warning information; The first vehicle, the awakened second vehicle, and the third vehicle assigned as the auxiliary perception role adjust the acquisition angle of the on-board sensors based on the location of the event to collect sensing data of the risk event from different angles. The sensor data collected by the first vehicle, the second vehicle, and the third vehicle are fused to obtain a multi-perspective joint evidence package for the risk event.

5. The vehicle sentry monitoring method according to claim 1, characterized in that, For any fourth vehicle in the temporary vehicle network, the steps for determining the comprehensive risk include: Based on the geofence information corresponding to the geographical location of the fourth vehicle, a first risk coefficient is determined; Based on the ambient sound signal of the fourth vehicle, abnormal acoustic events in the ambient sound signal are identified, and a second risk coefficient is determined according to the type and intensity of the abnormal acoustic event. Based on the visual perception data of the fourth vehicle, the density and speed of people around the fourth vehicle are determined, and a third risk coefficient is determined based on the density and speed of people. The first risk coefficient, the second risk coefficient, and the third risk coefficient are weighted and fused to obtain the comprehensive risk of the scenario in which the fourth vehicle is located.

6. The vehicle sentry monitoring method according to claim 1, characterized in that, The vehicle sentry monitoring method also includes: Obtain the feedback label of the risk event and the risk feature vector that triggered the risk event, wherein the feedback label includes a false alarm label and a real threat label; The feedback label and the risk feature vector are associated and stored in the local sample library of the first vehicle; When the number of samples in the local sample library reaches a preset sample threshold, and the first vehicle is in a charging or dormant state, the local risk perception model of the first vehicle is incrementally trained based on the samples in the local sample library to obtain the first model parameter update amount. The local risk perception model is updated based on the first model parameter update amount, wherein the local risk perception model is used to determine the comprehensive risk of the first vehicle.

7. The vehicle sentry monitoring method according to claim 1, characterized in that, The vehicle sentry monitoring method also includes: For any fifth vehicle in the temporary vehicle group network, the local collaborative strategy model of the fifth vehicle is trained based on the security event data stored locally in the fifth vehicle to obtain the second model parameter update amount. The security event data is the risk event record data generated by the fifth vehicle during historical monitoring. The local collaborative strategy model is used to realize the allocation of monitoring task roles for the fifth vehicle. The parameter update amount of the second model is encrypted to obtain the encrypted gradient; The encryption gradient is transmitted to the coordination node in the temporary vehicle group network, so that the coordination node can aggregate the encryption gradients from multiple vehicles in the temporary vehicle group network to generate model parameters of the global collaborative strategy model, wherein the global collaborative strategy model is used to allocate monitoring task roles to each vehicle in the temporary vehicle group network. The local collaborative strategy model of the fifth vehicle is updated based on the model parameters returned by the coordination node.

8. A vehicle sentry monitoring device, characterized in that, include: A network establishment unit is used to exchange information among multiple vehicles via short-range communication in order to build a temporary vehicle group network. The role allocation unit is used to assign monitoring task roles to each vehicle in the temporary vehicle group network based on the vehicle status information of each vehicle in the temporary vehicle group network. The monitoring task roles include a main monitoring role, an auxiliary sensing role, and a dormant role. The parameter determination unit is used to determine the operating parameters of the monitoring hardware of each vehicle in the temporary vehicle group network based on the monitoring task role of each vehicle in the temporary vehicle group network and the comprehensive risk of the scene in which they are located. The monitoring execution unit is used to broadcast early warning information through the temporary vehicle group network when a risk event is detected by the monitoring hardware, so as to trigger at least some vehicles in the temporary vehicle group network to perform collaborative monitoring operations.

9. A vehicle comprising: The memory and processor are characterized in that the processor, when executing a computer program stored in the memory, implements the steps of the vehicle sentry monitoring method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having stored thereon computer-executable instructions or a computer program, characterized in that, When the computer-executable instructions or the computer program are executed by a processor, the steps of the vehicle sentry monitoring method as described in any one of claims 1 to 7 are implemented.

Images