Intelligent agent interpretability information processing apparatus, method, device, and medium

Abstract

The embodiment of the invention provides an interpretable information processing device, method and equipment of an intelligent agent and a medium, which can be applied to the technical field of artificial intelligence. The interpretability information processing device of the intelligent agent can be applied to an information updating stage, a task planning stage and a task execution stage in a task planning execution process, and comprises an environment intention updating module, a planning reflection module and an information carrier module. The environment intention updating module is used for receiving environment updating information in the information updating stage of the current round so as to update planning intention information; the planning reflection module is used for updating the planning reflection information according to the updating planning intention information in the task planning stage of the current round; the information carrier module is used for recording the planning intention information and the planning reflection information in real time in a preset record identification data form, and generating interpretability information used for being called in the task planning stage of the current round, the task execution stage of the current round and the information updating stage of the next round.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, specifically to the field of image processing technology, and more specifically to an interpretable information processing method, device, equipment, medium and product for an intelligent agent. Background Art

[0002] Artificial Intelligence (AI for short) is an important driving force for the new round of scientific and technological revolution and industrial transformation. It is a new key technical science that studies, develops theories, methods, technologies and application systems for simulating, extending and expanding human intelligence. As an important part of intelligent science, artificial intelligence attempts to understand the essence of intelligence and produce a new intelligent machine (i.e., an intelligent agent) that can respond in a way similar to human intelligence.

[0003] An intelligent agent usually needs to have basic characteristics such as subjectivity, purposefulness, knowability, interpretability, value and realizability. Among them, interpretability means that the decision-making process and results of the intelligent agent can be understood and explained by humans, which is used to reflect the transparency of the intelligent agent to humans, enabling humans to understand the decision-making process and results of the intelligent agent, trace back and explain the decisions and behaviors of the intelligent agent. However, in the prior art, the traceability and explanation of the decisions and behaviors of intelligent agents still belong to the research difficulties in the field. The most important point is that it is impossible to better achieve the interpretability of real-time information of intelligent agents, resulting in problems such as inaccurate, untimely and poor transparency of information decisions in the interpretability of intelligent agents, making it difficult for humans to understand and explain the task execution intentions and feedback content of intelligent agents. Summary of the Invention

[0004] In view of at least one of the existing problems in the above-mentioned prior art, based on the current development status of interpretable artificial intelligence, the embodiments of the present invention aim to provide an interpretable information processing device, method, equipment, medium and product for an intelligent agent that can realize the interpretability of real-time information of the intelligent agent. A multi-stage interpretable information recording and calling scheme for intelligent agents based on a loop mechanism is provided. Through the recording and utilization of interpretable information in each stage of the intelligent agent loop program, the information transparency of the intelligent agent's decision-making behavior to developers is realized, the language interaction experience between the intelligent agent and the user is enhanced, so as to achieve the effect of tracing the origin of the intelligent agent's behavior and mutual trust between humans and machines, and realize the recording and utilization of interpretable information in the intelligent agent's planning and decision-making process, and realize the highly transparent interpretable information of the intelligent agent with high efficiency, accuracy and real-time.

[0005] One aspect of an embodiment of the present invention provides an interpretable information processing device for an agent, which is applied to the information update phase, the task planning phase, and the task execution phase during the execution of task planning. Among them, the device includes an environmental intention update module, a planning reflection module, and an information carrier module.

[0006] The environmental intention update module is used to receive environmental update information in the information update phase of the current round to update the planning intention information; the planning reflection module is used to update the planning reflection information according to the planning intention information in the task planning phase of the current round; the information carrier module is used to record the planning intention information and the planning reflection information in real time in the form of preset record identification data, and generate interpretable information to be called in the task planning phase of the current round, the task execution phase of the current round, and the information update phase of the next round.

[0007] According to an embodiment of the present invention, the environmental intention update module includes a mental cognition unit. The mental cognition unit is used to receive the environmental perception information of the environmental update information in the information update phase of the current round to update the mental cognition information.

[0008] According to an embodiment of the present invention, the environmental intention update module further includes an intention language update unit. The intention language update unit is used to receive the environmental sound information of the environmental update information in the information update phase of the current round to generate a natural language text.

[0009] According to an embodiment of the present invention, the environmental intention update module further includes an intention update unit. The intention update unit is used to generate planning intention information according to the historical feedback information, mental cognition information, natural language text, and its own value status information in the information update phase of the current round.

[0010] According to an embodiment of the present invention, the planning reflection module includes a batch simulation unit and an asynchronous simulation unit. The batch simulation unit is used to perform batch simulation according to the planning intention information to generate batch simulation information for the planning intention information; the asynchronous simulation unit is used to perform asynchronous simulation according to the planning intention information to generate asynchronous simulation information for the planning intention information.

[0011] According to an embodiment of the present invention, the planning reflection module further includes a planning reflection unit. The planning reflection unit is used to update the planning reflection information according to the batch simulation information and the asynchronous simulation information.

[0012] According to an embodiment of the present invention, the device further includes an information invocation module. The information invocation module is used to invoke interpretable information in the task execution phase of the current round for interactive scenario interpretation and self-intention planning.

[0013] Another aspect of the embodiments of the present invention provides a method for processing interpretable information of an agent, which is applied to the information update stage, task planning stage, and task execution stage during task planning and execution. Specifically, it includes: receiving environment update information in the information update stage of the current round to update the planning intention information; updating the planning reflection information according to the updated planning intention information in the task planning stage of the current round; and recording the planning intention information and planning reflection information in real time in the form of preset record identification data to generate interpretable information that can be called in the task planning stage of the current round, the task execution stage of the current round, and the information update stage of the next round.

[0014] Another aspect of the embodiments of the present invention provides an electronic device, including one or more processors and a memory. The memory is used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors are caused to execute the above-mentioned method for processing interpretable information of an agent.

[0015] Another aspect of the embodiments of the present invention provides a computer-readable storage medium, on which executable instructions are stored. When the instructions are executed by a processor, the processor is caused to execute the above-mentioned method for processing interpretable information of an agent.

[0016] Another aspect of the embodiments of the present invention provides a computer program product, including a computer program. When the computer program is executed by a processor, the above-mentioned method for processing interpretable information of an agent is implemented.

[0017] The interpretable information processing device of the agent provided by the embodiments of the present invention can at least partially solve the problem in the related art that it is impossible to better achieve the interpretability of the real-time information of the agent during the task planning and execution process of the agent. Therefore, it can at least achieve one of the following technical effects: (1) By establishing an interpretable information recording scheme for multiple stages of the agent based on a loop mechanism, through the recording of interpretable information in multiple stages in the internal program loop of the agent, the transparency of system decision-making information is realized, and the real-time traceability of the behavior information of the agent by humans can be achieved, improving the user's language communication experience.

[0018] (2) Through the recording of interpretable information generated based on its own value and its own state, the recording and utilization of interpretable information in the language interaction process based on its own value and its own state can be realized.

[0019] (3) The multi-round loop mechanism records historical events and related interpretable information. It can not only explain the information at the current moment, but also comprehensively explain this moment based on historical information. Therefore, the agent can comprehensively utilize the interpretable information of historical conversations and historical events to improve its language interaction ability.

[0020] In summary, the interpretable information processing device of the intelligent agent according to the embodiments of the present invention can scientifically record interpretable information related to each stage of the operation of the intelligent agent based on the loop mechanism in a real environment or a virtual environment, and give full play to the interpretability of the system. Secondly, it can enable the intelligent agent to record the relevant interpretable information obtained by reasoning based on the large model through its own value and state. Moreover, in the multi-round interaction between the intelligent agent and the environment, it can make full use of historical information to improve the current interpretable language communication, enhance the language interaction ability of the intelligent agent, and improve the user communication experience.

[0021] Therefore, the interpretable information processing device of the intelligent agent according to the embodiments of the present invention can support the interpretability for natural language. Based on the current development status of interpretable artificial intelligence, it provides a scheme for recording and utilizing interpretable information in the intelligent agent planning and decision-making process. Through the recording and utilization of interpretable information in each stage of the intelligent agent loop program, it realizes the information transparency of the intelligent agent's decision-making behavior to developers, enhances the language interaction experience between the intelligent agent and users, and thus achieves the effect of tracing the origin of the intelligent agent's behavior and mutual trust between humans and machines.

[0022] It should be understood that the above general description and the following specific embodiments are only exemplary and explanatory, and cannot limit the scope claimed by the present invention. Brief Description of the Drawings

[0023] Through the following description of the embodiments of the present invention with reference to the drawings, the above content and other objects, features and advantages of the present invention will become clearer. In the drawings: Figure 1 Schematically shows a structural block diagram of an interpretable information processing device of an intelligent agent according to an embodiment of the present invention; Figure 2 Schematically shows an application scenario diagram of an interpretable information processing device of an intelligent agent according to an embodiment of the present invention in the process of task planning and execution based on a loop mechanism; Figure 3A Schematically shows an interaction scenario diagram of each module of an interpretable information processing device of an intelligent agent according to an embodiment of the present invention in the process of task planning and execution based on a loop mechanism; Figure 3B Schematically shows another application scenario diagram of an interpretable information processing device of an intelligent agent according to an embodiment of the present invention in the process of task planning and execution based on a loop mechanism; Figure 4 Schematically shows an application scenario diagram of an interpretable information processing device, method, device, medium and program product of an intelligent agent according to an embodiment of the present invention; Figure 5The flowchart of the method for processing interpretable information of an agent according to an embodiment of the present invention is schematically shown; and Figure 6 The block diagram of an electronic device suitable for implementing the method for processing interpretable information of an agent according to an embodiment of the present invention is schematically shown.

[0024] The above-mentioned drawings are a part of the specification of the embodiments of the present invention, which illustrate the exemplary embodiments of the present invention. The attached drawings and the description of the specification are used together to explain the principles of the embodiments of the present invention. It should be understood that the above general description of the drawings and the following specific embodiments are only exemplary and explanatory, and they do not limit the scope that the present invention intends to claim. Detailed Embodiments

[0025] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer and more understandable, the spirit of what the present invention discloses will be clearly described below with reference to the drawings and in detail. After any person skilled in the art understands the embodiments of the content of the present invention, the techniques taught by the content of the present invention can be changed and modified, which does not deviate from the spirit and scope of the content of the present invention.

[0026] The schematic embodiments of the present invention and their descriptions are used to explain the present invention, but do not limit the present invention. In addition, the elements / components using the same or similar reference numerals in the drawings and embodiments are used to represent the same or similar parts.

[0027] Regarding the "first", "second",... etc. used in the present invention, they do not particularly refer to the meaning of order or sequence, nor are they used to limit the present invention. They are only used to distinguish elements or operations described with the same technical terms.

[0028] Regarding the directional terms used in the present invention, such as: up, down, left, right, front or back, etc., they are only references to the directions in the drawings. Therefore, the directional terms used are for explanation and not for limiting this creation.

[0029] Regarding the "comprising", "including", "having", "containing", etc. used in the present invention, they are all open-ended terms, that is, they mean including but not limited to.

[0030] Regarding the "and / or" used in the present invention, it includes any one or all combinations of the described things.

[0031] Regarding the "plurality" in the present invention, it includes "two" and "more than two"; regarding the "multiple groups" in the present invention, it includes "two groups" and "more than two groups".

[0032] Regarding terms such as "substantially" and "about" used in the present invention, they are used to modify any quantity or error that can vary slightly, but these slight variations or errors do not change its essence. Generally speaking, the range of such slight variations or errors modified by such terms can be 20% in some embodiments, 10% in some embodiments, 5% or other values in some embodiments. Those skilled in the art should understand that the aforementioned values can be adjusted according to actual needs and are not limited thereto.

[0033] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0034] In cases where expressions similar to "at least one of A, B, and C, etc." are used, generally, it should be interpreted according to the meaning commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include, but not be limited to, a system having only A, only B, only C, having A and B, having A and C, having B and C, and / or having A, B, and C, etc.). In cases where expressions similar to "at least one of A, B, or C, etc." are used, generally, it should be interpreted according to the meaning commonly understood by those skilled in the art (for example, "a system having at least one of A, B, or C" should include, but not be limited to, a system having only A, only B, only C, having A and B, having A and C, having B and C, and / or having A, B, and C, etc.). Those skilled in the art should also understand that substantially any disjunctive conjunction and / or phrase representing two or more alternative items, whether in the specification, claims, or drawings, should be understood as giving the possibility of including one of these items, either of these items, or both items. For example, the phrase "A or B" should be understood as including the possibility of "A" or "B", or "A and B".

[0035] In the existing intelligent agent natural language interaction scenario, after receiving the user's input of QA (question and answer) or other instructions, it will reply "received" and execute the instructions. However, if the instruction fails to execute, it will directly reply with a relatively general response such as "I can't do it". Among them, when there is a network failure, certain skills U do not match the preset conditions (such as precondition), or other errors occur during the task execution process, the current technology will not actively inform the user of the specific reasons. This will bring a bad experience to user interaction and also make the user feel too mechanical and unable to reflect the interpretability ability of general artificial intelligence.

[0036] In view of at least one of the existing problems in the above-mentioned prior art, based on the current development status of interpretable artificial intelligence, embodiments of the present invention aim to provide an interpretable information processing device, method, device, medium and product for an intelligent agent that can realize the interpretable information of the intelligent agent in real-time information, and provide a multi-stage interpretable information recording and calling scheme for the intelligent agent based on a loop mechanism. By recording and utilizing the interpretable information in each stage of the intelligent agent loop program, the information transparency of the intelligent agent's decision-making behavior to developers is realized, the language interaction experience between the intelligent agent and the user is enhanced, so as to achieve the effect of tracing the origin of the intelligent agent's behavior and mutual trust between humans and machines, and realize the recording and utilization of interpretable information in the intelligent agent's planning and decision-making process, and achieve high transparency of intelligent agent information that is efficient, accurate and real-time.

[0037] Figure 1 Schematically shows a structural block diagram of an interpretable information processing device 100 for an intelligent agent according to an embodiment of the present invention. The following will be combined with Figures 1 - 3B to make a further detailed description of the interpretable information processing device 100.

[0038] As Figure 1 shown, one aspect of the embodiments of the present invention provides an interpretable information processing device 100 for an intelligent agent, which is applied to the information update stage A, the task planning stage B, and the task execution stage C during the task planning and execution process. As Figure 1 shown, the interpretable information processing device 100 for the intelligent agent in this embodiment includes an environmental intention update module 110, a planning reflection module 120, and an information carrier module 130.

[0039] The environmental intention update module 110 is used to receive environmental update information in the current round of the information update stage to update the planning intention information; The planning reflection module 120 is used to update the planning reflection information according to the planning intention information in the current round of the task planning stage; The information carrier module 130 is used to record the planning intention information and the planning reflection information in real time in the form of preset record identification data, and generate interpretable information to be called in the current round of the task planning stage, the current round of the task execution stage, and the next round of the information update stage.

[0040] In the embodiments of the present invention, as Figure 1The interpretable information processing device 100 shown can be widely applied to the intelligent agent system architecture of multi-step loop programs, including the intelligent agent system applied to the simulation environment and the physical intelligent robot in the real environment. Whether in the simulation environment or the real environment, the intelligent agent can perform task planning and execution, which can be specifically reflected in the repeated cycle process of "information update - task planning - task execution" of the multi-round loop mechanism, enabling the intelligent agent to complete the expected action tasks, such as chatting with users, fetching water and cooking for users, etc.

[0041] As Figure 2 Shown is an application scenario diagram of the interpretable information processing device 100. The intelligent agent task planning and execution in the embodiments of the present invention can be understood as the process in which the intelligent agent plans and subsequently executes tasks based on the received and detected environmental interaction information. For example, according to the user's command (the received environmental interaction information), task generation (task planning) is performed, and operation execution (task execution) is performed according to the generated task. In this process, the intelligent agent will repeatedly perform the above task planning and execution process according to the above multi-round loop mechanism. Each round of execution will completely implement the three steps of "information update - task planning - task execution", and in the next round of execution process, a new round of task planning and execution process will be performed according to the task execution of the previous round.

[0042] Specifically, as Figure 2 Shown, based on the above multi-round loop mechanism, the intelligent agent can repeatedly perform the process of "information update - task planning phase - task execution", and thus can be divided into three main phases of "information update phase A (UpdateInfostate), task planning phase B (Plan), and task execution phase C (Execute)", which can be sequentially implemented depending on the update module, planning module, and execution module of the intelligent agent.

[0043] Specifically, as Figure 2 Shown, in the task execution phase C of the previous round, through the operation execution of the intelligent agent's execution action on the environment E (Environment), the intelligent agent can update the current round of environmental information according to this environmental interaction process and enter the information update phase A; in the current round of this information update phase A, the planning intention information of the intelligent agent (i.e., the execution intention Intent of the next round) is generated through the updated environmental information and enters the task planning phase B; in the current round of task planning phase B, the execution action is planned through the execution intention of the intelligent agent of the above planning intention information, and the optimal execution trajectory (traj_list) is generated and enters the task execution phase C; in the current round of task execution phase C, the corresponding operation action (Action) is executed on the environmental target according to the above optimal execution trajectory, thus completing the entire process of the current round of task planning and execution.

[0044] Furthermore, the Update infostate corresponds to the information update stage A, which is mainly responsible for reasoning based on the environmental perception information (belief_pg), and is used to complete the external environment interaction information such as vision, language, and sound and the corresponding cognitive reasoning information in the interpretable information processing device 100 (infostate). For example, it may include the inference information of the agent for unobserved information, the inference information for the future state of things, and the mental state information of the agent updated based on the inference.

[0045] The plan module corresponds to the task planning stage B. By using the plan based on the UV architecture (Skill U - Evaluation Value), the input information that the plan module can call may include the environmental perception state information (belief state) with the current cognition and focus of the agent; in addition, it also includes the U library describing the agent's skill library (such as the strategy of turning the Rubik's Cube and similar skill information such as the possible results caused by different strategies); further, it also includes the Action library that is called by different U libraries and constitutes a link in the U library strategy; finally, it may also include the V library representing the value evaluation of the agent (i.e., the Value library). Among them, the plan module can execute the selection of the skill library U and the relevant parameters of U, and the final output decision (i.e.,) is the execution path of the skill library U (such as traj_list). Therefore, in the current round of the task planning stage, the decision of the plan module can be different skills U selected based on simulation.

[0046] The execute module corresponds to the task execution stage C, which is mainly responsible for the execution of the above - mentioned planning decision. During its execution process, it can achieve real - time update of the relevant environmental perception information (corresponding to the next - round information update stage). At the same time, during the execution process, it can perform subtasks such as learning, reflection, and decision adjustment according to the latest information update content. That is, when the execute module controls the agent to output task - related actions for the environment, it can simultaneously update its own state and the model itself.

[0047] Such as Figures 1 - 3AAs shown, the environmental intention update module 110 can be equivalent to a component of the aforementioned update module. In the information update stage A, the environmental update information can be the interaction information generated during the current round of interaction between the agent and the environment. For example, the environmental visual information detected by the agent through various sensors during "seeing", "hearing", "touching" and other interaction actions (such as people, spaces, objects, sizes, shapes, colors, etc. in the room), environmental sound information (such as voiceprints, voices, noises, etc.), and environmental cognitive information (such as identifying who a person is, what a person is doing, what a person may do, etc.). In addition, the environmental update information can also include the environmental interaction information passively received by the agent, such as the instruction information input by the user through the interactive display touch interface of the agent.

[0048] The planning intention information can be the execution intention corresponding to the current round of information update stage generated by the agent through reasoning and cognition based on the above environmental update information. This execution intention can determine the task objective of the agent in the task execution stage of the current round. Since there was relevant historical planning intention information in the previous round of information update stage, in the current round of information update stage A, it is a further update of the historical planning intention information, such as adding a new execution intention and abandoning the historical execution intention on this basis, or modifying the historical execution intention to form a new execution intention.

[0049] As Figures 1 - 3A shown, the planning reflection module 120 can be understood as a component of the aforementioned planning module (plan). In the task planning stage B, the planning reflection information can usually be the execution reflection information generated by the agent through simulated execution according to the execution intention determined by the planning intention information. Specifically, it is to call the execution operator to perform execution simulation on each formed execution path (or called execution trajectory) in a virtual simulation environment, and select the optimal execution path according to the evaluation (value) of the simulation results generated by the execution simulation. The process of execution simulation and path evaluation can be understood as the "reflection" based on the execution intention, and the optimal execution path and related content output can be understood as the main components of the planning reflection information.

[0050] As Figures 1 - 3A shown, the information carrier module 130 (context) can be equivalent to the carrier for information storage and call management of each module unit such as the above environmental intention update module 110 and planning reflection module 120. It is the main information carrier in the loop stage of the above agent, and can be used to record almost all effective information of each link, each stage of the agent and the feedback from the external environment, including the interpretability information mentioned in the embodiments of the present invention.

[0051] The preset record identification data form can be the specified data format (such as the record identification) in which the above-mentioned planning intention information and planning reflection information are recorded in the information carrier module 130. It usually can have a certain record identification to separately record different types of information, so as to realize the classified recording and management (such as calling) of the agent information in different stages. Taking the currently applied agent framework process as an example, as Figure 3A shown, the information carrier module 130 (context) can be the main information management module in the interpretability information processing device 100 (infostate) responsible for recording and circularly transmitting historical information. Different interpretability information in the information carrier module 130 can be various information in the form of preset record identification data with similar record identifications stored by the agent in different loop stages. The interpretable information of the agent in each stage of the loop system mainly can cover three categories, including inappropriate operator logic (such as lfp_warm up), non-compliance of action preconditions (such as error code), and non-compliance of value orientation (such as low evaluation resulting in value conflict and corresponding action abandonment). The recording and calling of the interpretable information can specifically occur in the entire process loop such as the information update stage A, the task planning stage B, and the task execution stage.

[0052] In summary, in the interpretable information processing device 100 of the agent in the embodiment of the present invention, through the constructed multi-stage interpretable information recording scheme for the agent based on the loop mechanism, through the information carrier module 130, multi-stage interpretable information recording and calling can be realized in the internal program loop of the agent, so as to ensure that the system decision-making information is more transparent, enabling humans to trace the behavior information of the agent, thus significantly improving the user's language communication experience.

[0053] Moreover, recording historical events and related interpretable information based on the multi-round loop mechanism can not only explain the information at the current moment, but also comprehensively explain this moment based on historical information, thus ensuring the real-time and accuracy of the interpretable information at the same time, achieving a more transparent interpretable ability of the agent in the true sense, and thus achieving the effect of tracing the origin of the agent's behavior and mutual trust between humans and machines.

[0054] As Figures 1 - 3B shown, according to an embodiment of the present invention, the environmental intention update module 110 includes a mental cognition unit 311.

[0055] The mental cognition unit 311 is used to receive the environmental perception information of the environmental update information in the information update stage A of the current round to update the mental cognition information.

[0056] As Figure 3AAs shown, the mental cognitive unit 311 (believe step) corresponding to the information update phase A can process environmental perception information such as visual perception and auditory perception in the environmental update information transmitted by the external environment to the agent. This environmental perception information can also include information related to whether the task goal was successfully achieved in historical behaviors (such as the previous action execution process). Subsequently, it updates its mental cognitive information according to the relevant environmental update information (such as env_info_update) received from the outside world.

[0057] Among them, the mental cognitive information can include the external environment interaction information and corresponding cognitive reasoning information in the environmental update information. For example, it can include the agent's inference information about unobserved information, inference information about the future state of things, and the mental state information of the agent updated based on the inference.

[0058] In the mental cognitive unit 311 corresponding to this information update phase, all feedback from the outside world on the agent's previous decision-making behaviors can be recorded. For example, the generated error code can be recorded as mental cognitive information in the information carrier module 130 based on the preset record identification data form of evt_record. Specifically, for example, when the agent attempts the target task of unscrewing the bottle cap and finds that the bottle cap is too tight to unscrew during the execution of the action of unscrewing the bottle cap, the information generated during this execution process is then passed into the agent as environmental update information, and its environmental perception information (such as the bottle cap is too tight to unscrew) is also recorded as mental cognitive information in the information carrier module 130 (Context).

[0059] As Figures 1 - 3B shown, according to an embodiment of the present invention, the environmental intention update module 110 further includes an intention language update unit 312.

[0060] The intention language update unit 312 is used to receive the environmental sound information of the environmental update information in the current round of information update phase A to generate a natural language text.

[0061] As Figure 3AAs shown, the intention language update unit 312 (mind_nlp_update) corresponding to the information update phase A can extract the environmental update information during the agent interaction process in the external environment E to generate environmental sound information, where the environmental sound information can be environmental sound data (such as dialogue voice, voiceprint, etc.) during the agent-environment interaction process. The intention language update unit 312 can also selectively convert the above environmental sound information to generate natural language text, such as the dialogue text between humans and agents. The natural language text can be text information (current nlp message) that is more in line with human language expression generated after natural language processing (nature language processing). Among them, the natural language text information can be recorded in the information carrier module 130 in the form of a preset record identification data based on nlp_record. Therefore, the natural language text information can be used to verify and track the interaction information of the language communication between agents and humans, and between agents and agents.

[0062] As Figures 1 - 3B shown, according to an embodiment of the present invention, the environmental intention update module 110 further includes an intention update unit 313.

[0063] The intention update unit 313 is used to generate a planning intention information in the current round of information update phase A according to the historical feedback information, mental cognitive information, natural language text, and its own value status information.

[0064] As Figure 3AAs shown in the figure, the intent update unit 313 corresponding to the information update phase A is the last processing unit corresponding to this phase. The intent update unit 313 can comprehensively infer and judge the comprehensive information based on all the previous existing information to generate the planned intent information corresponding to the execution intent. Among them, these existing information can include the execution feedback information generated by the execution of the historical target task in the previous round of the loop (i.e., historical feedback information), can also include the environmental update information fed back from the outside world mentioned above, can also include the mental cognitive information updated by the mental cognitive unit 311, the natural language text updated by the intent language update unit 312, and the self-value state information. Among them, the self-value state information can include the value information generated by the evaluation of the agent's own behavior and the state information related to the agent's own behavior state in the current round, etc., and can mainly be used to trigger a new intent. For example, when the agent is executing the target task of moving from the living room to the bedroom to help the user get an item, and there is a stool obstacle in the aisle, making the target task unable to be completed according to the original actions, at this time, the agent can re-plan the task according to its own state (such as the distance from the obstacle) and the evaluation of its own behavior (whether the current behavior of itself causes the body to not pass smoothly, etc.), and obtain a new target subtask of bypassing or removing the obstacle in the current round. At this time, it also corresponds to the generation process of the execution intent. The planned intent information generated by the above historical feedback information, mental cognitive information, natural language text, and self-value state information can be recorded in the information carrier module 130 in the form of preset record identification data of int_record and evt_record respectively.

[0065] Therefore, the above planned intent information can include two types of interpretable information, namely intent information (Intent) and value information (Value). Intent information represents the agent's inner thoughts, motives, etc. For example, when receiving external dialogue information and having a relatively high degree of friendliness with the speaker, the intent to speak can be generated. Value information represents the reason for choosing not to execute the target task when there is a conflict with the agent's own value or state. For example, when the agent is commanded to break a fish tank, since this behavior conflicts with the existing value of protecting the user's property or even loving animals, the agent can choose not to execute this task. The two types of records in this stage explain the agent's thinking and inner thoughts in specific situations, facilitating humans to observe the agent's inner world.

[0066] Therefore, the interpretable information record of the planning intention information generated based on its own value and its own state can realize the interpretable information record and utilization based on the language interaction process of its own value and its own state, making... Moreover, in this process, historical events and other related interpretable information can be recorded based on the multi-round loop mechanism, which can not only explain the information at the current moment, but also comprehensively interpret this moment based on historical information, giving full play to the interpretability of the system and realizing the high transparency of the intelligent agent information with high efficiency, accuracy and real-time performance.

[0067] As Figures 1 - 3B shown, according to an embodiment of the present invention, the planning reflection module 120 includes a batch simulation unit 321 and an asynchronous simulation unit 322.

[0068] The batch simulation unit 321 is used to perform batch simulation according to the planning intention information to generate batch simulation information for the planning intention information; The asynchronous simulation unit 322 is used to perform asynchronous simulation according to the planning intention information to generate asynchronous simulation information for the planning intention information.

[0069] As Figure 3A shown, the batch simulation unit 321 (batch simulate) can perform batch planning simulation on the above-mentioned planning intention information in the current round of loop process, and can simultaneously perform batch simulation processing on multiple planning execution paths corresponding to the above-mentioned planning intention information. Specifically, for each execution path, the corresponding execution operator can be called to perform simulation execution in the simulation environment to obtain the corresponding simulation results, that is, batch simulation information.

[0070] Different from the above batch simulation, the asynchronous simulation unit 322 (asynchronous simulation unit 322) can perform independent asynchronous planning simulation on the above-mentioned planning intention information. This asynchronous simulation process is not restricted by the above intelligent agent loop process and is independent of this loop process without time conflict with this loop process. In this asynchronous simulation process, the corresponding execution operator can also be called to perform simulation execution on each execution path in the simulation environment to obtain the corresponding simulation results, that is, asynchronous simulation information.

[0071] In the corresponding task planning stage B, batch simulation and asynchronous simulation can start simultaneously. However, when the batch simulation completes the simulation of the current round in this stage, the asynchronous simulation may still be doing the simulation of the previous round, that is, the asynchronous simulation will not be restricted by the current round of task planning stage B and can be independent of this task planning stage B. When the batch simulation of a certain round is completed, the asynchronous simulation may also be completed at the same time. At this time, the corresponding batch simulation information and asynchronous simulation information can be generated simultaneously.

[0072] Specifically, when the user assigns a target task of "fetching water" to the agent, the intention update unit can provide the planning intention information (Intent) corresponding to "fetching water". At this time, in the task planning stage, the batch simulation unit 321 can perform action execution path planning based on this planning intention information, and can simultaneously generate multiple action execution path plans (such as the complete execution path of "finding the water dispenser and the cup - walking towards the cup - picking up the cup - turning around - walking towards the water dispenser - placing the cup at the water intake - turning on the water intake switch - waiting for the cup to be full - turning off the water intake switch - picking up the water cup - turning around - walking towards the user"), and perform batch simulation execution on these execution paths to obtain the batch simulation information corresponding to this stage. For the execution intention of specific non-timely planning intention information, the asynchronous simulation unit 322 does not need to consider the execution simulation stage of the batch simulation unit 321, and is completely independent of the "update - plan - execute" cycle of the agent's loop mechanism, and can perform parallel asynchronous planning and path simulation. For example, when in the above "fetching water" task, after the agent "walks towards the water dispenser - places the cup at the water intake - turns on the water intake switch", it is found that the water dispenser is not plugged in. At this time, the agent will generate a new task of "plugging in the water dispenser". At this time, this "plugging in the water dispenser" task can also correspond to a new execution path, such as the path of "finding the water dispenser wiring plug and the nearest socket - turning around - walking towards the plug - picking up the plug - turning around - walking towards the socket - inserting the plug into the socket". At this time, the path simulation of this new "plugging in the water dispenser" task can be understood as an asynchronous simulation relative to the above "fetching water" task, and then obtain the corresponding asynchronous simulation information.

[0073] As Figures 1 - 3B shown, according to an embodiment of the present invention, the planning reflection module 120 further includes a planning reflection unit 323.

[0074] The planning reflection unit 323 is used to update the planning reflection information according to the batch simulation information and the asynchronous simulation information.

[0075] As Figure 3A shown, corresponding to the task planning stage B, the planning reflection unit 323 (reflect) can perform information filtering and reflection after the execution simulation of the planning stage (plan) according to the above batch simulation information and asynchronous simulation information.

[0076] Among them, since multiple action execution paths of batch simulation and asynchronous simulation can be obtained during the aforementioned batch simulation and asynchronous simulation processes as decision-making trajectories for the subsequent execution phase, each trajectory can include a series of subsequent execution actions and action parameters, and these trajectories can be evaluated by the planning and reflection unit 323. Specifically, the simulation execution results of each trajectory can be evaluated separately. For example, evaluation can be performed based on the expected energy consumption value of the trajectory execution, or a comprehensive score can be given according to other preset criteria to form an evaluation result.

[0077] Based on the evaluation results, the planning and reflection unit 323 can also perform pruning and filtering of the decision-making trajectories (removing trajectories with the smallest evaluation value and those with value conflicts), and finally select the optimal decision-making trajectory for execution according to the score ranking. That is, among the multiple simulated decision-making trajectories, unreasonable trajectories will be eliminated due to their low scores. At the same time, interpretable information (such as lfp_warm up, Plan, etc.) is extracted from these simulated unreasonable and reasonable trajectories, and is recorded in the form of preset record identification data (such as evt_record) through the information carrier module 130, thereby greatly improving the interpretability of the agent's "not executing" or "executing" certain decisions. For example, when the user commands the agent to execute the task of "turning on the TV", if the TV in the environment has already been turned on, then the score of the trajectory related to simulating turning on the TV is relatively low, and finally the agent may choose not to execute any task or execute other tasks according to its own value status. In this case, the reason for not executing the command to turn on the TV is that the TV has already been turned on, and this reason and the execution operation can be recorded as interpretable information in the information carrier module 130.

[0078] Therefore, through the interpretable information recording scheme in multiple stages of the internal program loop of the agent, the transparency of system decision-making information can be effectively achieved, enabling humans to trace the behavior information of the agent, and greatly improving the user's language communication experience.

[0079] Among them, the interpretable function of the interpretable information processing device can be widely applied to, for example, Figure 2 the intelligent agent loop processing system framework as shown, and overall, the recording and utilization of interpretable information can be realized.

[0080] Thereby, the above-mentioned interpretable information processing device of the embodiment of the present invention can provide a technology that enables the agent to comprehensively utilize the interpretable information of historical conversations and historical events to improve the language interaction ability. By using a multi-round loop mechanism to record historical events and related interpretable information, it can not only explain the information at the current moment, but also effectively integrate historical information to explain this moment.

[0081] As shown Figures 1 - 3B In accordance with an embodiment of the present invention, the interpretability information processing apparatus 100 further includes an information invocation module 140.

[0082] The information invocation module 140 is configured to invoke interpretability information during the task execution phase C of the current round for interactive scenario interpretation and self-intention planning.

[0083] As shown Figure 3B In the embodiment of the present invention, the interpretability information recorded by the agent based on the cyclic mechanism architecture in each stage of "information update - task planning - task execution" can be utilized in two aspects: the first aspect is for scenario interpretation during the language communication between the agent and humans (i.e., external interpretation, which can be understood as interactive scenario interpretation); the second is for reference to historical information when the agent makes self-intention inference to generate planning intention information (i.e., internal generation of intention, which can be understood as self-intention planning). The former interactive scenario interpretation is for the external environment E and can explain the reasons for the decisions made by the agent; the latter self-intention planning is internal and can assist the agent in analyzing the historical process to generate new intentions.

[0084] The intention generated by the agent is used as an inner drive to simulate humans after receiving external feedback. It is an innovation point that the agent under this architecture can process relevant tasks through its own value state information. Due to the cyclic mechanism of the agent system architecture, the information carrier module 130 (context) can record the interpretability-related information in each stage of multiple rounds of cycles, enabling the agent to combine its own value and state during the information update stage A (such as intent_update), and using various types of information such as historical behaviors, conversations, and actions as inputs to a preset language model (such as Large Language Model, abbreviated as LLM) to generate different intentions. Among them, the system architecture for recording interpretability information in multiple rounds of cycles based on the above-mentioned information carrier module 130 can ensure the integrity of historical information, and at the same time can also ensure the rationality and comprehensiveness of the agent's interpretability information, thus playing a key role in generating reasonable agent intentions.

[0085] For the instructions given by the user to the agent, the agent can choose whether to execute them according to its own value and state. When the agent chooses not to execute the instructions, the decision-making system can use the collected interpretability information through the information invocation module 140 to make an external explanation. The system architecture for recording interpretability information in multiple rounds of cycles can comprehensively display the true inner monologue of the agent with historical information, and give a final summary answer that conforms to human expression habits with the help of a preset language model, so as to more clearly show the true decision-making reasons inside the agent system to the user and improve the language interaction experience effect.

[0086] Therefore, whether in a real environment or a virtual environment, at each stage of operation, the intelligent agent based on the loop mechanism can give full play to the interpretability ability of the system by scientifically recording interpretability-related information. Moreover, the intelligent agent can, based on its own value and state, infer relevant interpretability information records based on a preset language model, enabling the intelligent agent to make full use of historical information in multiple rounds of interaction with the environment to improve the current interpretability language communication process, significantly enhancing the intelligent agent's language interaction ability and improving the user's communication experience.

[0087] In summary, the interpretability information processing device for the intelligent agent provided by the embodiments of the present invention at least achieves one of the following technical effects: (1) By establishing a multi-stage interpretability information recording scheme for the intelligent agent based on the loop mechanism, through the recording of interpretability information in multiple stages in the internal program loop of the intelligent agent, the transparency of the system decision-making information is realized, enabling humans to trace the behavior information of the intelligent agent in real time and enhancing the user's language communication experience.

[0088] (2) Through the recording of interpretability information generated based on its own value and its own state, it is possible to record and utilize interpretability information in the language interaction process based on its own value and its own state.

[0089] (3) The multi-round loop mechanism records historical events and related interpretability information, which can not only explain the information at the current moment but also comprehensively explain this moment based on historical information. Therefore, the intelligent agent can comprehensively utilize the interpretability information of historical conversations and historical events to enhance its language interaction ability.

[0090] In summary, the above-mentioned interpretability information processing device for the intelligent agent in the embodiments of the present invention can, in a real environment or a virtual environment, scientifically record interpretability-related information at each stage of the operation of the intelligent agent based on the loop mechanism, giving full play to the system interpretability. Secondly, it enables the intelligent agent to obtain relevant interpretability information records through inference based on a large model according to its own value and state. Moreover, in multiple rounds of interaction between the intelligent agent and the environment, it can make full use of historical information to improve the current interpretability language communication, enhancing the intelligent agent's language interaction ability and improving the user's communication experience.

[0091] Therefore, the above-mentioned interpretability information processing device for the intelligent agent in the embodiments of the present invention can support interpretability for natural language. Based on the current development status of current interpretable artificial intelligence, it provides a scheme for recording and utilizing interpretability information in the intelligent agent's planning and decision-making process. Through the recording and utilization of interpretability information at each stage in the intelligent agent's loop program, the transparency of the intelligent agent's decision-making behavior to developers is realized, enhancing the language interaction experience between the intelligent agent and users, thereby achieving the effects of tracing the origin of the intelligent agent's behavior and mutual trust between humans and machines.

[0092] According to an embodiment of the present invention, any multiple of the environmental intention update module 110, the planning reflection module 120, the information carrier module 130, and the information invocation module 140 may be combined and implemented in one module, or any one of them may be split into multiple modules. Alternatively, at least part of the functions of one or more of these modules may be combined with at least part of the functions of other modules and implemented in one module. According to an embodiment of the present invention, at least one of the environmental intention update module 110, the planning reflection module 120, the information carrier module 130, and the information invocation module 140 may be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on chip, a system on substrate, a system on package, an application specific integrated circuit (ASIC), or any other reasonable way of integrating or packaging circuits, etc., implemented by hardware or firmware, or implemented in any one of the three implementation manners of software, hardware, and firmware, or in an appropriate combination of any several of them. Alternatively, at least one of the environmental intention update module 110, the planning reflection module 120, the information carrier module 130, and the information invocation module 140 may be at least partially implemented as a computer program module, and when the computer program module is run, the corresponding functions may be executed.

[0093] Based on the above-mentioned interpretable information processing device of the intelligent agent, the present invention also provides an interpretable information processing method for the intelligent agent. The following will be combined with Figures 4 - 5 to describe this method in detail.

[0094] One aspect of an embodiment of the present invention provides an interpretable information processing method for an intelligent agent, which is applied to the information update stage, the task planning stage, and the task execution stage during the task planning and execution process. Wherein, the method includes: receiving environmental update information in the information update stage of the current round to update the planning intention information; updating the planning reflection information according to the updated planning intention information in the task planning stage of the current round; and recording the planning intention information and the planning reflection information in real time in the form of preset record identification data to generate interpretable information for being invoked in the task planning stage of the current round, the task execution stage of the current round, and the information update stage of the next round.

[0095] Figure 4 Schematically shows an application scenario diagram of the interpretable information processing device, method, device, medium, and program product of the intelligent agent according to an embodiment of the present invention.

[0096] As Figure 4As shown, the application scenario 400 according to this embodiment may include terminal devices 401, 402, 403, a network 404, and a server 405. The network 404 is used to provide a medium for communication links between the terminal devices 401, 402, 403 and the server 405. The network 404 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0097] Users can use the terminal devices 401, 402, 403 to interact with the server 405 through the network 404 to receive or send messages, etc. Various communication client applications may be installed on the terminal devices 401, 402, 403, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (only for examples).

[0098] The terminal devices 401, 402, 403 may be various electronic devices with a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop portable computers, and desktop computers, etc.

[0099] The server 405 may be a server providing various services, such as a background management server that supports the websites browsed by users using the terminal devices 401, 402, 403 (only for examples). The background management server may analyze and process data such as user requests received, and feedback the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal devices.

[0100] It should be noted that the method for processing the interpretability information of the agent provided by the embodiments of the present invention can generally be executed by the server 405. Correspondingly, the device for processing the interpretability information of the agent provided by the embodiments of the present invention can generally be set in the server 405. The method for processing the interpretability information of the agent provided by the embodiments of the present invention can also be executed by a server or a server cluster different from the server 405 and capable of communicating with the terminal devices 401, 402, 403 and / or the server 405. Correspondingly, the device for processing the interpretability information of the agent provided by the embodiments of the present invention can also be set in a server or a server cluster different from the server 405 and capable of communicating with the terminal devices 401, 402, 403 and / or the server 405.

[0101] It should be understood that Figure 4 the numbers of the terminal devices, the network, and the server in

[0102] are merely illustrative. According to the implementation requirements, there may be any number of terminal devices, networks, and servers.

[0102] The following will be based on Figure 4 the scenario described Figure 5A detailed description is given of the method for processing interpretability information of the agent in the disclosed embodiments.

[0103] As Figure 5 shown, one aspect of the embodiments of the present invention provides a method for processing interpretability information of an agent, which is applied to the information update stage, the task planning stage, and the task execution stage in the process of task planning and execution. Among them, it includes operations S501 to S503.

[0104] In operation S501, environmental update information is received in the information update stage of the current round to update the planning intention information. In one embodiment, it can be implemented and executed through the environmental intention update module 110 described above, which will not be elaborated here. In operation S502, the planning reflection information is updated according to the updated planning intention information in the task planning stage of the current round. In one embodiment, it can be implemented and executed through the planning reflection module 120 described above, which will not be elaborated here. And In operation S503, the planning intention information and the planning reflection information are recorded in real time in the form of preset record identification data, and interpretability information is generated for being called in the task planning stage of the current round, the task execution stage of the current round, and the information update stage of the next round. In one embodiment, it can be implemented and executed through the information carrier module 130 described above, which will not be elaborated here.

[0105] The agent can be the execution subject of the above-mentioned method for processing interpretability information of the embodiments of the present invention, or an execution party controlled by the method for processing interpretability information. Specifically, it can be a humanoid intelligent robot or other AI devices, which usually have actuators capable of completing specific action tasks by themselves. For example, a humanoid robot can use a mechanical manipulator to complete the action task of picking up an object, and even perform action tasks such as product packaging on an industrial production line. Among them, these action tasks can be used as the target execution tasks of the agent. To complete these target execution tasks, the agent needs to complete a series of execution actions in sequence.

[0106] The agent can detect the environment it is in and its own status information in real time, and at the same time confirm the progress of the current target execution task. Specifically, it can judge the space environment and its own status in each iteration process. The current execution status can be at least one of the status information of the agent's own execution actions and the status of the environment it is in at the current moment. Among them, the status information of the agent's own execution actions can include the name of the currently executed action and the execution information corresponding to the action name (such as the torque, rotation angle of the wrist, estimated execution time, and the spatial positions of each key point of the fingers, etc.), and the status of the environment the agent is in can include the information of the people and objects around the space where the agent is at the current moment, such as the size, position (distance), and even shape and color of the obstacles on the travel route, or the dynamic change information of the people in the surrounding environment, such as the speed of people's movement and the emotional information on the face, etc. Compared with the historical environment information obtained from the previous environment detection, the environment information obtained from the current environment detection can be used as the aforementioned environment update information for relevant operations in the information update stage.

[0107] Figure 6 FIG. schematically shows a block diagram of an electronic device suitable for implementing the interpretable information processing method of an agent according to an embodiment of the present invention.

[0108] The above-mentioned electronic device provided by the embodiment of the present invention includes one or more processors and a memory. The memory is used to store one or more programs. Among them, when the one or more programs are executed by the one or more processors, the one or more processors are caused to execute the above-mentioned interpretable information processing method of the agent.

[0109] As Figure 6 shown, the electronic device 600 according to an embodiment of the present invention includes a processor 601, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 602 or the program loaded from the storage section 608 into the random access memory (RAM) 603. The processor 601 can include, for example, a general microprocessor (such as a CPU), an instruction set processor, and / or a related chipset, and / or a dedicated microprocessor (such as an application specific integrated circuit (ASIC)), etc. The processor 601 can also include on-board memory for caching purposes. The processor 601 can include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present invention.

[0110] In the RAM 603, various programs and data required for the operation of the electronic device 600 are stored. The processor 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. The processor 601 performs various operations of the method flow according to the embodiments of the present invention by executing the programs in the ROM 602 and / or the RAM 603. It should be noted that the programs may also be stored in one or more memories other than the ROM 602 and the RAM 603. The processor 601 may also perform various operations of the method flow according to the embodiments of the present invention by executing the programs stored in the one or more memories.

[0111] According to an embodiment of the present invention, the electronic device 600 may further include an input / output (I / O) interface 605, and the input / output (I / O) interface 605 is also connected to the bus 604. The electronic device 600 may further include one or more of the following components connected to the I / O interface 605: an input portion 606 including a keyboard, a mouse, etc.; an output portion 607 including, for example, a cathode ray tube (CRT), a liquid crystal display (LCD), etc. and a speaker, etc.; a storage portion 608 including a hard disk, etc.; and a communication portion 609 including a network interface card such as a LAN card, a modem, etc. The communication portion 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the I / O interface 605 as needed. A removable medium 611, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 610 as needed so that a computer program read from it can be installed into the storage portion 608 as needed.

[0112] The present invention also provides a computer-readable storage medium having executable instructions stored thereon, which when executed by a processor cause the processor to execute the above-described method for processing interpretability information of an agent.

[0113] Among them, the computer-readable storage medium may be included in the device / apparatus / system described in the above embodiments; or it may exist separately and not be assembled into the device / apparatus / system. The above computer-readable storage medium carries one or more programs, which when executed, implement the method according to the embodiments of the present invention.

[0114] According to an embodiment of the present invention, the computer-readable storage medium may be a non-volatile computer-readable storage medium, for example, it may include but is not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In the present invention, the computer-readable storage medium may be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, apparatus, or device. For example, according to an embodiment of the present invention, the computer-readable storage medium may include one or more memories other than the above-described ROM 602 and / or RAM 603 and / or ROM 602 and RAM 603.

[0115] An embodiment of the present invention further includes a computer program product, which includes a computer program that, when executed by a processor, implements the interpretable information processing method of the above-mentioned agent.

[0116] Among them, the computer program contains program code for executing the method shown in the flowchart. When the computer program product runs in a computer system, this program code is used to enable the computer system to implement the method provided by the embodiment of the present invention.

[0117] When this computer program is executed by the processor 601, it executes the above-mentioned functions defined in the system / apparatus of the embodiment of the present invention. According to an embodiment of the present invention, the above-described systems, apparatuses, modules, units, etc. can be implemented by computer program modules.

[0118] In one embodiment, this computer program can rely on tangible storage media such as optical storage devices and magnetic storage devices. In another embodiment, this computer program can also be transmitted and distributed in the form of a signal on a network medium, and be downloaded and installed through the communication part 609, and / or be installed from the removable medium 611. The program code contained in this computer program can be transmitted by any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0119] In such an embodiment, this computer program can be downloaded and installed from the network through the communication part 609, and / or be installed from the removable medium 611. When this computer program is executed by the processor 601, it executes the above-mentioned functions defined in the system of the embodiment of the present invention. According to an embodiment of the present invention, the above-described systems, devices, apparatuses, modules, units, etc. can be implemented by computer program modules.

[0120] According to embodiments of the present invention, program code for executing the computer programs provided by the embodiments of the present invention can be written in any combination of one or more programming languages. Specifically, these computing programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. The programming languages include, but are not limited to, programming languages such as Java, C++, Python, the "C" language, or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving a remote computing device, the remote computing device can be connected to the user's computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (for example, by connecting through the Internet using an Internet service provider).

[0121] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code that contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram or flowchart, and the combination of blocks in the block diagram or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0122] In addition, all actions of obtaining information, signals, or data in the present invention are carried out on the premise of complying with the corresponding data protection laws, regulations, and policies of the country where it is located and obtaining authorization from the owner of the corresponding device.

[0123] Those skilled in the art can understand that the features recited in the various embodiments and / or claims of the present invention can be combined or combined in various ways, even if such combinations or combinations are not explicitly recited in the present invention. In particular, without departing from the spirit and teachings of the present invention, the features recited in the various embodiments and / or claims of the present invention can be combined and combined in various ways. All such combinations and / or combinations fall within the scope of the present invention.

[0124] The embodiments of the present invention have been described above. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of the present invention. Although the embodiments have been described separately above, this does not mean that the measures in each embodiment cannot be used advantageously in combination. The scope of the present invention is defined by the appended claims and their equivalents. Without departing from the scope of the present invention, those skilled in the art can make various substitutions and modifications, and these substitutions and modifications should fall within the scope of the present invention.

Claims

1. An interpretable information processing device for an agent, which is applied to the information update stage, task planning stage, and task execution stage during the execution of task planning, and is characterized in that including: An environmental intention update module, configured to receive environmental update information during the information update phase of the current round to update the planning intention information; A planning reflection module, configured to update the planning reflection information according to the planning intention information during the task planning phase of the current round; and An information carrier module, configured to record the planning intention information and the planning reflection information in real time in the form of preset record identification data, and generate interpretable information to be called during the task planning phase of the current round, the task execution phase of the current round, and the information update phase of the next round.

2. The device according to claim 1, characterized in that The environmental intention update module includes: A mental cognition unit, configured to receive the environmental perception information of the environmental update information during the information update phase of the current round to update the mental cognition information.

3. The device according to claim 2, wherein The environmental intention update module further includes: An intention language update unit, configured to receive the environmental sound information of the environmental update information during the information update phase of the current round to generate a natural language text.

4. The device according to claim 3, characterized in that, The environmental intention update module further includes: An intention update unit, configured to generate planning intention information according to historical feedback information, the mental cognition information, the natural language text, and its own value status information during the information update phase of the current round.

5. The device according to claim 1, characterized in that, The planning reflection module includes: A batch simulation unit, configured to perform batch simulation according to the planning intention information to generate batch simulation information for the planning intention information; An asynchronous simulation unit, configured to perform asynchronous simulation according to the planning intention information to generate asynchronous simulation information for the planning intention information.

6. The device according to claim 5, characterized in that, The planning reflection module further includes: A planning reflection unit, configured to update the planning reflection information according to the batch simulation information and the asynchronous simulation information.

7. The device according to claim 1, characterized in that It further includes: An information invocation module, configured to invoke the interpretable information during the task execution phase of the current round for interactive scenario interpretation and self-intention planning.

8. An interpretable information processing method for an agent, which is applied to the information update stage, task planning stage, and task execution stage during the execution of task planning, and is characterized in that, including: Receiving environmental update information during the information update phase of the current round to update the planning intention information; Updating the planning reflection information according to the updated planning intention information during the task planning phase of the current round; and Recording the planning intention information and the planning reflection information in real time in the form of preset record identification data, and generating interpretable information to be called during the task planning phase of the current round, the task execution phase of the current round, and the information update phase of the next round.

9. An electronic device, including: One or more processors; A memory, configured to store one or more programs, wherein, when the one or more programs are executed by the one or more processors, the one or more processors execute the method according to claim 8.

10. A computer-readable storage medium, on which executable instructions are stored, and when the instructions are executed by a processor, the processor executes the method according to claim 8.

Images