A behavior tree modeling method, device and equipment supporting robot decision reuse

Abstract

The present application relates to a behavior tree modeling method, apparatus, and device that supports robot decision reuse. The method includes: constructing a decision library for robot software, selecting five mainstream decision modes from the decision library based on prior knowledge; the decision modes include prudent decision-making, reactive decision-making, hybrid decision-making, online probabilistic decision-making, and offline probabilistic decision-making; extracting three common behavior interaction modes from the decision modes based on the behavioral abstraction of perception behavior, effector behavior, and planning behavior, as well as the interactive relationship between behaviors; using the behavior interaction modes, selecting sequence control nodes and two action nodes in the behavior tree, and modeling with a sequence subtree structure as the basic model framework of the behavior interaction mode to obtain a behavior tree that supports robot decision reuse. This method can improve the reuse rate of robot decisions.

Description

Technical Field

[0001] The present application relates to the field of modeling technology, and in particular to a behavior tree modeling method, apparatus, computer equipment, and storage medium that support robot decision reuse. Background Art

[0002] Today, autonomous robots are expected to complete more complex tasks and operate increasingly in uncertain, open environments. Software design and development for such robots often involves two aspects: decision-making algorithm design and software architecture design. Because robot task decisions in open environments involve complex interactions and scheduling of perception, action, and effector behaviors, designing decision-making algorithms for such software becomes more challenging, which in turn increases the complexity of component functionality design and development for the underlying supporting software architecture. In recent years, software engineering reuse concepts and techniques have been gradually introduced into the robotics field, providing effective solutions for improving the efficiency of robotics software engineering development.

[0003] However, the current component-based robot software development method only provides reusable models and interfaces from the perspective of basic communication and control logic. Few works provide reusable design of complex decision-making mechanisms, making the current robot software development still inefficient and difficult. Summary of the Invention

[0004] Based on this, it is necessary to provide a behavior tree modeling method, device, computer equipment and storage medium that support robot decision reuse and can improve the reuse rate of robot decisions to address the above technical problems.

[0005] A behavior tree modeling method supporting robot decision reuse, the method comprising:

[0006] Build a decision library for the robot software and select five mainstream decision-making models from the library based on prior knowledge; the decision-making models include prudent decision-making, reactive decision-making, hybrid decision-making, online probabilistic decision-making, and offline probabilistic decision-making;

[0007] Based on the behavioral abstraction of perception behavior, effector behavior, and planning behavior, as well as the interactive relationship between behaviors, three common behavioral interaction patterns are extracted from the decision-making pattern. The sequence control node and two action nodes in the behavior tree are selected using the behavioral interaction pattern, and the sequence subtree structure is used as the basic model framework for modeling the behavior interaction pattern. A behavior tree model that supports robot decision reuse is obtained.

[0008] In one embodiment, the process of building a decision library for robot software includes: utilizing the massive software project resources gathered by the robot open source community on the Internet, and storing all software projects related to robot task decisions into the decision library through keyword matching and analysis.

[0009] In one embodiment, the three general behavioral interaction modes include a perception-effect behavioral interaction mode, a perception-planning behavioral interaction mode, and a planning-effect behavioral interaction mode.

[0010] In one embodiment, a behavior tree model supporting robot decision reuse is obtained by selecting a sequence control node and two action nodes in a behavior tree using a behavior interaction pattern and using a sequence subtree structure as a basic model framework for the behavior interaction pattern. The model includes:

[0011] The perception-planning behavior interaction model is used to select the sequence control node and two action nodes in the behavior tree. The sequence subtree structure is used as the basic model framework of the perception-planning behavior interaction model, and the left and right child nodes in the sequence subtree structure are implemented as the perception action node and the planning action node respectively.

[0012] An interface for interacting with robot sensors is provided in the perception-action node, allowing customized sensor drivers to serve as internal node programs. The classic planning algorithm POPF is embedded in the planning-action node as a general planning algorithm in a sequential subtree structure, resulting in a reusable behavior tree model for the perception-planning behavior interaction pattern.

[0013] In one embodiment, a sequence control node and two action nodes in a behavior tree are selected using a perception-effect behavior interaction model. A sequence subtree structure is used as the basic model framework of the perception-effect behavior interaction model, and the left and right child nodes in the sequence subtree structure are implemented as a perception action node and an effect action node, respectively.

[0014] An interaction interface with the robot sensor is provided in the perception action node, allowing customized sensor drivers to be used as node internal programs. An interaction interface with the robot effector is provided in the effect action node, allowing customized effector drivers to be used as node internal programs, thus obtaining a reusable behavior tree model of the perception-effect behavior interaction pattern.

[0015] In one embodiment, a sequence control node and two action nodes in a behavior tree are selected using a plan-effect behavior interaction model. The sequence subtree structure is used as the basic model framework of the plan-effect behavior interaction model. The left and right child nodes in the sequence subtree structure are implemented as a plan action node and an effect action node, respectively. The control logic of the sequence control node is modified, and a loop control condition is added, thereby changing the internal control logic of the sequence control node from single sequence control to loop sequence control.

[0016] The classic planning algorithm POPF is embedded in the planning action node as a general planning algorithm in the sequence subtree structure. An interaction interface with the robot effector is provided in the effector action node, allowing customized effector drivers to be used as node internal programs, thus obtaining a reusable behavior tree model of the planning-effector behavior interaction pattern.

[0017] A behavior tree modeling device supporting robot decision reuse, the device comprising:

[0018] The decision-making model selection module is used to build the decision library of the robot software. It selects five mainstream decision-making models from the decision library based on prior knowledge. The decision-making models include prudent decision-making, reactive decision-making, hybrid decision-making, online probabilistic decision-making, and offline probabilistic decision-making.

[0019] The behavior tree modeling module is used to extract three common behavior interaction patterns from the decision-making model based on the behavioral abstraction of perception behavior, effector behavior, and planning behavior, as well as the interaction relationship between behaviors. The behavior interaction pattern is used to select the sequence control node and two action nodes in the behavior tree and model the sequence subtree structure as the basic model framework of the behavior interaction pattern, thereby obtaining a behavior tree model that supports robot decision reuse.

[0020] A computer device includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0021] Build a decision library for the robot software and select five mainstream decision-making models from the library based on prior knowledge; the decision-making models include prudent decision-making, reactive decision-making, hybrid decision-making, online probabilistic decision-making, and offline probabilistic decision-making;

[0022] Based on the behavioral abstraction of perception behavior, effector behavior, and planning behavior, as well as the interactive relationship between behaviors, three common behavioral interaction patterns are extracted from the decision-making pattern. The sequence control node and two action nodes in the behavior tree are selected using the behavioral interaction pattern, and the sequence subtree structure is used as the basic model framework for modeling the behavior interaction pattern. A behavior tree model that supports robot decision reuse is obtained.

[0023] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the following steps:

[0024] Build a decision library for the robot software and select five mainstream decision-making models from the library based on prior knowledge; the decision-making models include prudent decision-making, reactive decision-making, hybrid decision-making, online probabilistic decision-making, and offline probabilistic decision-making;

[0025] Based on the behavioral abstraction of perception behavior, effector behavior, and planning behavior, as well as the interactive relationship between behaviors, three common behavioral interaction patterns are extracted from the decision-making pattern. The sequence control node and two action nodes in the behavior tree are selected using the behavioral interaction pattern, and the sequence subtree structure is used as the basic model framework for modeling the behavior interaction pattern. A behavior tree model that supports robot decision reuse is obtained.

[0026] The aforementioned behavior tree modeling method, apparatus, and device for supporting robot decision reuse reshape the behavior tree model, integrating the generalized and universal algorithmic flow of the decision algorithm with the control logic of the behavior tree structure. This makes the behavior tree model a reusable model with low coupling and high cohesion. Software engineers can flexibly combine these behavior tree structures to construct the desired robot software system, thereby increasing the reuse rate of robot decisions, the reuse rate of software components, and software development efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is an application scenario diagram of a behavior tree modeling method supporting robot decision reuse in one embodiment;

[0028] Figure 2 A structural block diagram of a behavior tree modeling device supporting robot decision reuse in one embodiment;

[0029] Figure 3 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0031] In one embodiment, Figure 1 As shown in FIG, a behavior tree modeling method supporting robot decision reuse is provided, including the following steps:

[0032] Step 102, build a decision library for the robot software, and select five mainstream decision modes from the decision library based on prior knowledge; the decision modes include prudent decision-making, reactive decision-making, hybrid decision-making, online probabilistic decision-making, and offline probabilistic decision-making.

[0033] First, a decision library for the robot software is constructed, and all software projects related to the robot task decision are stored in the decision library. Through manual interpretation and analysis, five mainstream decision algorithm modes are summarized from the decision library, namely: prudent decision-making, reactive decision-making, hybrid decision-making, online probabilistic decision-making, and offline probabilistic decision-making. The above decisions are existing technologies and will not be described in detail in this application.

[0034] Step 104: Based on the behavioral abstractions of perception behavior, effector behavior, and planning behavior, as well as the interaction relationships between behaviors, three common behavioral interaction patterns are extracted from the decision-making pattern. Using the behavioral interaction patterns, a sequence control node and two action nodes in the behavior tree are selected to model the behavior interaction pattern using a sequence subtree structure as the basic model framework, thereby obtaining a behavior tree model that supports robot decision reuse.

[0035] This application is based on five decision-making algorithm patterns. According to the behavioral abstraction of "perception behavior, effect behavior, and planning behavior" and the interaction relationship between behaviors, three general behavioral interaction patterns are extracted, namely: "perception-effect behavior" interaction pattern, "perception-planning behavior" interaction pattern, and "planning-effect behavior" interaction pattern. The behavior interaction pattern is then used to select the sequence control node and two action nodes in the behavior tree, and the sequence subtree structure is used as the basic model framework for modeling the behavior interaction pattern to obtain a behavior tree model that supports robot decision reuse. First, according to the behavior abstraction and the interaction relationship between behaviors, the general behavior interaction pattern is extracted and integrated with the control logic of the behavior tree, so that the behavior tree model becomes a low-coupling, high-cohesion reusable model. The software system provides model assembly rules based on the three types of reusable behavior tree models developed in the previous step, so that users can flexibly select two of the tree structures and integrate them to construct a complete software decision structure, thereby developing a robot software through software component reuse.

[0036] In the aforementioned behavior tree modeling method that supports robot decision reuse, the present invention remodels the behavior tree model, integrating the generalized and universal algorithmic flow of the decision algorithm with the control logic of the behavior tree structure. This makes the behavior tree model a low-coupling, highly cohesive, reusable model. Software engineers can flexibly combine these behavior tree structures to construct the desired robot software system, thereby increasing the reuse rate of robot decisions, the reuse rate of software components, and software development efficiency.

[0037] In one embodiment, the process of building a decision library for robot software includes: utilizing the massive software project resources gathered by the robot open source community on the Internet, and storing all software projects related to robot task decisions into the decision library through keyword matching and analysis.

[0038] In one embodiment, the three general behavioral interaction modes include a perception-effect behavioral interaction mode, a perception-planning behavioral interaction mode, and a planning-effect behavioral interaction mode.

[0039] In one embodiment, a behavior tree model supporting robot decision reuse is obtained by selecting a sequence control node and two action nodes in a behavior tree using a behavior interaction pattern and using a sequence subtree structure as a basic model framework for the behavior interaction pattern. The model includes:

[0040] The perception-planning behavior interaction model is used to select the sequence control node and two action nodes in the behavior tree. The sequence subtree structure is used as the basic model framework of the perception-planning behavior interaction model, and the left and right child nodes in the sequence subtree structure are implemented as the perception action node and the planning action node respectively.

[0041] An interface for interacting with robot sensors is provided in the perception-action node, allowing customized sensor drivers to serve as internal node programs. The classic planning algorithm POPF is embedded in the planning-action node as a general planning algorithm in a sequential subtree structure, resulting in a reusable behavior tree model for the perception-planning behavior interaction pattern.

[0042] In a specific embodiment, based on the "perception-planning-behavior" interaction model, a sequence control node and two action nodes are selected from the behavior tree, and a sequence subtree structure is used as the basic model framework for this interaction model. The left and right subnodes in this subtree structure are implemented as a "perception-action node" and a "planning-action node," respectively. An interface for interacting with the robot's sensors is provided in the "perception-action node," allowing customized sensor drivers to serve as internal node programs. The classic planning algorithm POPF is embedded within the "planning-action node" as a general planning algorithm within this subtree structure. This substructure serves as a specific implementation model for the "perception-planning-behavior" interaction model and also as a reusable software component model.

[0043] In one embodiment, a sequence control node and two action nodes in a behavior tree are selected using a perception-effect behavior interaction model. A sequence subtree structure is used as the basic model framework of the perception-effect behavior interaction model, and the left and right child nodes in the sequence subtree structure are implemented as a perception action node and an effect action node, respectively.

[0044] An interaction interface with the robot sensor is provided in the perception action node, allowing customized sensor drivers to be used as node internal programs. An interaction interface with the robot effector is provided in the effect action node, allowing customized effector drivers to be used as node internal programs, thus obtaining a reusable behavior tree model of the perception-effect behavior interaction pattern.

[0045] In a specific embodiment, based on the "perception-effect behavior" interaction model, a sequence control node and two action nodes are selected from the behavior tree, and a sequence subtree structure is used as the basic model framework for this interaction model. The left and right subnodes in this subtree structure are implemented as "perception action nodes" and "effect action nodes," respectively. The "perception action node" provides an interface for interacting with the robot's sensors, allowing customized sensor drivers to be implemented as internal node programs. The "effect action node" provides an interface for interacting with the robot's effectors, allowing customized effector drivers to be implemented as internal node programs. The above substructure serves as a specific implementation model for the "perception-effect behavior" interaction model and also as a reusable software component model.

[0046] In one embodiment, a sequence control node and two action nodes in a behavior tree are selected using a plan-effect behavior interaction model. The sequence subtree structure is used as the basic model framework of the plan-effect behavior interaction model. The left and right child nodes in the sequence subtree structure are implemented as a plan action node and an effect action node, respectively. The control logic of the sequence control node is modified, and a loop control condition is added, thereby changing the internal control logic of the sequence control node from single sequence control to loop sequence control.

[0047] The classic planning algorithm POPF is embedded in the planning action node as a general planning algorithm in the sequence subtree structure. An interaction interface with the robot effector is provided in the effector action node, allowing customized effector drivers to be used as node internal programs, thus obtaining a reusable behavior tree model of the planning-effector behavior interaction pattern.

[0048] In a specific embodiment, based on the "planning-effect behavior" interaction mode, a sequence control node and two action nodes in the behavior tree are selected, and a sequence subtree structure is used as the basic model framework of the interaction mode. The left and right subnodes in the subtree structure are respectively implemented as "planning action nodes" and "effect action nodes". First, the control logic of the sequence control node is transformed, and loop control conditions are added to change the internal control logic of the sequence control node from "single sequence control" to "loop sequence control". The classic planning algorithm POPF is embedded in the "planning action node" as a general planning algorithm in the subtree structure. An interaction interface with the robot effector is provided in the "effect action node", allowing customized effector drivers to be used as node internal programs. The above substructure serves as a specific implementation model of the "planning-effect behavior" interaction mode, and is also a reusable software component model.

[0049] It should be understood that although Figure 1The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. In addition, Figure 1 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0050] In one embodiment, Figure 2 As shown, a behavior tree modeling device supporting robot decision reuse is provided, including: a decision mode selection module 202 and a behavior tree modeling module 204, wherein:

[0051] The decision mode selection module 202 is used to build a decision library for the robot software and select five mainstream decision modes from the decision library based on prior knowledge; the decision modes include prudent decision-making, reactive decision-making, hybrid decision-making, online probabilistic decision-making, and offline probabilistic decision-making;

[0052] The behavior tree modeling module 204 is used to extract three common behavior interaction patterns from the decision model based on the behavioral abstractions of perception behavior, effector behavior, and planning behavior, as well as the interactive relationships between behaviors. Using the behavior interaction patterns, the module selects the sequence control node and two action nodes in the behavior tree, using the sequence subtree structure as the basic model framework for modeling the behavior interaction patterns, thereby obtaining a behavior tree that supports robot decision reuse.

[0053] In one embodiment, the behavior tree modeling module 204 is further configured to use the behavior interaction pattern to select a sequence control node and two action nodes in the behavior tree and model the sequence subtree structure as the basic model framework of the behavior interaction pattern to obtain a behavior tree that supports robot decision reuse, including:

[0054] The perception-planning behavior interaction model is used to select the sequence control node and two action nodes in the behavior tree. The sequence subtree structure is used as the basic model framework of the perception-planning behavior interaction model, and the left and right child nodes in the sequence subtree structure are implemented as the perception action node and the planning action node respectively.

[0055] An interface for interacting with robot sensors is provided in the perception-action node, allowing customized sensor drivers to serve as internal node programs. The classic planning algorithm POPF is embedded in the planning-action node as a general planning algorithm in a sequential subtree structure, resulting in a reusable behavior tree model for the perception-planning behavior interaction pattern.

[0056] In one embodiment, the behavior tree modeling module 204 is further configured to select a sequence control node and two action nodes in the behavior tree using a perception-effect behavior interaction model, using a sequence subtree structure as a basic model framework for the perception-effect behavior interaction model, and implementing the left and right child nodes in the sequence subtree structure as a perception-action node and an effector-action node, respectively.

[0057] An interaction interface with the robot sensor is provided in the perception action node, allowing customized sensor drivers to be used as node internal programs. An interaction interface with the robot effector is provided in the effect action node, allowing customized effector drivers to be used as node internal programs, thus obtaining a reusable behavior tree model of the perception-effect behavior interaction pattern.

[0058] In one embodiment, the behavior tree modeling module 204 is further configured to utilize the plan-effect behavior interaction model to select a sequence control node and two action nodes in the behavior tree, using the sequence subtree structure as the basic model framework for the plan-effect behavior interaction model. The left and right child nodes in the sequence subtree structure are implemented as a plan action node and an effect action node, respectively. The control logic of the sequence control node is modified, and a loop control condition is added to change the internal control logic of the sequence control node from single sequence control to loop sequence control.

[0059] The classic planning algorithm POPF is embedded in the planning action node as a general planning algorithm in the sequence subtree structure. An interaction interface with the robot effector is provided in the effector action node, allowing customized effector drivers to be used as node internal programs, thus obtaining a reusable behavior tree model of the planning-effector behavior interaction pattern.

[0060] The specific definitions of a behavior tree modeling device supporting robot decision reuse can be found in the definitions of a behavior tree modeling method supporting robot decision reuse above and will not be elaborated here. Each module in the aforementioned behavior tree modeling device supporting robot decision reuse can be implemented in whole or in part via software, hardware, or a combination thereof. Each of these modules can be embedded in or independent of a processor in a computer device in hardware form, or stored in a computer device memory in software form, allowing the processor to call and execute the corresponding operations of each module.

[0061] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 3As shown. The computer device includes a processor, memory, network interface, display screen and input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a behavior tree modeling method that supports robot decision reuse is implemented. The display screen of the computer device can be a liquid crystal display or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad provided on the computer device housing, or an external keyboard, touchpad or mouse.

[0062] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0063] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0064] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0065] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art could make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present patent application shall be determined by the appended claims.

Claims

1. A behavior tree modeling method that supports robot decision reuse, characterized in that: The method comprises: Build a decision library for the robot software and select five mainstream decision modes from the library based on prior knowledge; the decision modes include prudent decision-making, reactive decision-making, hybrid decision-making, online probabilistic decision-making, and offline probabilistic decision-making; Based on the behavioral abstraction of perception behavior, effector behavior, and planning behavior, as well as the interaction relationship between behaviors, three common behavioral interaction patterns are extracted from the decision-making model. Sequence control nodes and two action nodes in the behavior tree are selected based on the behavioral interaction patterns to model the behavior interaction patterns using a sequence subtree structure as the basic model framework, thereby obtaining a behavior tree model that supports robot decision reuse. The three general behavioral interaction modes include perception-effect behavior interaction mode, perception-planning behavior interaction mode and planning-effect behavior interaction mode; The behavior interaction pattern is used to select a sequence control node and two action nodes in the behavior tree and model the sequence subtree structure as the basic model framework of the behavior interaction pattern. A behavior tree model that supports robot decision reuse is obtained, including: The perception-planning behavior interaction model is used to select a sequence control node and two action nodes in the behavior tree, and a sequence subtree structure is used as the basic model framework of the perception-planning behavior interaction model. The left and right child nodes in the sequence subtree structure are respectively implemented as a perception action node and a planning action node; An interface for interacting with robot sensors is provided in the perception-action node, allowing customized sensor drivers to be used as internal node programs. The classic planning algorithm POPF is embedded in the planning-action node as a general planning algorithm in the sequence subtree structure, resulting in a reusable behavior tree model for the perception-planning behavior interaction pattern. The perception-effect behavior interaction model is used to select a sequence control node and two action nodes in the behavior tree, and a sequence subtree structure is used as the basic model framework of the perception-effect behavior interaction model. The left and right child nodes in the sequence subtree structure are respectively implemented as a perception action node and an effect action node; The perception-action node provides an interface for interacting with the robot's sensors, allowing customized sensor drivers to be used as internal node programs. The effector-action node provides an interface for interacting with the robot's effectors, allowing customized effector drivers to be used as internal node programs. This results in a reusable behavior tree model for the perception-effect behavior interaction pattern. Utilizing the plan-effect behavior interaction model, a sequence control node and two action nodes are selected from the behavior tree. The sequence subtree structure is used as the basic model framework of the plan-effect behavior interaction model. The left and right child nodes in the sequence subtree structure are implemented as a plan action node and an effect action node, respectively. The control logic of the sequence control node is modified, and a loop control condition is added to change the internal control logic of the sequence control node from single sequence control to loop sequence control. The classical planning algorithm POPF is embedded in the planning action node as a universal planning algorithm in the sequence subtree structure. An interaction interface with the robot effector is provided in the effector action node, allowing a customized effector driver to be used as an internal program of the node, thereby obtaining a reusable behavior tree model of the planning-effector behavior interaction mode.

2. The method according to claim 1, characterized in that The process of building a decision library for robot software includes: By utilizing the massive software project resources gathered by the robot open source community on the Internet, all software projects related to robot task decision-making are stored in the decision library through keyword matching and analysis.

3. A behavior tree modeling device supporting robot decision reuse, characterized in that: The device comprises: A decision-making mode selection module is used to build a decision library for the robot software and select five mainstream decision-making modes from the library based on prior knowledge; the decision-making modes include prudent decision-making, reactive decision-making, hybrid decision-making, online probabilistic decision-making, and offline probabilistic decision-making; A behavior tree modeling module is configured to extract three common behavior interaction patterns from the decision model based on the behavioral abstractions of perception behavior, effector behavior, and planning behavior and the interaction relationships between the behaviors, and to use the behavior interaction patterns to select a sequence control node and two action nodes in the behavior tree and model them using a sequence subtree structure as a basic model framework for the behavior interaction patterns, thereby obtaining a behavior tree model that supports robot decision reuse. The three common behavior interaction patterns include a perception-effector behavior interaction pattern, a perception-planning behavior interaction pattern, and a planning-effector behavior interaction pattern. The behavior interaction patterns are used to select a sequence control node and two action nodes in the behavior tree and model them using a sequence subtree structure as a basic model framework for the behavior interaction patterns, thereby obtaining a behavior tree model that supports robot decision reuse, including: The perception-planning behavior interaction model is used to select a sequence control node and two action nodes in the behavior tree, and a sequence subtree structure is used as the basic model framework of the perception-planning behavior interaction model. The left and right child nodes in the sequence subtree structure are respectively implemented as a perception action node and a planning action node; An interface for interacting with robot sensors is provided in the perception-action node, allowing customized sensor drivers to be used as internal node programs. The classic planning algorithm POPF is embedded in the planning-action node as a general planning algorithm in the sequence subtree structure, resulting in a reusable behavior tree model for the perception-planning behavior interaction pattern. The perception-effect behavior interaction model is used to select a sequence control node and two action nodes in the behavior tree, and a sequence subtree structure is used as the basic model framework of the perception-effect behavior interaction model. The left and right child nodes in the sequence subtree structure are respectively implemented as a perception action node and an effect action node; The perception-action node provides an interface for interacting with the robot's sensors, allowing customized sensor drivers to be used as internal node programs. The effector-action node provides an interface for interacting with the robot's effectors, allowing customized effector drivers to be used as internal node programs. This results in a reusable behavior tree model for the perception-effect behavior interaction pattern. Utilizing the plan-effect behavior interaction model, a sequence control node and two action nodes are selected from the behavior tree. The sequence subtree structure is used as the basic model framework of the plan-effect behavior interaction model. The left and right child nodes in the sequence subtree structure are implemented as a plan action node and an effect action node, respectively. The control logic of the sequence control node is modified, and a loop control condition is added to change the internal control logic of the sequence control node from single sequence control to loop sequence control. The classical planning algorithm POPF is embedded in the planning action node as a universal planning algorithm in the sequence subtree structure. An interaction interface with the robot effector is provided in the effector action node, allowing a customized effector driver to be used as an internal program of the node, thereby obtaining a reusable behavior tree model of the planning-effector behavior interaction mode.

4. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 2 are implemented.

Images