Method for managing an agent, vehicle and storage medium

Abstract

This application provides a management method for intelligent agents, a vehicle, and a storage medium, relating to the field of intelligent cockpit technology. The method includes: when a first intelligent agent is in a service state, in response to receiving a service request for a second intelligent agent, obtaining lock status information of the first intelligent agent, the lock status information being used to characterize the interruption permission attribute of the first intelligent agent; based on the lock status information, generating a processing decision for the service request, the processing decision being used to determine whether the second intelligent agent is allowed to interrupt the first intelligent agent; and managing the second intelligent agent and the first intelligent agent based on the processing decision; thus solving the technical problem of chaotic intelligent agent management in multi-agent services in related technologies.

Description

Technical Field

[0001] This application relates to the field of intelligent cockpit technology, and more particularly to a method for managing intelligent agents, a vehicle, and a storage medium. Background Technology

[0002] With the continuous iteration and upgrading of intelligent vehicle and intelligent cockpit technologies, in-vehicle software systems are gradually integrating multiple types of intelligent agents to provide diversified proactive services such as vehicle control, navigation, entertainment, communication, life services and system services. The concurrent execution of multiple intelligent agents and the collaborative service of multiple scenarios have become the mainstream form of in-vehicle interaction. The complexity and concurrency of in-vehicle services have increased significantly, which puts forward higher requirements for service scheduling order, conflict management capabilities and operational stability.

[0003] Current vehicle-mounted intelligent agent service management and scheduling technologies still have many shortcomings. Intelligent agents lack a unified and standardized classification system, resulting in chaotic overall management and collaborative scheduling. Service interruption strategies are simple, and services that are being executed are easily interrupted without rules, which seriously affects service continuity and user experience.

[0004] In summary, the relevant technologies suffer from the technical problem of chaotic agent management when multiple agent services are used. Summary of the Invention

[0005] In view of the above problems, this application provides a method, vehicle, and storage medium for managing intelligent agents to improve the management effect of intelligent agents in multi-agent services. The technical solution is as follows: A method for managing intelligent agents, the method comprising: When the first agent is in a service state, in response to receiving a service request for the second agent, the locking status information of the first agent is obtained, and the locking status information is used to characterize the interruption permission attribute of the first agent. Based on the locked state information, a processing decision is generated for the service request, the processing decision being used to determine whether the second agent is allowed to interrupt the first agent; Based on the processing decision, the second intelligent agent and the first intelligent agent are managed.

[0006] In this way, when the first agent is in a service state and receives a service request for the second agent, the locking status information of the first agent is first obtained, and then a processing decision is generated based on the locking status information. Finally, the second agent and the first agent are managed according to the processing decision. By establishing a step-by-step execution logic based on the locking status information, a standardized agent interruption judgment and management process is established, avoiding arbitrary interruptions without basis or irregular service scheduling. This makes the management process when multiple agents provide services concurrently orderly and controllable, while ensuring the scheduling order of services that are being executed and new service requests, and adapting to the needs of scenarios where multiple agents run concurrently.

[0007] Optionally, the lock status information includes interrupting the permission mapping relationship, and the step of generating a processing decision for the service request based on the lock status information includes: If the locking status information indicates that the first agent is in a forced locking state, a decision is generated to prevent the second agent from interrupting the processing of the first agent. If the locking status information indicates that the first agent is in a non-forced locking state, the processing decision is generated based on the interruption of the permission mapping relationship.

[0008] In this way, the locked state information is clearly defined, including the interruption permission mapping relationship. At the same time, the locked state is divided into a mandatory locked state and a non-mandatory locked state. For the mandatory locked state, a processing decision is directly generated to prohibit the second agent from interrupting the first agent. For the non-mandatory locked state, a processing decision is generated based on the interruption permission mapping relationship. By classifying the locked state and matching the differentiated judgment logic, clear uninterruptible protection is provided for agents with mandatory locking attributes. At the same time, the interruption of permission mapping relationship provides a clear judgment basis for non-mandatory locking scenarios, making the processing decision generation logic more targeted and reasonable.

[0009] Optionally, generating the processing decision based on the interruption permission mapping relationship includes: If the interruption permission mapping relationship is to prohibit interruption, a processing decision is generated to prohibit the second agent from interrupting the first agent. If the interruption permission mapping relationship is set to allow interruption, a processing decision is generated that allows the second agent to interrupt the first agent. If the interruption permission mapping relationship is a priority decision, obtain the second priority value corresponding to the second agent and the first priority value corresponding to the first agent; The processing decision is generated based on the first priority value and the second priority value.

[0010] In this way, three types of branch processing logic are set up for interruption permission mapping relationship. When the interruption permission mapping relationship is "disruption prohibited", a processing decision of "disruption prohibited" is generated. When the interruption permission mapping relationship is "interruption allowed", a processing decision of "interruption allowed" is generated. When the interruption permission mapping relationship is "priority adjudication", a processing decision is generated based on the first priority value corresponding to the first agent and the second priority value corresponding to the second agent. Through hierarchical judgment logic, it not only realizes rapid decision-making in scenarios with clear permissions, improving the response efficiency of service requests, but also covers scenarios where the permission mapping relationship is not clearly defined through the priority adjudication fallback logic, making the processing decision generation logic more comprehensive and the judgment process more hierarchical.

[0011] Optionally, generating the processing decision based on the first priority value and the second priority value includes: If the first priority value is less than the second priority value, a decision is generated that allows the second agent to interrupt the processing of the first agent. If the first priority value is greater than or equal to the second priority value, a decision is generated to prohibit the second agent from interrupting the processing of the first agent.

[0012] In this way, for priority adjudication scenarios, a comparison and judgment rule based on the first priority value and the second priority value is clearly defined. When the first priority value is less than the second priority value, a processing decision is generated that allows the second agent to interrupt the first agent. When the first priority value is greater than or equal to the second priority value, a processing decision is generated that prohibits the second agent from interrupting the first agent. By comparing the quantified priority values, a standardized priority adjudication judgment standard is established, avoiding subjective arbitrariness in the priority judgment process. This makes the processing decision generation process in priority adjudication scenarios predictable and reproducible, ensuring the fairness and orderliness of the multi-agent scheduling process.

[0013] Optionally, the management of the second intelligent agent and the first intelligent agent based on the processing decision includes: If the processing decision is to allow interruption, execute the interruption operation, suspend or stop the service of the first agent, and start the service of the second agent; If the processing decision is to disallow interruption, the service request of the second agent is added to the waiting queue, while the first agent continues to provide service.

[0014] In this way, corresponding agent management actions are matched for different processing decisions. When the processing decision allows interruption, the interruption operation is executed, pausing or stopping the service of the first agent and starting the service of the second agent. When the processing decision does not allow interruption, the service request of the second agent is added to the waiting queue and the first agent continues to provide service. Through management actions that correspond one-to-one with the processing decisions, the processing decisions can be fully implemented. At the same time, the setting of the waiting queue ensures that new service requests are not lost in scenarios where interruption is not allowed. This achieves both orderly service switching in scenarios where interruption is allowed and ensures the continuity of the current service and proper handling of new service requests in scenarios where interruption is not allowed.

[0015] Optionally, after performing the interruption operation, the method further includes: Save the service progress status of the first intelligent agent; After the second agent finishes its service, the first agent's service is restored based on the service progress status, and a restoration prompt is output.

[0016] In this way, after the interruption operation is performed, the service progress status of the first agent is saved. After the service of the second agent is completed, the service of the first agent is restored based on the saved service progress status, and a restoration prompt is output. By saving and restoring the service progress status, the loss of service progress caused by the interruption operation is avoided, and the interrupted service can continue to execute after the new service is completed, ensuring the continuity of service. At the same time, by outputting the restoration prompt, users can know the service restoration status in a timely manner, improving the user experience of the service process.

[0017] Optionally, the method further includes: Obtain feedback data from users regarding service interruption and disruption operations; Based on the feedback data, the user's usage habits and service preferences are determined; Based on the aforementioned usage habits and service preferences, adjust the lock status information of the intelligent agent.

[0018] In this way, by obtaining feedback data from users regarding service interruptions and disruptions, the user's usage habits and service preferences are determined based on the feedback data. The locking status information of the intelligent agent is then adjusted according to the user's usage habits and service preferences, establishing a dynamic adjustment mechanism for the locking status information. This allows the locking status information of the intelligent agent to adapt to the personalized needs of different users, avoiding the problem that fixed locking status rules cannot match the usage habits of different users. This ensures that the management rules of the intelligent agent can continuously meet the actual needs of the users.

[0019] Optionally, the method further includes: Multiple intelligent agents are divided into multiple types according to their functional attributes, and each type corresponds to an independent service domain; Within each type, the locking status information is pre-stored. The locking status information includes at least hierarchical locking rules and interruption permission mapping relationships. The hierarchical locking rules include the locking status information setting of the first intelligent agent and the interruption rule setting when the service request of the second intelligent agent is triggered. The interruption permission mapping relationship is the interruptible relationship, non-interruptible relationship and priority mapping between the first intelligent agent and other intelligent agents.

[0020] In this way, multiple intelligent agents are divided into multiple types according to their functional attributes, with each type corresponding to an independent service domain. Locking status information, including at least hierarchical locking rules and interruption permission mapping relationships, is pre-stored within each type. The specific content covered by hierarchical locking rules and interruption permission mapping relationships is clearly defined. By managing intelligent agents by type and service domain according to functional attributes, classified control of intelligent agents with different functions is achieved, avoiding rule overlap and management confusion between intelligent agents of different functional types. At the same time, the standardized rules pre-stored within the same type provide a unified and clear basis for the management of intelligent agents within the same domain, facilitating the independent configuration and maintenance of rules for each type in the future, and improving the scalability and maintainability of the intelligent agent management system.

[0021] A management device for an intelligent agent, the device comprising: The acquisition module is configured to, when the first agent is in a service state, in response to receiving a service request for the second agent, acquire the lock status information of the first agent, wherein the lock status information is used to characterize the interruption permission attribute of the first agent. A generation module is used to generate a processing decision for the service request based on the locking status information, the processing decision being used to determine whether the second agent is allowed to interrupt the first agent; The management module is used to manage the second intelligent agent and the first intelligent agent based on the processing decision.

[0022] Optionally, the generation module is also used for: If the locking status information indicates that the first agent is in a forced locking state, a decision is generated to prevent the second agent from interrupting the processing of the first agent. If the locking status information indicates that the first agent is in a non-forced locking state, the processing decision is generated based on the interruption of the permission mapping relationship.

[0023] Optionally, the generation module is also used for: If the interruption permission mapping relationship is to prohibit interruption, a processing decision is generated to prohibit the second agent from interrupting the first agent. If the interruption permission mapping relationship is set to allow interruption, a processing decision is generated that allows the second agent to interrupt the first agent. If the interruption permission mapping relationship is a priority decision, obtain the first priority value corresponding to the first agent and the second priority value corresponding to the second agent; The processing decision is generated based on the first priority value and the second priority value.

[0024] Optionally, the generation module is also used for: If the first priority value is less than the second priority value, a decision is generated that allows the second agent to interrupt the processing of the first agent. If the first priority value is greater than or equal to the second priority value, a decision is generated to prohibit the second agent from interrupting the processing of the first agent.

[0025] Optionally, the management module is also used for: If the processing decision is to allow interruption, execute the interruption operation, suspend or stop the service of the first agent, and start the service of the second agent; If the processing decision is to disallow interruption, the service request of the second agent is added to the waiting queue, while the first agent continues to provide service.

[0026] Optionally, after the interruption operation is performed, the agent's management device also includes a save module: A storage module is used to save the service progress status of the first intelligent agent; After the second agent finishes its service, the first agent's service is restored based on the service progress status, and a restoration prompt is output.

[0027] Optionally, the management device for the intelligent agent may also include an adjustment module: The adjustment module is used to obtain feedback data from users related to service interruption and disruption operations; Based on the feedback data, the user's usage habits and service preferences are determined; Based on the aforementioned usage habits and service preferences, adjust the lock status information of the intelligent agent.

[0028] Optionally, the management device for the intelligent agent may also include a storage module: The storage module is used to divide multiple intelligent agents into multiple types according to their functional attributes, with each type corresponding to an independent service domain. Within each type, the locking status information is pre-stored. The locking status information includes at least hierarchical locking rules and interruption permission mapping relationships. The hierarchical locking rules include the locking status information setting of the first intelligent agent and the interruption rule setting when the service request of the second intelligent agent is triggered. The interruption permission mapping relationship is the interruptible relationship, non-interruptible relationship and priority mapping between the first intelligent agent and other intelligent agents.

[0029] A vehicle includes an onboard controller, the onboard controller including a memory, a processor and a computer program stored in the memory, wherein when the processor executes the computer program, it implements a management method for any of the optional intelligent agents described above.

[0030] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a management method for any of the optional intelligent agents described above.

[0031] By utilizing the aforementioned technical solution, this application provides a management method for intelligent agents. By acquiring the locking status information of the first intelligent agent, a clear and unified core judgment criterion is established for determining whether subsequent services should be interrupted. This fundamentally avoids arbitrary interruption behavior without basis or rules when multiple intelligent agents provide concurrent services. Based on this locking status information, a processing decision is generated to determine whether the second intelligent agent is allowed to interrupt the first intelligent agent. This ensures that the interruption permission judgment process revolves entirely around the interruption permission attribute of the first intelligent agent, guaranteeing the consistency and predictability of the judgment logic. This solves the service conflict problems caused by disordered multi-agent service scheduling and ambiguous interruption rules in related technologies. Finally, based on the generated processing decision, corresponding management actions are executed on the second and first intelligent agents. This not only ensures the service continuity of the currently executing first intelligent agent but also realizes the orderly processing of newly initiated service requests from the second intelligent agent. It fully covers the entire process of service scheduling in multi-agent concurrent scenarios, effectively improving the stability of multi-agent operation and the rationality of scheduling.

[0032] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0033] 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 the scope of this application. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1One of the flowcharts of the intelligent agent management method provided in this application is shown; Figure 2 This is a second schematic flowchart illustrating the intelligent agent management method provided in an embodiment of this application; Figure 3 The third flowchart illustrates the intelligent agent management method provided in this application embodiment; Figure 4 A schematic diagram of the structure of a smart agent management device provided in an embodiment of this application is shown. Detailed Implementation

[0034] Exemplary embodiments of the present application will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this application will be thorough and complete, and will fully convey the scope of the present application to those skilled in the art.

[0035] To address the technical problem of chaotic agent management in multi-agent services in related technologies, this application provides an agent management method, such as... Figure 1 As shown, Figure 1 This is a schematic flowchart illustrating a method for managing intelligent agents provided in an embodiment of this application. The method includes: S11. When the first intelligent agent is in a service state, in response to receiving a service request for the second intelligent agent, obtain the lock status information of the first intelligent agent.

[0036] The first intelligent agent refers to the in-vehicle intelligent service unit currently in the service execution state and occupying the in-vehicle service execution resources in the vehicle system, such as the body control intelligent agent performing window lifting control, the driver assistance intelligent agent performing automatic parking, and the audio and video playback intelligent agent playing music. The second intelligent agent refers to the in-vehicle intelligent service unit corresponding to the new service request received by the vehicle system that has not yet started the service. It is the initiator that requests to occupy the in-vehicle service execution resources and may interrupt the service of the first intelligent agent, such as the audio and video playback intelligent agent corresponding to the music playback request initiated by the user's voice, and the call function intelligent agent corresponding to the incoming call request triggered by the vehicle. The lock status information is a set of pre-configured attributes that are strongly bound to the functional attributes of the first intelligent agent. It is used to clarify the core rules for whether the service of the intelligent agent can be interrupted by other intelligent agents during the service operation, that is, to characterize the interruption permission attribute of the first intelligent agent.

[0037] Specifically, the system monitors the operational status of all in-vehicle intelligent agents in real time, while continuously listening for service requests triggered by users, the vehicle system, or in-vehicle sensors. When two triggering conditions are met simultaneously, the execution process of this step is initiated: there is at least one intelligent agent in the process of performing a service, i.e., the first intelligent agent; and a service request is received for another intelligent agent that is not in a service state, i.e., a service request for the second intelligent agent.

[0038] Specifically, after completing the trigger condition verification, the first intelligent agent currently executing the service is locked, and its unique identity, service domain, and current service scenario are determined; at the same time, the second intelligent agent initiating the new service request is locked, and its unique identity, service domain, and service function type are determined; based on the unique identity of the first intelligent agent, the locking status information corresponding to the first intelligent agent is retrieved from the pre-configured storage area.

[0039] For example, while driving, a user initiates a request to "open the driver's side window" via voice command, activating the vehicle control agent to perform the window raising / lowering operation. At this time, the vehicle control agent is the first agent, in a service-ready state. During the window raising / lowering process, the vehicle detects an incoming Bluetooth call from the mobile phone, triggering a service activation request for the call function agent. At this time, the call function agent is the second agent. After receiving the service request, the vehicle system meets the triggering conditions, locks the vehicle control agent (first agent) that is currently performing the service, and retrieves the corresponding lock status information of the vehicle control agent from the storage area.

[0040] In one specific embodiment, all in-vehicle intelligent agents with independent service capabilities within the vehicle system are identified, the core functional attributes of each intelligent agent are extracted, and all intelligent agents are divided into multiple non-overlapping types according to the consistency of functional attributes. Each type corresponds to an independent service domain, and the intelligent agents within each service domain have the same type of in-vehicle functional attributes.

[0041] For each independent service domain, the locking status information is configured within the domain according to the configuration file. The locking status information includes two core categories: hierarchical locking rules and interruption permission mapping relationships. The hierarchical locking rules are divided into two parts: first, configuring exclusive locking status information settings for each agent within the domain, clarifying the interruption permission attributes of each agent; second, configuring interruption rules for all agents within the domain when service requests are triggered, clarifying the triggering conditions, verification process, and execution logic when a new service request is initiated. The interruption permission mapping relationship configures corresponding interruptible, non-interruptible, and priority mapping relationships between every two agents within the domain, clarifying the interruption permission agreement between any two agents within the domain.

[0042] Specifically, each intelligent agent within the vehicle's infotainment system is configured with its own list of interruptible objects, list of non-interruptible objects, and priority value. The list of interruptible objects records the unique identifiers of other intelligent agents that are allowed to actively interrupt the intelligent agent when it initiates an interruption request. The list of non-interruptible objects records the unique identifiers of other intelligent agents that are absolutely prohibited from actively interrupting the intelligent agent when it initiates an interruption request. The priority value is a quantified value strongly bound to the functional safety level of the intelligent agent, and its magnitude is positively correlated with the intelligent agent's security priority and functional importance. Within each service domain, a unified interruption permission mapping relationship is established by integrating the list of interruptible objects, list of non-interruptible objects, and priority values ​​of all intelligent agents within the domain.

[0043] The interruptible list is a whitelist corresponding to each agent, recording the unique identifiers of all other agents that the agent has the right to actively interrupt when initiating a service request. In other words, the agent can interrupt other services in the list. The non-interruptible list is a blacklist corresponding to each agent, recording the unique identifiers of all other agents that the agent absolutely has no right to actively interrupt when initiating a service request. In other words, the agent absolutely cannot interrupt other services in the list. The priority value is a quantified value tied to the functional importance and security level of each agent. The higher the value, the higher the service priority of the agent. For example, in an in-vehicle scenario, agents related to driving safety have the highest priority value, while entertainment and lifestyle agents have relatively lower priority values.

[0044] For example, taking the call function agent within the communication service domain as an example, assuming the interruptible object list records the unique identifiers of the vehicle control agent, audio / video playback agent, and navigation broadcast agent, it means that when the call function agent initiates a service request, it can actively interrupt the services of these agents; the non-interruptible object list records the unique identifiers of the automatic parking agent and emergency driving assistance agent, meaning that the call function agent absolutely cannot interrupt the services of these safety-related agents; the priority value is 8, and the vehicle control agent is configured with a priority value of 5, the audio / video playback agent with a priority value of 3, the navigation broadcast agent with a priority value of 6, and the automatic parking agent with a priority value of 10; within the communication service domain, vehicle control service domain, and entertainment service domain, an interruption permission mapping relationship is established and integrated within each domain. Taking the vehicle control service domain as an example, the mapping relationship clearly defines the automatic parking agent and all other agents. The rules for all entities are that interruption is not allowed. The rule between the window control agent and the call function agent is that interruption is allowed. The rule between the window control agent and the audio / video playback agent is based on priority. When the user is using voice control to raise or lower the window, the vehicle body control agent (the first agent) is providing service. At this time, the vehicle receives a Bluetooth call, and the call function agent (the second agent) initiates a service request. First, it checks the interruptible list of the call function agent. If the vehicle body control agent is found in the list, it is directly determined that interruption is allowed, completing a quick judgment. If the user initiates a music playback request at this time, the audio / video playback agent, as the second agent, checks its interruptible and non-interruptible lists. If the vehicle body control agent is not listed, it proceeds to judge based on priority value. The priority value of the audio / video playback agent (3) is compared with the priority value of the vehicle body control agent (5), and the final judgment is completed according to the rules.

[0045] Specifically, the lock status information configured in each service domain is stored in an independent storage partition corresponding to that service domain. When the vehicle system starts, the lock status information of all service domains is synchronously loaded into memory, and can be retrieved directly from memory in subsequent calls, thus improving the efficiency of rule retrieval.

[0046] Among them, functional attributes refer to the category characteristics of the core in-vehicle functions possessed by the in-vehicle intelligent agent, which is the basis for classifying the service capabilities of the intelligent agent. For example, functions such as controlling vehicle hardware, providing navigation, and providing audio and video entertainment are all different functional attributes. Service domains refer to the independent management units corresponding to the set of intelligent agents divided according to the same type of functional attributes. Each service domain corresponds to a type of in-vehicle function, and all intelligent agents within the domain have the same type of functional attributes. The functional attributes between different service domains do not overlap, such as vehicle control and navigation. Hierarchical locking rules refer to the hierarchical management rules formulated within the service domain for the locking status and interruption behavior of intelligent agents. They are divided into two levels: setting the locking status of the intelligent agent itself and setting the interruption behavior when a new service request is initiated. Interruption permission mapping relationship refers to the rule agreement between any two intelligent agents within the service domain regarding whether the intelligent agent initiating the interruption request can interrupt the service of the interrupted intelligent agent. It is a one-to-one correspondence of interruption permissions between two intelligent agents.

[0047] In this embodiment, all vehicle-mounted intelligent agents are classified and managed by functional attributes. Locking status information containing hierarchical locking rules and interruption permission mapping relationships is pre-stored in each service domain. This not only realizes the classified control of intelligent agents of different functional types, avoiding rule overlap and management chaos across functional types, but also provides a unified and clear basis for judging interruption permissions of intelligent agents within the same domain through standardized rules pre-configured within the same domain. At the same time, the domain-based architecture design facilitates the independent configuration, iteration and maintenance of rules for each functional type, greatly improving the scalability and maintainability of the intelligent agent management system.

[0048] For example, suppose all in-vehicle intelligent agents are divided into 6 types according to their functional attributes, corresponding to 6 independent service domains: vehicle control service domain, navigation service domain, entertainment service domain, communication service domain, lifestyle service domain, and system service domain. The vehicle control service domain includes intelligent agents such as body control, power control, driving assistance, and energy management; the navigation service domain includes intelligent agents such as route planning, traffic information, and charging services; and the entertainment service domain includes intelligent agents such as audio / video playback, game applications, and interactive functions. Then, for the vehicle control service domain, the locking status information within the domain is configured. The hierarchical locking rules section configures locking status settings for each intelligent agent within the domain, such as setting a forced locking attribute for the automatic parking intelligent agent and a lock attribute for window control... The intelligent agent is configured with the attribute that it can be interrupted by high-priority services; then the interruption rules are configured when a service request is triggered, specifying that when a new service request is initiated within the domain, the lock status of the interrupted intelligent agent must be checked first, and then the execution flow of the interruption permission mapping relationship is queried; the interruption permission mapping relationship part: configure corresponding interruption rules for every two intelligent agents within the domain, for example, set the call function intelligent agent to interrupt the window control intelligent agent, and the window control intelligent agent cannot interrupt the automatic parking intelligent agent, thus completing the domain rule configuration; finally, the lock status information of the 6 service domains is stored in the corresponding independent storage partitions, and when the vehicle system is powered on and started, all rules are loaded into memory, so that when rules in the vehicle control service domain need to be called later, they can be quickly retrieved directly from memory.

[0049] In this implementation, each agent is configured with a list of interruptible objects, a list of non-interruptible objects, and a unique priority value. A hierarchical interruption permission judgment logic is established, which prioritizes list matching and uses priority as a fallback. The blacklist and whitelist design enables rapid matching and judgment of interruption permissions in normal scenarios, improving the response efficiency of service requests. The priority value provides a fallback for scenarios not covered by the list, balancing the efficiency of the judgment process with the flexibility of rule configuration. At the same time, the unified interruption permission mapping relationship within the domain ensures the consistency and predictability of interruption rules for agents within the same domain, adapting to the differentiated management needs of agents with different functions in the vehicle scenario.

[0050] In this embodiment, by selectively acquiring lock status information representing the current service interruption permission attribute in a specific scenario where the current service is running and a new service request is initiated, it provides a core basis for subsequent interruption permission judgment, fundamentally avoiding arbitrary service interruption behavior without rules or basis. Furthermore, by selectively acquiring the exclusive attribute information of the currently running service, it ensures a strong correlation between the judgment process and the current service running state, avoiding rule overlap and judgment logic confusion in the multi-agent scheduling process. At the same time, it establishes a compliant trigger entry point for the entire agent management process, ensuring the orderliness and predictability of the entire multi-agent service scheduling process.

[0051] S12. Based on the lock status information, generate a processing decision for the service request.

[0052] Among them, the processing decision is used to determine whether the second agent is allowed to interrupt the first agent; the lock state type refers to the category divided according to the interruption permission attribute of the first agent, which can specifically include three categories: forced lock state, normal lock state, and interruptible state.

[0053] Specifically, the completeness and validity of the locking status information of the first agent obtained in step S11 are verified. It is confirmed that the hierarchical locking rules and interruption permission mapping relationships contained in the locking status information are all in a valid and effective state, and are completely matched with the identity identifier and the service domain to which the first agent belongs. If the verification is successful, the hierarchical judgment stage is entered. If the verification fails, a default decision to prohibit interruption is generated.

[0054] Specifically, according to the pre-set judgment levels, the judgment process is executed in sequence through three levels: lock state type identification, interruption permission mapping relationship matching, and priority fallback judgment. First, the lock state type of the first agent is identified, and the corresponding judgment rule is matched according to the lock state type. Then, the interruption permission between the second agent and the first agent is matched based on the interruption permission mapping relationship. If there is no clear matching rule, the priority value comparison judgment is entered.

[0055] Specifically, based on the final result of the hierarchical judgment, standardized processing decisions are generated. These processing decisions can be divided into two categories: allowing the second intelligent agent to interrupt the first intelligent agent and prohibiting the second intelligent agent from interrupting the first intelligent agent. At the same time, each processing decision is attached with a corresponding decision basis identifier, and the lock state information entry on which the processing decision is based is recorded.

[0056] In this embodiment, based on the pre-acquired lock status information, a processing decision is generated to determine whether the second agent is allowed to interrupt the first agent. A standardized and reproducible interruption permission judgment process is established, which ensures that the generation process of the processing decision is completely matched with the interruption permission attribute of the first agent, avoiding the arbitrariness and unpredictability of the judgment logic. Furthermore, the judgment mode based on a single core criterion makes the decision generation logic clear and controllable, fundamentally solving the service conflict problem caused by the ambiguity of multi-agent interruption rules and the lack of uniform judgment standards in the prior art. At the same time, it provides a clear execution basis for subsequent agent management actions.

[0057] S13. Based on processing decisions, manage the second intelligent agent and the first intelligent agent.

[0058] Specifically, the processing decision generated in step S12 is analyzed to extract the core conclusions (interruption allowed / interruption prohibited), the basis for the decision, and the identity identifiers of the corresponding first and second intelligent agents, thereby clarifying the execution objectives and rules of this management action.

[0059] Specifically, based on the decision conclusion obtained from the analysis, pre-set management actions corresponding one-to-one with the decision conclusion are matched. If the decision conclusion allows interruption, the matched management actions include: sending an interrupt command to the first intelligent agent, saving the service progress status of the first intelligent agent, releasing the service resources occupied by the first intelligent agent, and starting the service of the second intelligent agent. If the decision conclusion prohibits interruption, the matched management actions include: adding the service request of the second intelligent agent to the waiting queue of the corresponding service domain, keeping the service of the first intelligent agent running continuously, and sending a prompt message to the user that the service request has entered the waiting queue. The matched management actions are executed in sequence. After the execution is completed, the service status and resource occupation status of the first and second intelligent agents are updated synchronously, and the status change information is synchronized to all relevant modules of the vehicle system to ensure the consistency of the intelligent agent status within the system.

[0060] Among them, the service progress status refers to the data set of completed service execution progress, running parameters, execution environment, etc. when the first intelligent agent's service is interrupted. It is used to continue the execution state before the interruption when the first intelligent agent's service is restored, so as to avoid loss of service progress. The waiting queue refers to the ordered queue set up for each service domain in the vehicle system to store unexecuted service requests. The service requests in the queue are sorted from high to low priority values. After the system resources are released, the service requests in the queue are executed in order. Service resources refer to the hardware and software resources required by the intelligent agent to execute services in the vehicle system, including audio output channels, voice interaction channels, vehicle hardware control permissions, processor computing resources, memory resources, etc.

[0061] For example, suppose a user is playing music in the car, and the audio / video playback agent (first agent) is in service mode. At this time, the vehicle receives a Bluetooth call, and the call function agent (second agent) initiates a service request. The preceding steps generate a processing decision that allows the call function agent to interrupt the audio / video playback agent. The system analyzes the processing decision and extracts the core conclusion that interruption is allowed. The corresponding first agent is the audio / video playback agent, and the second agent is the call function agent. The decision is based on the interruption permission mapping relationship between the two. Subsequently, the system matches the standardized management actions corresponding to allowing interruption, in the following order: sending an interrupt command to the audio / video playback agent, saving the current music playback progress, playback mode, volume, and other service progress status, releasing the audio output channel occupied by the audio / video playback agent, and starting the call function agent's incoming call service. The above management actions are executed in sequence. After the execution is completed, the status of the audio / video playback agent is updated to paused service, the status of the call function agent is updated to service in progress, and the status information is synchronized to the vehicle's audio management module and interactive display module. At the same time, the incoming call interface is displayed on the vehicle's screen.

[0062] In one specific embodiment, feedback data related to service interruption and disruption operations is continuously collected from users through the vehicle's interactive interface. This feedback data includes users' proactive evaluation of disruption operations, data on service recovery operations, data on manual adjustments to the execution order of service requests, data on repeatedly initiated service requests, and user behavior data recorded by the vehicle system after the disruption operation. The collected feedback data is cleaned, categorized, and statistically analyzed to extract users' acceptance of intelligent agent disruption operations in different scenarios, their priority preferences for different types of intelligent agent services, and their demand for service continuity. This helps determine users' stable usage habits and service preferences. Based on the analyzed usage habits and service preferences, the locking status information of the intelligent agents is adjusted accordingly. Adjustments include modifying the locking status type of the intelligent agent, updating the list of interruptible and non-interruptible objects, adjusting the priority value of the intelligent agent, and optimizing the disruption permission mapping relationship. After the adjustments are completed, the rule data in the system storage and memory is updated synchronously.

[0063] Feedback data refers to all behavioral data and proactive evaluation data generated by users during the use of the vehicle system related to service interruptions and disruptions; usage habits and service preferences are stable and reproducible behavioral tendencies formed by users during long-term use of the vehicle system's intelligent agent services, including priority preferences for different types of intelligent agent services, acceptance of service interruptions, and needs for service continuity.

[0064] For example, during daily operation, the vehicle's infotainment system continuously collects user feedback data. It was found that in parked scenarios, users repeatedly initiated music playback requests while the navigation assistant was broadcasting route information, and each time they manually paused the navigation broadcast. Furthermore, users repeatedly gave positive feedback regarding the interruption of navigation broadcasts by music. This feedback data was then statistically analyzed. A data threshold of 5 instances was set; when the same operation occurred 6 times, the threshold requirement was met. The analysis concluded that in parked scenarios, the user prioritized entertainment-related services over navigation broadcast services and accepted the habit of the music playback assistant interrupting the navigation broadcast assistant. Based on this preference conclusion, the system adaptively adjusted the lock status information. In the interruption permission mapping relationship between the entertainment and navigation service domains, the rules for the music playback assistant and the navigation broadcast assistant in parked scenarios were modified to allow interruption. Simultaneously, the list of interruptible objects for the music playback assistant was updated to include the navigation broadcast assistant identifier in parked scenarios. After the adjustment, the rules stored in the system and in memory were updated synchronously, and a rule adjustment prompt message was sent to the user. The rules took effect immediately upon user confirmation.

[0065] In this embodiment, based on the generated processing decisions, corresponding standardized management actions are executed on the first and second intelligent agents, forming a complete closed loop for multi-agent service scheduling. This ensures the consistency and controllability of the multi-agent scheduling process through management actions that correspond one-to-one with the processing decisions. Furthermore, through differentiated management actions, it simultaneously considers the continuity of current services and the processing of new service requests, avoiding the problems of progress loss or the discarding of new service requests due to service interruptions. This effectively improves the stability of multi-agent service scheduling and the user experience.

[0066] In the above scheme, by obtaining the lock status information of the first agent, a clear and unified core judgment basis is established for determining whether subsequent services should be interrupted, fundamentally avoiding arbitrary interruption behavior without basis or rules when multiple agents provide concurrent services. Based on this lock status information, a processing decision is generated to determine whether the second agent is allowed to interrupt the first agent. The interruption permission judgment process revolves entirely around the interruption permission attribute of the first agent, ensuring the consistency and predictability of the judgment logic. This solves the service conflict problem caused by disordered multi-agent service scheduling and ambiguous interruption rules in related technologies. Finally, based on the generated processing decision, corresponding management actions are executed on the second agent and the first agent. This not only ensures the service continuity of the currently executing first agent, but also realizes the orderly processing of newly initiated service requests from the second agent. It fully covers the entire process of service scheduling in multi-agent concurrent scenarios, effectively improving the stability of multi-agent operation and the rationality of scheduling.

[0067] In some embodiments, such as Figure 2As shown, the lock status information includes the interruption of permission mapping relationships. Based on the lock status information, a processing decision for the service request is generated, including: S121. If the locking status information indicates that the first agent is in a forced locking state, generate a decision to prohibit the second agent from interrupting the processing of the first agent.

[0068] Among them, the forced lock state is the highest level of vehicle intelligent agent lock state. It means that during the service execution process, the service cannot be interrupted regardless of the priority of the second intelligent agent that initiated the request. Examples of intelligent agent service scenarios directly related to driving safety include automatic parking, emergency driving assistance, highway cruise control, and emergency braking execution.

[0069] Specifically, from the first agent's lock status information obtained in step S11, the lock status type identifier of the agent is extracted. By matching the lock status type identifier, it is confirmed whether the first agent is in a forced lock state. At the same time, it is confirmed that the service scenario currently being executed by the first agent matches the safety-critical scenario corresponding to the forced lock state, eliminating abnormal situations of mismatched identifiers. At the same time, the current driving status of the vehicle is verified to confirm that the safety-critical scenario is running continuously, and there are no cases of user-initiated termination, service completion, or safety risk resolution.

[0070] Specifically, after completing the above two verifications, a standardized decision to prohibit the second agent from interrupting the first agent's processing is directly generated. At the same time, the decision is attached with a corresponding security rule number and scenario verification results as traceability basis, and the decision generation time is recorded synchronously.

[0071] In this embodiment, a rigid service protection barrier of the highest level is established for vehicle safety-critical scenarios. By identifying the forced lock state of the first intelligent agent, a processing decision that prohibits any second intelligent agent from interrupting is directly generated. This not only avoids driving safety risks caused by unnecessary interruptions of safety-critical services from the root, meeting the core requirements of vehicle functional safety, but also simplifies the judgment process of the highest safety level scenario through the single-condition direct decision mode, shortening the decision-making time. At the same time, through clear rule boundaries, the judgment logic of safety scenarios and normal scenarios is avoided from being confused, ensuring the safety bottom line of the multi-agent scheduling system.

[0072] S122. If the locking status information indicates that the first agent is in a non-forced locking state, a processing decision is generated based on the interruption of the permission mapping relationship.

[0073] Among them, the non-forced locking state is the locking state of the vehicle intelligent agent other than the forced locking state. It includes two categories: regular locking state and interruptible state. For example, regular service scenarios that are not the highest level of driving safety, such as window control, air conditioning adjustment, music playback, and navigation broadcast, can be interrupted by other intelligent agents during the service execution process according to pre-configured rules.

[0074] Specifically, when the first agent's lock state is non-forced lock state, the security level of the scenario currently served by the first agent is checked to confirm that there are no triggering conditions for a temporary upgrade to a forced lock state. Based on the unique identity and service domain information of the first and second agents, the exclusive interruption permission mapping relationship between the two agents is retrieved. If the two agents belong to different service domains, the cross-domain service general rule is retrieved first, and then the corresponding cross-domain interruption permission mapping relationship is matched. The retrieved interruption permission mapping relationship is used as the core judgment basis to enter the subsequent hierarchical judgment process, and the corresponding processing decision is matched according to the different types of mapping relationships.

[0075] Specifically, if the interruption permission mapping relationship is "disruption prohibited", a processing decision is generated to prohibit the second agent from interrupting the first agent; that is, from the interruption permission mapping relationship retrieved in step S122, the permission type identifier between the first agent and the second agent is extracted, and it is confirmed that the permission type identifier is the "disruption prohibited" type. That is, the pre-configured rules clearly stipulate that the second agent has no right to actively interrupt the service being executed by the first agent; it is verified whether there is an exemption condition for the "disruption prohibited" rule in the current scenario. Exemption conditions can include two categories: vehicle triggering emergency safety warning and user manually confirming priority execution of the second agent's service. If there is no exemption condition, the "disruption prohibited" rule is directly matched; if there is an exemption condition, the priority decision is executed.

[0076] Specifically, if the permission type between the first and second intelligent agents is marked as interruptible, that is, the interruption permission mapping relationship is marked as interruptible, the current service execution progress and resource usage status of the first intelligent agent are obtained, and it is verified whether the current service is at an interruptible execution node to avoid functional abnormalities caused by interruption at critical nodes of service execution; after the verification is completed, a processing decision that allows the second intelligent agent to interrupt the first intelligent agent is directly generated.

[0077] Specifically, if the interruption of the permission mapping relationship is a priority decision, the first priority value corresponding to the first intelligent agent is retrieved based on the identity of the first intelligent agent; the second priority value corresponding to the second intelligent agent is retrieved based on the identity of the second intelligent agent; both priority values ​​are quantitative values ​​that are pre-bound to the functional attributes of the intelligent agent, the frequency of user use, and the security importance level, and the higher the value, the higher the service priority.

[0078] Specifically, the two priority values ​​are compared, and the corresponding decision rules are matched according to the comparison results. When the first priority value is less than the second priority value, it means that the second agent that initiated the request has a higher service priority, and a processing decision is generated to allow the second agent to interrupt the first agent. When the first priority value is greater than or equal to the second priority value, it means that the second agent that initiated the request has a service priority equal to or lower than the first agent currently providing the service, and a processing decision is generated to prohibit the second agent from interrupting the first agent. At the same time, the specific values ​​of the two values, the comparison results, and the rule basis are recorded to form a complete decision log.

[0079] In this embodiment, a standardized decision generation process based on interrupting permission mapping relationships is established for routine service scenarios that are not safety-critical in vehicles. This process provides a flexible permission judgment space for routine service scenarios such as window control, music playback, and navigation broadcast by classifying non-mandatory locking states, adapting to the diverse daily service needs of vehicles. Furthermore, by using a unified interruption permission mapping relationship as the core of judgment, the consistency and predictability of decision generation logic in routine scenarios are ensured, avoiding arbitrary interruptions without rules. At the same time, a balance is achieved between the rigid constraints of safety scenarios and the flexible scheduling of routine scenarios, thus improving the full-scenario coverage of the multi-agent scheduling system.

[0080] In the above scheme, the mandatory locking state of S121 directly prohibits interruption rules, establishing protection rules for key driving safety scenarios such as automatic parking and emergency driving assistance. In the non-mandatory locking state of S122, the hierarchical decision-making logic based on the interruption permission mapping relationship provides standardized permission judgment basis for conventional service scenarios such as window control, entertainment interaction, and navigation broadcast. This achieves a balance between rigid constraints in safety scenarios and flexible scheduling in conventional scenarios, solving the core problems of ambiguous interruption rules, chaotic judgment logic, and insufficient protection in safety scenarios in related technologies. At the same time, it provides an execution basis for subsequent service execution links, making the decision-making link of the entire multi-agent scheduling process form a logically rigorous and fully adaptable closed loop.

[0081] In some embodiments, such as Figure 3 As shown, based on processing decisions, the management of the second and first intelligent agents includes: S131. If the processing decision is to allow interruption, execute the interruption operation, suspend or stop the service of the first intelligent agent, and start the service of the second intelligent agent.

[0082] The interruption operation refers to the standardized interruption operation performed on the first intelligent agent that is performing the service in order to start the service of the second intelligent agent. It includes the entire process of sending the interruption command, locking the execution permission, handling the service status, and releasing shared resources, rather than a single service termination command.

[0083] Specifically, if the decision is to allow interruption, a standardized interrupt command is sent to the first intelligent agent. The interrupt command may include an interrupt type identifier, an interrupt execution time limit, and progress saving requirements. At the same time, the service execution permissions of the first intelligent agent are locked, preventing it from continuing to occupy core service resources. Based on the type of interrupt command and the service attributes of the first intelligent agent, the operation of pausing or stopping the service is selected. Pausing the service preserves the service runtime environment and basic data, only terminating the foreground execution process. Stopping the service terminates the entire service process and releases all occupied resources. After completing the service handling and resource release of the first intelligent agent, the service startup interface of the second intelligent agent is called, passing in the core parameters of the service request, starting the service execution process of the second intelligent agent, and updating the service status of the second intelligent agent to "in execution".

[0084] In one specific embodiment, while sending an interrupt command to the first intelligent agent and performing an interruption operation, a progress saving process is triggered. The core execution data, running parameters, and environment configuration of the first intelligent agent's current service are extracted, encapsulated into a service progress status data packet, and stored in the storage area. At the same time, an identifier number bound to the first intelligent agent and this interruption event is added to the data packet. After the second intelligent agent service starts, the service running status of the second intelligent agent is continuously monitored, and the triggering conditions for service termination are detected in real time. The triggering conditions for service termination include the completion of the entire process of the second intelligent agent service, the user actively terminating the second intelligent agent service, and abnormal termination of service execution. As long as any one of these conditions is met, the subsequent service recovery process is triggered.

[0085] Specifically, after the recovery process is triggered, the corresponding service progress status data packet is retrieved from the local storage area based on the data packet's identifier. The integrity of the service progress status data packet is first verified. If the verification passes, the execution data and running parameters within the data packet are written back to the first intelligent agent's operating environment. The necessary system resources are then reallocated to the first intelligent agent, and the service execution process of the first intelligent agent is restarted, allowing it to continue running from the interrupted progress node. After the service recovery execution is completed, a service recovery prompt is displayed to the user via the vehicle's infotainment screen and voice announcement, informing the user that the interrupted service has been resumed and the current service progress. If automatic recovery is not performed, a prompt is sent to the user asking whether to resume the service, providing an option to confirm or cancel recovery, and the corresponding operation is executed based on the user's selection.

[0086] In this embodiment, a switching execution process is established for the processing decision that allows interruption. The orderly suspension or termination of the first intelligent agent service is achieved through standardized interruption operations, and the startup of the second intelligent agent service is completed simultaneously. Through standardized step-by-step execution, system anomalies and data loss caused by irregular and abrupt service interruptions are avoided. At the same time, it ensures that high-priority services initiated by users can be responded to and executed quickly, which meets the requirements of security, orderliness and timeliness of service switching in the vehicle scenario.

[0087] S132. If the processing decision is to disallow interruption, add the service request of the second agent to the waiting queue and keep the first agent providing service.

[0088] The waiting queue refers to an ordered data set set up separately for each service domain in the vehicle system to store service requests that have not yet been executed. Service requests in the queue are sorted according to preset rules and are retrieved and executed in order after system resources are released and the current service is completed.

[0089] Specifically, if the decision is to disallow interruption, the service execution permission and resource usage status of the first agent will be maintained continuously, its service execution process will not be interrupted, and its service running status will be continuously monitored to ensure that the service can be executed normally until completion. During this period, no interruption command will be sent to it, and its resource usage priority will not be adjusted to ensure the continuity of the current service execution.

[0090] Specifically, the service requests of the second intelligent agent are added to the waiting queue corresponding to the service domain to which the second intelligent agent belongs, according to the preset sorting rules. At the same time, attribute tags such as identifier, priority value, enqueue time, and request validity period are added to the service requests to complete the enqueue operation.

[0091] Optionally, after the service request is enqueued, the system displays a message to the user that the service request has entered the waiting queue, informing the user of the estimated execution time and the reason for waiting. At the same time, the service status of the second intelligent agent is updated to "waiting to execute" while the service status of the first intelligent agent remains "in execution".

[0092] In this embodiment, for processing decisions that do not allow interruption, a service request handling and current service protection process is established. By adding new service requests to an orderly managed waiting queue while maintaining the continuous execution of the first intelligent agent's service, service execution anomalies caused by irregular interruptions are avoided. Furthermore, the waiting queue ensures that newly initiated service requests are not directly discarded, thus achieving orderly scheduling of multiple service requests. At the same time, it takes into account the service security requirements and proper handling of user needs in the vehicle scenario, enabling the entire multi-agent scheduling system to achieve closed-loop coverage across all scenarios.

[0093] In the above solution, the standardized interruption operation, service switching process, and supporting progress saving and recovery mechanism of S131 enable rapid response of high-priority services and ensure the continuity of interrupted services, avoiding data loss and repeated user operations caused by service interruption. The current service keep-alive mechanism and orderly waiting queue management mode of S132 not only ensure the complete execution of uninterruptible services and driving safety, but also ensure that newly initiated service requests are not arbitrarily discarded, realizing the orderly scheduling of multiple service requests. This solves the problems of disordered service switching, abnormal execution process, and improper handling of new requests in related technologies, taking into account both the functional safety requirements of the vehicle scenario and the user experience.

[0094] In addition, such as Figure 4 As shown, Figure 4 This is a schematic diagram of the structure of a smart agent management device 400 provided in an embodiment of this application. The smart agent management device 400 includes: The acquisition module 401 is used to acquire the lock status information of the first intelligent agent in response to receiving a service request for the second intelligent agent when the first intelligent agent is in a service state. The lock status information is used to characterize the interruption permission attribute of the first intelligent agent. The generation module 402 is used to generate a processing decision for the service request based on the lock status information. The processing decision is used to determine whether the second agent is allowed to interrupt the first agent. Management module 403 is used to manage the second intelligent agent and the first intelligent agent based on processing decisions.

[0095] In the above scheme, by obtaining the lock status information of the first agent, a clear and unified core judgment basis is established for determining whether subsequent services should be interrupted, fundamentally avoiding arbitrary interruption behavior without basis or rules when multiple agents provide concurrent services. Based on this lock status information, a processing decision is generated to determine whether the second agent is allowed to interrupt the first agent. The interruption permission judgment process revolves entirely around the interruption permission attribute of the first agent, ensuring the consistency and predictability of the judgment logic. This solves the service conflict problem caused by disordered multi-agent service scheduling and ambiguous interruption rules in related technologies. Finally, based on the generated processing decision, corresponding management actions are executed on the second agent and the first agent. This not only ensures the service continuity of the currently executing first agent, but also realizes the orderly processing of newly initiated service requests from the second agent. It fully covers the entire process of service scheduling in multi-agent concurrent scenarios, effectively improving the stability of multi-agent operation and the rationality of scheduling.

[0096] In one specific embodiment, the generation module 402 is further configured to: If the lock status information indicates that the first agent is in a forced lock state, a decision is generated to prevent the second agent from interrupting the processing of the first agent. If the lock status information indicates that the first agent is in a non-forced lock status, a processing decision is generated based on the interruption of the permission mapping relationship.

[0097] In one specific embodiment, the generation module 402 is further configured to: If the interruption permission mapping relationship is set to prohibit interruption, generate a processing decision to prohibit the second agent from interrupting the first agent. If the interruption permission mapping relationship is set to allow interruption, generate a processing decision that allows the second agent to interrupt the first agent. If the interruption of the permission mapping relationship is a priority decision, obtain the first priority value corresponding to the first agent and the second priority value corresponding to the second agent. A processing decision is generated based on the first priority value and the second priority value.

[0098] In one specific embodiment, the generation module 402 is further configured to: If the first priority value is less than the second priority value, a decision is generated that allows the second agent to interrupt the first agent's processing decision. If the first priority value is greater than or equal to the second priority value, a decision is generated to prevent the second agent from interrupting the first agent's processing.

[0099] In one specific embodiment, the management module 403 is further configured to: If the decision is to allow interruption, execute the interruption operation, suspend or stop the service of the first agent, and start the service of the second agent; If the decision is to disallow interruption, the service request of the second agent is added to the waiting queue, while the first agent continues to provide service.

[0100] In one specific embodiment, after the interruption operation is performed, the agent management device 400 further includes a storage module: The storage module is used to save the service progress status of the first intelligent agent; After the second agent finishes its service, the first agent's service is restored based on the service progress status, and a restoration prompt is output.

[0101] In one specific embodiment, the management device 400 for the intelligent agent further includes an adjustment module: The adjustment module is used to obtain feedback data from users related to service interruption and disruption operations; Based on feedback data, determine the user habits and service preferences of the target audience; Adjust the lock status information of the intelligent agent based on usage habits and service preferences.

[0102] In one specific embodiment, the management device 400 for the intelligent agent further includes a storage module: The storage module is used to divide multiple intelligent agents into multiple types according to their functional attributes, with each type corresponding to an independent service domain. Within each type, lock status information is pre-stored. The lock status information includes at least hierarchical lock rules and interruption permission mapping relationships. The hierarchical lock rules include the lock status information settings of the first agent and the interruption rule settings when the service request of the second agent is triggered. The interruption permission mapping relationship is the interruptible relationship, non-interruptible relationship and priority mapping between the first agent and other agents.

[0103] Regarding the apparatus in the above embodiments, the specific manner in which each unit performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0104] This embodiment also provides a vehicle, including an on-board controller. The on-board controller includes a memory, a processor, and a computer program stored in the memory. When the processor executes the computer program, it implements the management method of any of the optional intelligent agents described above, thus achieving the same effect as the above implementation method.

[0105] This embodiment also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a management method for any of the optional intelligent agents described above.

[0106] The beneficial effects of the above embodiments can be referred to the beneficial effects of the corresponding methods provided above, and will not be repeated here.

[0107] Through the above description of the embodiments, those skilled in the art will understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0108] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0109] In the description of this application, it should be understood that if the terms "upper", "lower", "front", "rear", "left" and "right" are used to indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the position or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application.

[0110] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0111] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for managing intelligent agents, characterized in that, The method includes: When the first agent is in a service state, in response to receiving a service request for the second agent, the locking status information of the first agent is obtained, and the locking status information is used to characterize the interruption permission attribute of the first agent. Based on the locked state information, a processing decision is generated for the service request, the processing decision being used to determine whether the second agent is allowed to interrupt the first agent; Based on the processing decision, the second intelligent agent and the first intelligent agent are managed.

2. The management method according to claim 1, characterized in that, The locking status information includes the interruption of permission mapping relationships. The step of generating a processing decision for the service request based on the locking status information includes: If the locking status information indicates that the first agent is in a forced locking state, a decision is generated to prevent the second agent from interrupting the processing of the first agent. If the locking status information indicates that the first agent is in a non-forced locking state, the processing decision is generated based on the interruption of the permission mapping relationship.

3. The management method according to claim 2, characterized in that, The process of generating the processing decision based on the interruption permission mapping relationship includes: If the interruption permission mapping relationship is to prohibit interruption, a processing decision is generated to prohibit the second agent from interrupting the first agent. If the interruption permission mapping relationship is set to allow interruption, a processing decision is generated that allows the second agent to interrupt the first agent. If the interruption permission mapping relationship is a priority decision, obtain the first priority value corresponding to the first agent and the second priority value corresponding to the second agent; The processing decision is generated based on the first priority value and the second priority value.

4. The management method according to claim 3, characterized in that, The step of generating the processing decision based on the first priority value and the second priority value includes: If the first priority value is less than the second priority value, a decision is generated that allows the second agent to interrupt the processing of the first agent. If the first priority value is greater than or equal to the second priority value, a decision is generated to prohibit the second agent from interrupting the processing of the first agent.

5. The management method according to claim 1, characterized in that, The management of the second and first intelligent agents based on the processing decision includes: If the processing decision is to allow interruption, execute the interruption operation, suspend or stop the service of the first agent, and start the service of the second agent; If the processing decision is to disallow interruption, the service request of the second agent is added to the waiting queue, while the first agent continues to provide service.

6. The management method according to claim 5, characterized in that, After performing the interruption operation, the method further includes: Save the service progress status of the first intelligent agent; After the second agent finishes its service, the first agent's service is restored based on the service progress status, and a restoration prompt is output.

7. The management method according to claim 1, characterized in that, The method further includes: Obtain feedback data from users regarding service interruption and disruption operations; Based on the feedback data, the user's usage habits and service preferences are determined; Based on the aforementioned usage habits and service preferences, adjust the lock status information of the intelligent agent.

8. The management method according to claim 1, characterized in that, The method further includes: Multiple intelligent agents are divided into multiple types according to their functional attributes, and each type corresponds to an independent service domain; Within each type, the locking status information is pre-stored. The locking status information includes at least hierarchical locking rules and interruption permission mapping relationships. The hierarchical locking rules include the locking status information setting of the first intelligent agent and the interruption rule setting when the service request of the second intelligent agent is triggered. The interruption permission mapping relationship is the interruptible relationship, non-interruptible relationship and priority mapping between the first intelligent agent and other intelligent agents.

9. A vehicle, comprising an on-board controller, the on-board controller including a memory, a processor, and a computer program stored in the memory, characterized in that, When the processor executes the computer program, it implements the management method for intelligent agents as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the management method of the intelligent agent as described in any one of claims 1 to 8.

Images