Multi-robot coordination method, system, and multi-robot system

Abstract

The present application relates to the technical field of multi-robot cooperation, and provides a multi-robot cooperation method, a multi-robot cooperation system and a multi-robot system. The multi-robot cooperation method based on service state triggering comprises the following steps: continuously sensing environment state and user state information through a multi-modal sensor to construct a service state vector for quantitatively describing the interaction between the user, the environment and the robot; in response to the deviation degree of the service state vector and the expected service state meeting a cooperation triggering condition, generating a cooperation demand indicating a target service state and broadcasting the cooperation demand to other robots; based on the cooperation demand, jointly negotiating with the other robots to determine a cooperation structure of the cooperation demand, the cooperation structure comprising cooperation robots participating in execution and role allocation of the cooperation robots; and continuously monitoring the service state vector when the cooperation robots execute corresponding behaviors based on the role allocation.

Description

Technical Field

[0001] This invention relates to the field of multi-robot collaboration technology, and in particular to a multi-robot collaboration method, system, and multi-robot system. Background Technology

[0002] In the field of multi-robot systems, existing technologies typically employ centralized or quasi-centralized task allocation and scheduling methods. In a typical scheme, a central control unit or master robot is responsible for collecting environmental perception information, task requirement information, and the status data of each robot. Based on preset rules or optimization algorithms, it generates task allocation results and then issues execution instructions to each robot node. However, in this typical existing technology, task requirements are explicitly given by human users or higher-level systems; the task is broken down into several sub-tasks and then uniformly allocated by the central node; each robot mainly acts as an execution unit, passively receiving the allocation results and completing the corresponding operations. This scheme can work normally in scenarios with a small number of robots and relatively slow environmental changes, but as the number of robots increases or the dynamics of the environment increase, the central node easily becomes a performance bottleneck, and the system's dependence on the central node leads to low overall robustness.

[0003] To overcome the single point of failure problem in centralized architectures, existing technologies have proposed distributed bidding or contract-like multi-robot collaborative schemes. The basic process is as follows: a robot or system module generates a clear task description and broadcasts it to all robots. Each robot calculates the task execution cost based on its own location, load capacity, energy consumption, and other factors, and competes for the task execution right through bidding or auction. Ultimately, the system that issued the task selects the robot with the optimal cost to undertake the task. However, in existing technologies, the tasks issued are clearly described before bidding begins. The core basis for bidding decisions is physical or resource-related parameters such as time, energy consumption, path, or risk. The collaborative goal is usually to determine the optimal or near-optimal robot to execute the single issued task. Therefore, distributed bidding or contract-like multi-robot collaborative schemes are typically applied in scenarios with clear objectives and task boundaries, such as logistics warehousing and industrial handling. Furthermore, this collaborative scheme only achieves decentralization in form; essentially, it is still a task-centric allocation mechanism that treats each robot as an execution unit. Moreover, in this collaborative scheme, robots need to wait for external instructions or task assignments to execute collaborative tasks. These methods lack mechanisms to proactively trigger collaboration, cannot autonomously determine whether collaboration is needed based on environmental changes, and cannot trigger collaborative behavior in real time.

[0004] In fields where multiple robots are needed to provide long-term, continuous, and adaptive services, such as elderly care, smart homes, public service spaces (such as hospitals and banks), and smart exhibition halls, user needs are usually not given directly in the form of clear instructions or specific tasks. Therefore, existing technologies cannot cope with dynamic human-machine coexistence service scenarios where task objectives are unclear.

[0005] With the development of robotics, robots can achieve human-computer interaction through multimodal perception. Robots acquire environmental and user interaction information through multimodal perception modules such as vision, speech, and touch, and use intent recognition models to analyze user needs or emotional states, thereby driving a single robot to achieve human-computer interaction. However, current technologies focus on the closed loop of perception, decision-making, and execution within a single robot, remaining at the level of individual intelligence. Simply improving the intent recognition capabilities of a single robot cannot solve the problem of collaborative organization of multiple robots in complex interaction scenarios.

[0006] In order to overcome the above-mentioned defects of the existing technology, there is an urgent need in this field for a multi-robot collaboration technology based on service state triggering, which can autonomously form a reasonable collaborative division of labor relationship to complete complex service behaviors in human-computer interaction scenarios where the task objectives are unclear and the needs gradually become apparent, without the need for central control. Summary of the Invention

[0007] The following provides a brief overview of one or more aspects to offer a basic understanding of them. This overview is not an exhaustive summary of all conceived aspects, nor is it intended to identify key or decisive elements of all aspects, nor to define the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to the more detailed descriptions that follow.

[0008] To overcome the aforementioned deficiencies in existing technologies, this invention provides a multi-robot collaboration method, system, and system. Based on service state triggering, the multi-robot collaboration method, system, and system can autonomously form a reasonable collaborative division of labor relationship to complete complex service behaviors in human-computer interaction scenarios where task objectives are unclear and needs gradually emerge, without the need for central control.

[0009] Specifically, the above-described service state-triggered multi-robot collaboration method according to the first aspect of the present invention includes the following steps: continuously sensing environmental and user state information through multimodal sensors to construct a service state vector for quantitatively describing the interaction between the user, the environment, and the robot; generating a collaboration requirement indicating the target service state and broadcasting the collaboration requirement to other robots in response to the degree of deviation between the service state vector and the desired service state satisfying the collaboration triggering condition; jointly negotiating with the other robots based on the collaboration requirement to determine the collaboration structure of the collaboration requirement, the collaboration structure including participating collaborative robots and role assignments of the collaborative robots; and continuously monitoring the service state vector when the collaborative robots perform corresponding behaviors based on the role assignments.

[0010] Furthermore, in some embodiments of the present invention, the joint negotiation is achieved through a bidding negotiation mechanism.

[0011] Furthermore, in some embodiments of the present invention, the multi-robot collaboration method further includes the steps of: listening to broadcasts from other robots to receive the collaboration request; and, in response to receiving the collaboration request, participating in the joint negotiation based on the state contribution of the collaboration request to jointly determine the collaboration structure of the collaboration request with the other robots.

[0012] Furthermore, in some embodiments of the present invention, the step of generating a collaborative requirement indicating a target service state and broadcasting the collaborative requirement to other robots in response to the deviation between the service state vector and the desired service state satisfying a collaborative triggering condition includes: calculating the deviation between the service state vector and the desired service state as a state tension value; and generating the collaborative requirement and broadcasting the collaborative requirement to other robots when the state tension value exceeds a preset threshold.

[0013] Furthermore, in some embodiments of the present invention, the step of generating a collaborative requirement indicating the target service state and broadcasting the collaborative requirement to other robots in response to the deviation between the service state vector and the desired service state satisfying the collaborative triggering condition further includes: when the state tension value exceeds a preset threshold, inputting service state vector data within a first time range into a world model to predict the service state curve within a future second time range; and generating the collaborative requirement and broadcasting the collaborative requirement to other robots in response to the service state curve within the second time range indicating a deterioration in the service state.

[0014] Furthermore, in some embodiments of the present invention, the service state vector includes a security risk concern parameter and an interaction activity parameter.

[0015] Furthermore, in some embodiments of the present invention, the multi-robot collaboration method further includes the steps of: recording historical collaboration events in a collaboration effect history database, wherein the historical collaboration events include a first service state vector when the collaboration triggering condition is triggered, the collaboration structure negotiated under the first service state vector, and a second service state vector after performing corresponding actions based on the collaboration structure; and in response to the degree of deviation between the service state vector and the desired service state satisfying the collaboration triggering condition and the service state vector being similar to the first service state vector, preferentially referring to the collaboration structure negotiated under the first service state vector during the joint negotiation.

[0016] Furthermore, in some embodiments of the present invention, the step of continuously monitoring the service state vector when the collaborative robot performs the corresponding behavior based on the role assignment includes: dissolving the collaborative structure in response to the service state vector of the collaborative robot performing the corresponding behavior based on the role assignment reaching the target service state.

[0017] Furthermore, the service state-triggered multi-robot collaborative system provided according to the second aspect of the present invention includes a memory and a processor. The memory stores computer instructions. The processor is connected to the memory and configured to execute the computer instructions stored in the memory to implement the service state-triggered multi-robot collaborative method provided in any of the above embodiments.

[0018] Furthermore, the above-described service state-triggered multi-robot system provided according to the third aspect of the present invention includes multiple robots, wherein the robots are configured with the service state-triggered multi-robot collaborative system provided in the second aspect of the present invention.

[0019] This invention introduces a state-driven collaboration mechanism to perceive the environment and user states to determine the service state. Each robot then autonomously determines whether to trigger collaboration based on the service state and automatically assesses the collaboration requirements. After triggering collaboration, the invention further negotiates among multiple robots based on the service state offsets indicated by the requirements to determine the executing robot and its role. This naturally allows each robot to participate and adjust in a timely manner according to the situation, with each executing robot independently deciding the specific task it will perform. In this way, robots are not merely passively executing tasks, but can proactively initiate collaboration based on the state assessment results, achieving more flexible, real-time, and efficient multi-robot collaborative work.

[0020] This invention achieves at least the following beneficial effects: It shifts from "task allocation" to "collaborative structure formation": This invention no longer relies on predefined tasks for collaboration, but instead uses service state triggering as the core to drive multiple robots to form collaborative relationships, significantly improving the system's adaptability to complex interaction scenarios; It enhances the flexibility and self-organization capabilities of multi-robot collaboration: Through autonomous bidding and allocation of collaborative roles, this invention enables robots to dynamically participate in different collaborative roles based on their capabilities and state-determined state contributions, avoiding the limitations of a single executor model; It enhances the robustness and scalability of multi-robot systems under decentralized conditions: The collaborative structure of this invention is formed through autonomous negotiation by robots, without relying on a central node, reducing the risk of single-point failures and making it suitable for application scenarios with dynamically changing robot numbers; It improves service continuity and naturalness in human-computer interaction scenarios: Through the dynamic adjustment and evolution of the collaborative structure, this invention enables multiple robots to continuously optimize their collaborative behavior as service states change, improving the overall service experience. Attached Figure Description

[0021] The above-described features and advantages of the present invention will be better understood after reading the following detailed description of embodiments of the present disclosure in conjunction with the accompanying drawings. In the drawings, components are not necessarily drawn to scale, and components having similar related characteristics or features may have the same or similar reference numerals.

[0022] Figure 1 A schematic diagram of a multi-robot collaborative system based on service state triggering, according to some embodiments of the present invention, is shown.

[0023] Figure 2 A flowchart of a multi-robot collaboration method based on service state triggering according to some embodiments of the present invention is shown;

[0024] Figure 3 A flowchart of a multi-robot collaboration method based on service state triggering according to some embodiments of the present invention is shown;

[0025] Figure 4 A flowchart of a multi-robot collaborative method based on service status triggering in a health and wellness care scenario, according to Embodiment 3 of the present invention, is shown.

[0026] Figure label:

[0027] 100: A multi-robot collaborative system triggered by service status;

[0028] 110: Memory;

[0029] 111: Computer-readable storage medium;

[0030] 120: Processor;

[0031] S310~S340: Steps. Detailed Implementation

[0032] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. Although the description of the present invention is presented in conjunction with preferred embodiments, this does not mean that the features of the invention are limited to these embodiments. On the contrary, the purpose of describing the invention in conjunction with embodiments is to cover other options or modifications that may be derived based on the claims of the present invention. To provide a thorough understanding of the invention, many specific details will be included in the following description. The invention may also be implemented without using these details. Furthermore, to avoid confusion or obscuring the focus of the invention, some specific details will be omitted in the description.

[0033] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0034] Furthermore, the terms "upper," "lower," "left," "right," "top," "bottom," "horizontal," and "vertical" used in the following description should be understood as the orientations shown in the relevant paragraphs and accompanying drawings. These relative terms are for illustrative purposes only and do not imply that the described apparatus must be manufactured or operated in a specific orientation, and therefore should not be construed as limiting the invention.

[0035] It is understood that although terms such as "first," "second," and "third" may be used herein to describe various components, regions, layers, and / or parts, these components, regions, layers, and / or parts should not be limited by these terms, and these terms are only used to distinguish different components, regions, layers, and / or parts. Therefore, the first components, regions, layers, and / or parts discussed below may be referred to as second components, regions, layers, and / or parts without departing from some embodiments of the present invention.

[0036] As mentioned above, existing multi-robot collaborative technologies based on bidding or contract-like networks only achieve decentralization in form; the collaborative object is still the predefined task itself. This technological framework assumes the following premises: the task objective is clear before collaboration begins; the task execution method and completion criteria are determined; and the core issue of collaboration is simply "who will execute the task." However, in complex service scenarios involving human-machine coexistence, these premises are often difficult to uphold. User needs are often vague, progressive, and uncertain, making it difficult to accurately articulate them as directly executable tasks in the initial stage, thus hindering the effective application of existing bidding-based collaborative mechanisms.

[0037] Existing embodied robots can only achieve human-robot interaction in single-robot scenarios. In multi-robot systems, simply using the intent recognition results of a single robot as input for task generation and applying existing task bidding and allocation mechanisms still presents the following problems: a lack of modeling of the collaborative relationships among multiple robots; an inability to express the different functions or roles undertaken by different robots in the collaborative process; and difficulty in dynamically adjusting the division of labor based on the progress of the interaction. Therefore, simply improving intent recognition capabilities cannot fundamentally solve the problem of collaborative organization of multiple robots in complex interaction scenarios.

[0038] To overcome the aforementioned deficiencies in existing technologies, this invention provides a multi-robot collaboration method, system, and system. Based on service state triggering, the multi-robot collaboration method, system, and system can autonomously form a reasonable collaborative division of labor relationship to complete complex service behaviors in human-computer interaction scenarios where task objectives are unclear and needs gradually emerge, without the need for central control.

[0039] The service state-triggered multi-robot system described above, provided in the third aspect of this invention, employs a fully distributed peer-to-peer network architecture. This service state-triggered multi-robot system consists of multiple independent robot nodes, all of which are physically and logically equal, without a central control node to centrally command all robots. Each robot can be both an initiator and a participant in the negotiation process.

[0040] The service state-triggered multi-robot collaborative system provided in the second aspect of the present invention can be configured in each robot of a service state-triggered multi-robot system, and each robot is connected to other robots via a system network. Here, the system network can be a wide area network or a local area network, or a combination of both.

[0041] Each robot can proactively collect user status information data with specific user authorization, or it can receive relevant status information data actively input by the user via a human-machine interface, enabling the robots to perceive the environmental and user states. Depending on implementation needs, a multi-robot system triggered by service states can have any number of robots.

[0042] The multi-robot system formed by the robots based on service state triggering does not have a unified control or scheduling node. Each robot node is logically equal and is equipped with the multi-robot collaborative system based on service state triggering provided by this invention. This enables each robot to have the ability to identify service state, make collaborative decisions, and execute roles. Multi-robot collaboration and role emergence are achieved through peer-to-peer communication between the robots.

[0043] Please refer to Figure 1 , Figure 1 A schematic diagram of a multi-robot collaborative system based on service state triggering, according to some embodiments of the present invention, is shown.

[0044] exist Figure 1 In the illustrated embodiment, the service state-triggered multi-robot collaborative system 100 may be configured with a memory 110 and a processor 120. The memory 110 includes, but is not limited to, a computer-readable storage medium 111 on which computer instructions are stored. The processor 120 is connected to the memory 110 and is configured to execute the computer instructions stored in the memory 110 to implement the service state-triggered multi-robot collaborative method provided in the first aspect of the present invention.

[0045] In some embodiments, a service state-triggered multi-robot collaborative system may include multiple program modules. These program modules may be stored in memory. Program modules include, but are not limited to, an operating system, one or more applications, other program modules, and program data; each or some combination of these examples may include an implementation of a network environment. The program modules typically execute the service state-triggered multi-robot collaborative method described in the embodiments of this invention.

[0046] Preferably, the program modules of the multi-robot collaborative system based on service state triggering may include a multimodal active perception module, a service state modeling module, a service state tension assessment module, a collaborative state target release module, a state contribution assessment and collaborative negotiation module, and a collaborative structure execution and evolution module.

[0047] The following will first describe the working principle of the service state-triggered multi-robot collaborative system using some embodiments of service state-triggered multi-robot collaborative methods. Those skilled in the art will understand that these embodiments of service state-triggered multi-robot collaborative methods are merely non-limiting implementations provided by this invention, intended to clearly demonstrate the main concepts of the invention and provide specific solutions convenient for public implementation, rather than limiting all functions or operating methods of the service state-triggered multi-robot collaborative system. Similarly, this service state-triggered multi-robot collaborative system is also only one non-limiting implementation provided by this invention, and does not limit the executing entities and execution order of the steps in these service state-triggered multi-robot collaborative methods.

[0048] Please refer to the reference. Figure 2 and Figure 3 , Figure 2 and Figure 3 A flowchart of a multi-robot collaboration method based on service state triggering, according to some embodiments of the present invention, is shown.

[0049] A multi-robot collaborative system triggered by service status can operate independently in each robot node when the multi-robot system is started and distributed monitoring is implemented. This multi-robot system has no central control.

[0050] Then, as Figure 2 and Figure 3 As shown, the multi-robot collaborative system can execute step S310 through the multimodal active perception module and the service state modeling module: continuously perceive environmental and user state information through multimodal sensors to construct a service state vector for quantitatively describing the interaction between users, environment and robots.

[0051] Here, multimodal sensors refer to sensing systems integrated with various types of perception modules mounted on a robot. Multimodal sensors can acquire environmental information and human-robot interaction information in real time. Multimodal sensors include, but are not limited to, vision sensors, hearing sensors, LiDAR, and inertial measurement units.

[0052] User status information can include user behavior, user voice, and user spatial trajectory.

[0053] Service status is typically reflected indirectly through changes in user behavior, semantic interaction fragments, spatial activity characteristics, or environmental interaction methods. In dynamic human-machine coexistence service scenarios where task objectives are unclear and user needs gradually emerge during interaction, user needs are usually not directly given in the form of explicit instructions or specific tasks, but rather expressed indirectly and continuously evolving through changes in behavioral patterns, scattered semantic information, and activity trajectories in space. For example, in a senior care scenario, an elderly person may only frequently get up, pace around the room without a clear destination, or show signs of low mood, but without verbally requesting assistance or giving specific instructions; in a smart exhibition hall, a visitor may linger in front of an exhibit for a long time, but without actively requesting a guided tour.

[0054] The multi-robot collaborative system continuously and proactively senses environmental and user status information through multimodal sensors, constructs a service state vector to quantitatively describe the interaction between users, the environment, and robots, and identifies and evaluates potential service states that have not yet been explicitly expressed as specific tasks or operation instructions based on the perception results, thereby determining whether there are service needs that require multi-robot collaboration.

[0055] In an optional embodiment, the multi-robot collaborative system can communicate with other robots to synchronously acquire relevant data collected by the multimodal sensors on other robots, and combine the environmental state and user state information sensed by the synchronously acquired multimodal sensors of other robots to understand and identify the current service state.

[0056] In some embodiments, the multi-robot collaborative system generates quantifiable index parameters by combining the perceived environmental state and user state information to form a service state vector. The quantifiable index parameters may include, but are not limited to, safety risk concern parameters and interaction activity parameters.

[0057] Specifically, the multimodal active perception module of the multi-robot collaborative system is responsible for continuously and proactively collecting multi-source information about the environment and the user. The inputs to the multimodal active perception module include, but are not limited to, visual information acquired by multimodal sensors (such as user position and posture), voice information (such as voice content and tone), user behavior trajectories (such as movement path and speed), and changes in environmental conditions (such as light and abnormal sounds). For example, the multimodal active perception module can determine whether the user's movement frequency within the room is abnormal through data streams collected by cameras, and detect whether voice interaction is interrupted for extended periods through a microphone array.

[0058] Then, based on the real-time data provided by the multimodal active perception module, the service state modeling module can construct a service state vector that can quantitatively describe the current overall interaction between "human-environment-robot". In some embodiments, the service state vector may include key state indicators such as service continuity, safety risk awareness, and interaction activity. For example, in a companionship scenario, the service state modeling module can output a service state vector that uses a decrease in the interaction activity indicator to represent "the user has not had any effective interaction with any robot for 10 consecutive minutes" and an increase in the safety risk awareness indicator to represent "the user's activity range is expanding towards the risk area".

[0059] In this way, the multi-robot collaborative system, through its onboard multimodal sensors, continuously and autonomously collects and analyzes environmental and user interaction data without explicit task instructions, in order to construct a service state vector that quantitatively describes the interaction between users, environment and robots. Based on the service state vector, it identifies user state, environmental context and service state changes, transforming from the traditional "passive response to instructions" to "active collaboration based on service state triggering".

[0060] Multi-robot collaborative systems provide a data foundation for service status identification through multimodal information fusion. The identified service status is used to guide the direction of multi-robot collaborative decision-making and is the basis for triggering collaboration and role emergence in multi-robot collaborative systems.

[0061] Please continue to refer to this. Figure 2 and Figure 3 The service state tension assessment module and the collaborative state target release module of the multi-robot collaborative system can execute step S320: in response to the degree of deviation between the service state vector and the desired service state satisfying the collaborative triggering condition, generate a collaborative requirement indicating the target service state and broadcast the collaborative requirement to other robots.

[0062] The collaboration triggering condition is the condition that triggers multi-robot collaboration among robots in a multi-robot system. When the deviation between the service state vector and the desired service state meets the collaboration triggering condition, a collaboration requirement indicating the target service state is generated and broadcast to other robots. Under the premise of sharing the service state, multiple robots autonomously form a negotiation result based on their understanding of the service state through communication and a distributed decision-making mechanism.

[0063] In some embodiments, the coordination condition may be that the degree to which the service state deviates from the expected service state is greater than a threshold.

[0064] Specifically, the service state tension assessment module of a multi-robot collaborative system can compare the service state vector S with the desired service state S. And calculate the service state vector S and the desired service state Si. The degree of deviation is used as the state tension value T. The desired service state can be preset by the multi-robot collaborative system, or it can be obtained through learning and optimization.

[0065] Preferably, the state tension value can be determined by the state deviation magnitude, duration weight, and deterioration trend factor.

[0066] When the calculated state tension value T does not exceed the preset threshold, the multi-robot collaborative system can determine that the current service state meets expectations, and there is no need to trigger collaboration between robots. The current service state can be maintained and each robot can continue to provide services.

[0067] When the calculated state tension value T exceeds a preset threshold, the multi-robot collaborative system can determine that the current service status has deviated from the ideal trajectory and requires triggering collaboration between robots to correct it. For example, when the expected service status of the multi-robot collaborative system is "the user should be in a state of continuous attention", but the actual detected interaction continuity index indicated in the service status vector continues to decline, the service state tension assessment module will calculate a high state tension value.

[0068] When the deviation between the service state vector and the desired service state meets the collaborative triggering condition, it indicates that the multi-robot collaborative system determines that the current service state deviates from the desired state and requires robot collaboration to correct it, thereby triggering multi-robot collaboration and activating the collaborative state target publishing module of the multi-robot collaborative system.

[0069] The collaborative state goal publishing module can generate a collaborative requirement indicating the service state goal using methods such as large models, and broadcast the collaborative requirement to other robots in the multi-robot system. This collaborative requirement differs from specific, directly executable task instructions (such as "patrol point A"). The collaborative requirement is a state optimization description indicating the service state goal; that is, a state description to be improved, such as "increase attention to users and reduce safety risks in their surrounding environment."

[0070] For example, by inputting service state vectors representing deviations from expectations in security risk awareness indicators (i.e., high security risk) and interaction activity indicators (i.e., low interactivity) into pre-trained models such as large models, the collaborative state target release module can determine target service states such as "reducing security risk awareness parameters and interaction activity parameters" as collaborative requirements.

[0071] In other words, when a robot recognizes that a service state meets the preset collaboration trigger conditions, it does not directly generate a specific executable task. Instead, it generates a collaboration requirement description for that service state and sends collaboration invitations to other robots in the network based on the collaboration requirement description. Thus, after triggering multi-robot collaboration in the multi-robot collaboration system, the collaboration will start by negotiating around the service state rather than around a predefined task. That is, the goal of guiding multi-robot collaboration is state optimization rather than task execution.

[0072] Furthermore, by continuously modeling and assessing the overall interaction state of "human-environment-robot," this invention can trigger collaborative behavior in advance before service quality significantly declines, thus avoiding the response lag caused by waiting for explicit task instructions in traditional technologies. Moreover, since the service state itself is continuous and quantifiable, this triggering mechanism can naturally support adaptive adjustments to the collaborative scale, role division, and intervention intensity, which is difficult to achieve with collaborative methods based on discrete task triggering.

[0073] Even better, when the state tension value exceeds a preset threshold and the collaborative state target release module is activated to generate a collaborative requirement, the collaborative state target release module can further determine whether to trigger collaboration based on a world model. The collaborative state target release module can introduce a world model to dynamically model and predict the service state vector reflecting the interaction between users, environment, and robots. It not only judges anomalies based on historical service state change trends, but also further predicts multiple possible evolution trajectories of the service state in the future time range.

[0074] Specifically, the collaborative state target publishing module inputs the service state vector data within the first time range into the world model to predict the service state curve within the second time range. When the service state curve within the second time range indicates a deterioration in the service state, i.e., when the service state is unlikely to naturally revert to a stable range in most evolution paths, the collaborative state target publishing module determines that the state has an irreversible deterioration trend in a predictive sense. At this time, the collaborative state target publishing module generates a collaborative requirement and broadcasts the collaborative requirement to other robots to trigger multi-robot collaborative intervention.

[0075] The preset threshold can be a relatively broad threshold, so that by introducing a world model, it is possible to determine more early whether to trigger multi-robot collaborative intervention.

[0076] This collaborative approach, based on negotiation of collaborative needs, shifts the focus of collaboration from "executing pre-defined tasks" to "jointly optimizing service states." This fundamentally reduces the reliance on precise task modeling, decomposition, and description, significantly enhancing the adaptability and robustness of multi-robot systems in open, dynamic, and unstructured environments. Even when faced with unforeseen new scenarios, multi-robot collaborative systems can flexibly organize their collaborative behavior around state objectives, without being limited by a fixed task library.

[0077] In this way, the multi-robot collaborative system uses the identification and change of service status as the trigger and decision-making basis for multi-robot collaboration, rather than using a clear task or operation instruction as the only trigger condition. This allows the multi-robot collaborative system to drive collaboration through service status, enabling multiple robots to form collaborative behavior before the task is clearly defined.

[0078] Compared to the traditional approach in existing technologies that relies on explicit task instructions or specific events to trigger collaboration, this invention proposes a novel triggering mechanism centered on service status. This mechanism quantifies service status, compares it with a preset expected service status, and calculates the degree of deviation. When the deviation exceeds a set threshold, it determines that the current service situation has "tension" or a "gap," thereby autonomously triggering a multi-robot collaborative process. This triggering mechanism allows for proactive collaboration to be initiated in the early stages, before user needs are clear and tasks are formed, preventing service quality degradation at the status level. Furthermore, the early intervention of the multi-robot system effectively avoids response delays caused by waiting for explicit instructions in open service scenarios, greatly improving the timeliness of response to potential and ambiguous needs, thus significantly enhancing service continuity, proactivity, and overall security.

[0079] Please continue to refer to this. Figure 2 and Figure 3 The state contribution assessment and collaborative negotiation module of the multi-robot collaborative system can perform step S330: based on collaborative needs, negotiate with other robots to determine the collaborative structure, which includes the collaborative robots participating in the execution and the role allocation of the collaborative robots.

[0080] The robot that triggers the collaboration conditions and generates the collaboration request acts as the party initiating and publishing the collaboration request, and sends the collaboration request to each robot in the multi-robot system.

[0081] In a multi-robot system, each robot is equipped with a state contribution assessment and collaborative negotiation module that continuously listens to broadcasts from other robots to receive collaborative requests. Upon receiving a collaborative request, each robot can participate in negotiation based on its own state contribution to the request, in order to jointly determine the collaborative structure with other robots.

[0082] During the negotiation process, each robot can predict its contribution to the achievement of the service state goal based on its own capability profile, current resource status (such as battery level, location sensor configuration, and current load), environmental context, and historical collaborative performance, thus determining its own contribution to the collaborative requirements. Based primarily on this contribution, each robot autonomously evaluates different collaborative roles within the collaborative requirements and submits bids for one or more roles. Through distributed negotiation, it autonomously determines the participants and role assignments, forming the collaborative structure. This collaborative structure emerges through negotiation based on each node's assessment of its contribution to the global goal, rather than being assigned by a central node or pre-defined by fixed rules.

[0083] For example, during collaborative negotiations, robots can leverage large-scale models and their own profiles and state contribution information to achieve an "emergent" outcome. For instance, a robot might be a patrol dog. When a collaborative requirement indicates a need to "improve environmental safety," the robot can input this requirement and its own profile (as a robot dog) into a pre-trained large-scale model to obtain its state contribution information. Based on this state contribution information, it can then participate in a bidding process to determine the final collaborative structure.

[0084] In some embodiments, the state contribution can be determined by the number of dimensions that the robot can improve, the rate of improvement, and the execution stability.

[0085] In a multi-robot system, the robots negotiate together through a bidding-based negotiation mechanism.

[0086] The bidding-based negotiation mechanism refers to a distributed collaborative decision-making method inspired by bidding mechanisms, such as market-based auction algorithms, Contract Net Protocol (CNP), and consensus protocols. In this invention, this bidding-based negotiation mechanism is used to support service state-driven collaborative processes. The bidding object is the "right to assume collaborative roles" rather than the "right to execute tasks." Each robot participates in the negotiation based on its own state and capabilities to form a collaborative role structure adapted to the current service state, rather than simply being used for task allocation.

[0087] Through a bidding-based negotiation mechanism, each robot conducts distributed negotiation around the collaborative goals corresponding to its service state. Without predefined role assignments or central scheduling, a collaborative decision is formed based on an assessment of its own state, capabilities, and potential impact after participating in the collaboration. This decision determines the collaborative structure comprised of multiple robots. The collaborative structure includes the participating collaborative robots and their assigned roles. In some embodiments, the set of robots participating in the collaboration may include one or more collaborative robots, and multiple collaborative robots can form a collaborative team.

[0088] The collaborative structure formed by the autonomous emergence of service states is entirely based on the assessment of each robot's contribution to achieving the "state goal" and distributed negotiation, rather than completing pre-set fixed tasks. While achieving self-organization and optimized resource allocation within the collaborative structure, it makes collaborative combinations more targeted and flexible, reducing the rigidity and overhead of centralized scheduling, and avoiding single-point bottlenecks, computational burdens, and unreasonable allocation due to incomplete information. Furthermore, this collaborative structure can prevent task overlap or omission, improving overall efficiency.

[0089] After that, as Figure 2 and Figure 3 As shown, the collaborative structure execution and evolution module of the multi-robot collaborative system can execute step S340: continuously monitor the service state vector of the collaborative robot when it performs corresponding behaviors based on role assignment.

[0090] The collaborative structure execution and evolution module of the robot that publishes collaborative requests can maintain continuous monitoring of the service status during the process of the collaborative robot performing the corresponding behavior.

[0091] A robot that issues a collaboration request may succeed in the negotiation process and become a participating collaborative robot, or it may fail and not participate. When a robot that issues a collaboration request negotiates with other robots and the negotiation result is participation in execution, it performs the corresponding behavior based on the collaboration structure. Specifically, collaborative robots can determine their assigned collaborative roles based on role allocation and further determine the collaborative tasks performed by those roles, as well as the corresponding perception, interaction, or operational behaviors. Each robot can use relevant models to understand its assigned collaborative role in order to determine the specific tasks it will perform.

[0092] During execution, the collaborative robot can continuously perceive environmental conditions and user behavior through multimodal sensors to obtain changes in service status. Preferably, the collaborative robots in the collaborative group can maintain synchronization of service status.

[0093] Based on changes in service status, the collaborative structure execution and evolution module of the collaborative robot can adjust the collaborative structure, so that the formed collaborative roles evolve with changes in service status and service process, in order to adapt to changes in demand in continuous service scenarios.

[0094] In some embodiments, the collaborative structure is disbanded in response to the service state vector of the collaborative robot performing corresponding actions based on role assignment reaching the target service state.

[0095] In some embodiments, when a significant change in the service status is detected (such as the disappearance of state tension due to the satisfaction of user needs), the collaborative structure execution and evolution module of the collaborative robot will re-evaluate the current collaborative structure to achieve dynamic adjustment or disbandment of the collaborative group.

[0096] In some embodiments, the current collaborative structure can be reassessed when one or more collaborative robots in a collaborative group detect a change in service status or a deviation from expected collaborative performance. Based on the results of the reassessment, adjustments to collaborative roles are triggered. The collaborative state target publishing module of the collaborative robot that first triggers the collaborative structure adjustment in the collaborative group sends the request for adjustment of the collaborative structure to other collaborative robots in the collaborative group. Alternatively, based on the results of the reassessment, renegotiation is triggered. The collaborative state target publishing module of the collaborative robot that first triggers the collaborative conditions in the collaborative group sends the renegotiation collaborative request to other robots in the multi-robot system.

[0097] Therefore, multi-robot systems can achieve dynamic evolution of their cooperative structure. Here, the conditions that trigger adjustments to the cooperative structure and the cooperative conditions that trigger renegotiation can be preset or obtained through learning and optimization.

[0098] Through monitoring the execution process and dynamically adjusting and evolving the collaborative structure, the roles of each robot in the collaboration are not pre-defined statically (such as fixed "patrolmen" or "companions"). Roles emerge dynamically during the distributed negotiation process, based on each robot node's contribution to different state objectives. More importantly, the multi-robot collaborative system continuously monitors changes in service status during collaborative execution. Once the service status changes (such as goal achievement or new demands), the established collaborative structure can automatically trigger adjustments or renegotiations, thereby adjusting, adding, deleting, or reconstructing role assignments, achieving dynamic evolution of the collaborative structure.

[0099] Thus, this invention endows multi-robot systems with strong long-term operational adaptability and scenario flexibility. Like an organism, the multi-robot system can dynamically adjust its organizational form according to changes in the service process and environment, ensuring service stability and a high-quality user experience in long-term, complex, and ever-changing scenarios such as companionship and guidance.

[0100] Even better, the multi-robot collaborative system can also introduce a world model into the collaborative structure execution and evolution module to predict the state evolution results under different collaborative structures, thereby making a forward-looking assessment and selection of collaborative intervention methods, and realizing a proactive collaborative decision-making mechanism that prevents the continuous deterioration of service status with minimal collaborative cost.

[0101] Furthermore, in some embodiments, the multi-robot collaborative system can also possess continuous learning and optimization capabilities through collaborative memory functionality.

[0102] Multi-robot collaborative systems can record historical collaborative events in a collaborative effect history database to form an experience knowledge base. Historical collaborative events include the first service state vector when the collaborative trigger condition is triggered, the collaborative structure formed under the first service state vector, and the second service state vector after executing the corresponding action based on the collaborative structure. By using the first service state vector that triggers the collaborative condition and the second service state vector after executing the corresponding action based on the collaborative structure, the multi-robot collaborative system can further record the execution effects of historical collaborative events.

[0103] Multi-robot collaborative systems can learn and optimize based on historical collaborative events in a historical database of collaborative effects.

[0104] Based on successful historical collaborative events, multi-robot collaborative systems can refine the models for predicting state contributions and making predictions based on state contributions. When encountering a service state with triggering collaborative conditions similar to the first service state vector, the multi-robot collaborative system can prioritize referencing the collaborative structure corresponding to the first service state vector as a priori suggestion when negotiating with other robots through a bidding-based negotiation mechanism to determine the collaborative structure, and when each robot predicts its state contribution. This guides the multi-robot collaborative system to more quickly form a collaborative structure that has been historically proven to be efficient.

[0105] Based on the execution results of historical collaborative events, a multi-robot collaborative system can adjust the collaborative conditions that trigger multi-robot collaboration. For example, when the execution results of historical collaborative events deviate from the expected service state, the collaborative structure can be readjusted, and the collaborative conditions that trigger collaboration can also be adjusted.

[0106] As a result, with the accumulation of operating time, the collaborative decision-making of multi-robot collaborative systems will become more accurate and efficient, and the overall service performance will continue to improve, achieving the evolution from "able to collaborate" to "good at collaborating".

[0107] In summary, existing technologies lack a collaborative mechanism that can guide multiple robots to dynamically form a collaborative relationship with differentiated roles around the service state before the task is clearly defined as a specific operation. This results in insufficient adaptability and low collaborative efficiency of existing technologies in complex service scenarios.

[0108] To address the problems existing in the prior art, this invention introduces a service state-based collaborative triggering mechanism and a bidding and negotiation method oriented towards collaborative roles, which can achieve at least the following beneficial effects:

[0109] Achieving a shift from "task allocation" to "collaborative structure formation": This invention no longer relies on predefined tasks for collaboration, but instead uses service status as the core to drive multiple robots to form collaborative relationships, significantly improving the system's adaptability to complex interactive scenarios;

[0110] Enhancing the flexibility and self-organization of multi-robot collaboration: This invention enables robots to dynamically participate in different collaborative roles based on their capabilities and state-determined state contribution through autonomous bidding and allocation of collaborative roles, thus avoiding the limitations of the single executor mode.

[0111] Enhancing the robustness and scalability of multi-robot systems under decentralized conditions: The collaborative structure of this invention is formed by autonomous negotiation among robots, without relying on a central node, reducing the risk of single point of failure, and is suitable for application scenarios where the number of robots changes dynamically.

[0112] Improving service continuity and naturalness in human-computer interaction scenarios: This invention enables multiple robots to continuously optimize their collaborative behavior as the service status changes through dynamic adjustment and evolution of the collaborative structure, thereby improving the overall service experience.

[0113] The following are several specific embodiments in the context of elderly care and companionship, which will be used to elaborate on the multi-robot collaboration mechanism based on service status triggering provided by the present invention.

[0114] In the initial service state of Example 1, each robot node in the multi-robot system continuously collects multimodal data about the environment and the user through its onboard multimodal sensors. Then, each robot node can run the service state modeling module based on the multimodal data, fusing and abstracting the multimodal data into a numerical current service state vector.

[0115] In Example 1, Robot A detects that the user is repeatedly moving between the living room and the balcony using a visual sensor, and simultaneously detects that there has been no voice interaction over a period of time using a voice module. Robot A uses a service state modeling module to adjust the service state vector, mapping "repeated movement" and "no voice interaction" to a decrease in the interaction continuity index and an increase in the safety risk awareness index in the service state vector. Thus, each robot can transform fuzzy, unstructured environmental and behavioral information into calculable and comparable quantifiable states.

[0116] Next, each robot in the multi-robot system can use the service state tension assessment module to compare its service state vector with the internally stored expected service state to determine whether the state tension has exceeded a threshold. For example, in the expected service state, the interaction continuity index should be higher than a certain threshold. When robot A calculates and finds that the interaction continuity index has been below the threshold for 5 minutes and the safety risk concern index has risen simultaneously, it determines that the state tension value calculated by the deviation between the service state and the expected service state has exceeded the limit, thus autonomously triggering a collaborative request. This achieves autonomous collaborative triggering driven by the internal service state of the multi-robot system without explicit external task instructions.

[0117] Once the collaboration condition is triggered, the collaboration state target publishing module of robot A, which first detected the state tension exceeding the limit, is activated. As the collaboration initiating robot, robot A's collaboration state target publishing module generates a descriptive message as a collaboration requirement. This collaboration requirement indicates the service state target, such as "Target: Increase the attention to user A to Lv.4 and reduce the safety risk of its surrounding environment to Lv.1".

[0118] Next, Robot A's collaborative state goal publishing module can broadcast this collaborative request, containing the service state goal, to other robots in the multi-robot system via a local network (such as Wi-Fi). Figure 4 In the illustrated embodiment, the multi-robot system may include multiple robots such as robot A, robot B, and robot C.

[0119] Then, robot A and other robots that receive the collaboration request can activate the state contribution assessment and collaboration negotiation module to achieve collaboration negotiation within the multi-robot system.

[0120] For example, based on its own location information and minimizing the distance to the user's location, Robot B assesses that its participation in the interaction can significantly improve the attention index; Robot C, based on its onboard high-definition panoramic camera, assesses that its role in environmental monitoring can effectively reduce the safety risk index. Each robot calculates its own predicted state contribution value, and through a distributed negotiation algorithm, a consensus is reached: Robot B will assume the role of active interaction, and Robot C will assume the role of safety monitoring, forming a temporary collaborative team. That is, the collaborative structure formed through negotiation involves Robot B and Robot C as the participating collaborative robots, with Robot B assigned the role of interactor and Robot C assigned the role of monitor.

[0121] The collaborative structure determined through consultation among all robots is not pre-specified or centrally allocated, but rather emerges autonomously through consultation based on the assessment of each node's contribution to the global state objective.

[0122] Then, based on the determined collaborative structure, each collaborative robot performs corresponding actions. In Implementation 1, Robot B begins to proactively approach the user and engage in voice interaction, while Robot C focuses its perception on the user's surrounding environment.

[0123] Meanwhile, as each collaborative robot performs its corresponding action, it continuously monitors changes in the service state vector. Once the user's emotions stabilize and they sit down to rest (indicated by the return of interaction continuity and safety risk awareness indicators to normal), the collaborative team automatically assesses that the state tension in the current state has been eliminated. At this point, the multi-robot collaborative system can adjust its collaborative structure, allowing it to evolve dynamically. Robot C may automatically exit, leaving only Robot B to provide light companionship.

[0124] Example 2 is an extension of Example 1. Based on Example 1, Example 2 introduces a collaborative effect history database to realize the collaborative memory function of the multi-robot collaborative system.

[0125] In Embodiment 2, the multi-robot collaborative system can record each historical collaborative event, including the first service state when the collaborative conditions are triggered in each historical collaborative event, the collaborative structure negotiated in the first service state, the second service state after the corresponding behavior is executed based on the collaborative structure, and the actual execution effect after collaborative execution.

[0126] When a service state similar to the first service state of a historical collaborative event in the collaborative effect history database appears, the robots will prioritize referencing the robot combinations with high contributions in the historical data during the joint negotiation and state contribution evaluation phases, thereby forming a collaborative structure more quickly and effectively. For example, records from a multi-robot collaborative system show that in the state of "an elderly person's anxious wandering," the collaborative group composed of robot B (good at emotional interaction) and robot D (good at playing soothing music) achieved the best improvement results, and in similar scenarios thereafter, this collaborative combination will be prioritized for formation.

[0127] In this way, by recording and learning the performance of different collaborative structures under specific conditions, the multi-robot collaborative system can continuously optimize the collaborative strategy formed over time, thereby continuously improving the overall service efficiency.

[0128] Please refer to Figure 4 , Figure 4 A flowchart of a multi-robot collaborative method based on service status triggering in a health and wellness care scenario, according to Embodiment 3 of the present invention, is shown.

[0129] like Figure 4 As shown, in the initial service state of Embodiment 3, each robot node in the multi-robot system continuously collects multimodal data of the environment and the user through its onboard multimodal sensors. Then, each robot node can run the service state modeling module based on the multimodal data, fusing and abstracting the multimodal data into a numerical current service state vector.

[0130] In this embodiment, the multimodal active perception module in the multi-robot collaborative system of robot R1 can adjust the service state vector by using the service state modeling module when it detects that the elderly frequently get up, speak less, or walk unsteadily. The detected situation of the elderly frequently getting up, speaking less, or walking unsteadily is mapped to an increase in the safety risk attention index and a decrease in the interaction continuity index in the vector.

[0131] Next, each robot in the multi-robot system can use the service state tension assessment module to compare its service state vector with the internally stored desired service state to determine whether the state tension has exceeded a threshold. Robot R1's service state tension assessment module can compare its service state vector with the desired state to determine if the state tension exceeds the threshold. When the state tension exceeds the threshold, it indicates that multi-robot collaboration needs to be triggered; otherwise, the current service state is maintained.

[0132] When the cooperative condition is triggered, the cooperative state target publishing module of robot R1, which first detected the state tension exceeding the limit, is activated. For example... Figure 4 As shown, the service status objective for this non-specific task can be: "Objective G: Improve safety and maintain continuity of care".

[0133] Robot R1's collaborative state goal publishing module can broadcast collaborative requests containing service state goals to other robots in a multi-robot system. Figure 4 In the illustrated embodiment, the multi-robot system may include multiple robots such as robot R1, robot R2, and robot R3.

[0134] Then, robot R1 and other robots that receive the collaboration request can activate the state contribution assessment and collaboration negotiation module to achieve collaboration negotiation within the multi-robot system. Through a distributed negotiation algorithm, the negotiation structure is determined: robot R1 assumes the role of caregiver, accompanying and interacting with the elderly; robot R2 assumes the role of environmental inspector, inspecting the environment to reduce safety risks; and robot R3 assumes the role of environmental regulator, adjusting lighting, temperature, and comfort levels, thus forming a temporary collaboration group. That is, the collaborative structure formed through negotiation involves robots R1, R2, and R3, with robot R1 assigned the role of caregiver, robot R2 the role of environmental inspector, and robot R3 the role of environmental regulator.

[0135] Then, based on the determined collaborative structure, each collaborative robot performs corresponding actions. In Implementation 3, robot R1 can approach the elderly person to provide companionship, robot R2 can inspect for obstacles on the ground, and robot R3 can adjust the lighting and air conditioning.

[0136] Meanwhile, as each collaborative robot performs its corresponding actions, it continuously monitors changes in the service state vector. When the safety concern index decreases and the interaction continuity index recovers, it indicates an improvement in the service state. The multi-robot collaborative system can adjust its collaborative structure, allowing it to evolve dynamically.

[0137] Therefore, the service state-triggered multi-robot collaboration method provided by this invention is particularly suitable for continuous, long-term accompanying service scenarios. This service state-triggered multi-robot collaboration method focuses on the continuous maintenance and optimization of service state, rather than the sequential completion of discrete tasks, which is highly compatible with application scenarios such as elderly care companionship and intelligent tour guide, which require long-term monitoring of the user's overall state.

[0138] In summary, through the organic combination of the various steps of this invention, this invention achieves outstanding overall and groundbreaking beneficial effects compared to existing multi-robot collaborative technologies.

[0139] This invention breaks through the limitations of the traditional task-driven paradigm. The multi-robot collaborative system enables collaboration to be triggered even when the task instructions are not yet clear and the user needs are only initially apparent. It fundamentally solves the problem of collaboration lag in existing technologies when dealing with open and progressive service needs, and realizes a paradigm shift from "passively responding to tasks" to "actively maintaining the state".

[0140] This invention enables the formation of a decentralized, self-organizing collaborative structure. The multi-robot system does not rely on a central control node, nor does it require predefined, complex task allocation rules. The collaborative structure emerges autonomously through distributed negotiation based on the state contributions of each robot node. This not only improves the reliability and resilience of the multi-robot system (no single point of failure) but also greatly enhances its scalability, making the addition of new robot nodes much easier.

[0141] This invention significantly enhances the overall adaptability in complex and dynamic scenarios. It is particularly suitable for open scenarios requiring long-term, continuous interaction, such as elderly care companionship, intelligent navigation, and public services. It possesses a natural adaptability to the ambiguous needs, unstructured environments, and dynamic changes inherent in these scenarios, enabling it to provide more coherent and contextualized services.

[0142] This invention also enables the long-term evolution of multi-robot systems and continuous optimization of user experience. Thanks to the dynamic evolution mechanism of the collaborative structure and the optimized collaborative memory learning mechanism, the multi-robot collaborative system of this invention can continuously self-adjust and optimize throughout its entire lifecycle. This not only ensures the long-term stability and efficiency of the multi-robot system but also allows the service experience received by users to become increasingly improved and intelligent over time.

[0143] Although the methods described above are illustrated and depicted as a series of actions for the sake of simplicity, it should be understood and appreciated that these methods are not limited by the order of the actions, as some actions may occur in a different order and / or concurrently with other actions from the illustrations and descriptions herein or not illustrated and described herein but which may be understood by those skilled in the art, according to one or more embodiments.

[0144] Those skilled in the art will understand that information, signals, and data can be represented using any of a variety of different techniques and skills. For example, the data, instructions, commands, information, signals, bits, symbols, and chips described throughout the above description can be represented by voltage, current, electromagnetic waves, magnetic fields or magnetic particles, light fields or optical particles, or any combination thereof.

[0145] Those skilled in the art will further appreciate that the various illustrative logic blocks, modules, circuits, and algorithm steps described in conjunction with the embodiments disclosed herein can be implemented as electronic hardware, computer software, or a combination of both. To clearly illustrate this interchangeability between hardware and software, the various illustrative components, blocks, modules, circuits, and steps are described above in a generalized manner in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the specific application and the design constraints imposed on the overall system. Those skilled in the art may implement the described functionality in different ways for each specific application, but such implementation decisions should not be construed as departing from the scope of the invention.

[0146] The various illustrative logic modules and circuits described in conjunction with the embodiments disclosed herein may be implemented or performed using a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. The general-purpose processor may be a microprocessor, but in alternatives, it may be any conventional processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors cooperating with a DSP core, or any other such configuration.

[0147] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of both. The software module may reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium is coupled to a processor such that the processor can read and write information to / from the storage medium. In an alternative, the storage medium may be integrated into the processor. The processor and storage medium may reside in an ASIC. The ASIC may reside in a user terminal. In an alternative, the processor and storage medium may reside as discrete components in the user terminal.

[0148] In one or more exemplary embodiments, the described functionality may be implemented in hardware, software, firmware, or any combination thereof. If implemented in software as a computer program product, the functionality may be stored or transmitted as one or more instructions or code on or through a computer-readable medium. A computer-readable medium includes both computer storage media and communication media, encompassing any medium that facilitates the transfer of a computer program from one location to another. A storage medium may be any available medium accessible to a computer. By way of example and not limitation, such a computer-readable medium may include RAM, ROM, EEPROM, CD-ROM or other optical disc storage, disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and is accessible to a computer. Any connection is also legitimately referred to as a computer-readable medium. For example, if the software is transmitted from a website, server, or other remote source using coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave, then the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of a medium. As used in this article, disk and disc include compact discs (CDs), laser discs, optical discs, digital multi-purpose discs (DVDs), floppy disks, and Blu-ray discs. Disks typically reproduce data magnetically, while discs reproduce data optically using lasers. Combinations of these should also be included within the scope of computer-readable media.

[0149] The prior description of this disclosure is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to this disclosure will be apparent to those skilled in the art, and the general principles defined herein may be applied to other variations without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not intended to be limited to the examples and designs described herein, but should be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for service state trigger-based multi-robot coordination, characterized in that, The method comprises steps of: continuously perceiving environment state and user state information through multi-modal sensors to construct a service state vector for quantitatively describing interaction conditions among the user, the environment and the robot; in response to a deviation between the service state vector and an expected service state satisfying a collaborative trigger condition, generating a collaborative demand indicating a target service state and broadcasting the collaborative demand to other robots; based on the collaborative demand, jointly negotiating with the other robots to determine a collaborative structure, the collaborative structure including collaborative robots participating in execution and role allocation of the collaborative robots; and continuously monitoring service state vectors when the collaborative robots execute corresponding behaviors based on the role allocation.

2. The multi-robot collaboration method of claim 1, wherein, The joint negotiation is implemented through a bidding negotiation mechanism.

3. The multi-robot collaboration method of claim 2, wherein, The multi-robot collaborative method further comprises steps of: listening to broadcasts of the other robots to receive the collaborative demand; and in response to reception of the collaborative demand, participating in the joint negotiation according to a state contribution degree to the collaborative demand to jointly determine the collaborative structure of the collaborative demand with the other robots.

4. The multi-robot collaboration method of claim 1, wherein, The step of generating the collaborative demand indicating the target service state and broadcasting the collaborative demand to the other robots in response to the deviation between the service state vector and the expected service state satisfying the collaborative trigger condition comprises: calculating the deviation between the service state vector and the expected service state as a state tension value; and generating the collaborative demand and broadcasting the collaborative demand to the other robots when the state tension value exceeds a preset threshold.

5. The multi-robot collaboration method of claim 4, wherein, The step of generating the collaborative demand indicating the target service state and broadcasting the collaborative demand to the other robots in response to the deviation between the service state vector and the expected service state satisfying the collaborative trigger condition further comprises: when the state tension value exceeds the preset threshold, inputting service state vector data in a first time range into a world model to predict a service state curve in a future second time range; in response to the service state curve in the second time range indicating deterioration of the service state, generating the collaborative demand and broadcasting the collaborative demand to the other robots.

6. The multi-robot collaboration method of claim 1, wherein, The service state vector comprises a safety risk attention parameter and an interaction activity parameter.

7. The multi-robot collaboration method of claim 1, wherein, The multi-robot collaborative method further comprises steps of: recording a historical collaborative event in a collaborative effect history library, the historical collaborative event including a first service state vector when the collaborative trigger condition is triggered, the collaborative structure formed by negotiation under the first service state vector, and a second service state vector after corresponding behaviors are executed based on the collaborative structure; and in response to the deviation between the service state vector and the expected service state satisfying the collaborative trigger condition and the service state vector being similar to the first service state vector, preferentially referring to the collaborative structure formed by negotiation under the first service vector when the joint negotiation is performed. The step of continuously monitoring service state vectors when the collaborative robots execute corresponding behaviors based on the role allocation comprises:

8. The multi-robot collaboration method of claim 7, wherein, in response to the service state vector when the collaborative robots execute corresponding behaviors based on the role allocation reaching a target service state, dissolving the collaborative structure. ​ 9. A service state trigger based multi-robot coordination system, characterized in that, comprising: a memory having computer instructions stored thereon; and a processor connected to the memory and configured to execute the computer instructions stored on the memory to implement the method of claim 1-8.

10. A service state trigger based multi-robot system, characterized in that, comprising: a plurality of robots configured with the system of claim 9.

Images