A mind map-based pddl problem calculation method and system

Abstract

The application discloses a mind map-based PDDL problem calculation method and system. The method comprises the following steps: obtaining a domain.pddl file based on a PDDL problem to be edited, and determining component information required for editing a PDDL problem corresponding to the domain.pddl file according to the domain.pddl file; performing mind map editing on the PDDL problem by using the component information; parsing the mind map to obtain problem.pddl information; generating a problem.pddl file according to the problem.pddl information and a domain corresponding to the PDDL problem; and obtaining a solution of the PDDL problem based on the domain.pddl file and the problem.pddl file. The application can provide convenience for users to solve planning problems by using PDDL.

Description

Technical Field

[0001] This invention belongs to the field of computer software engineering technology and the field of automatic planning and scheduling in artificial intelligence; it relates to the scenario of editing and solving PDDL problems in real time on a web page or locally, and proposes a PDDL problem calculation method and system based on mind maps. Background Technology

[0002] "AI Planning" is a hallmark of intelligence and a branch of artificial intelligence. It aims to automate reasoning by combining "AI knowledge representation" and "AI computing technology." To automatically solve planning problems using AI Planning, users first need to represent the problem in a way that AI can understand. In academia, there is a specialized language for representing AI Planning problems called PDDL (Planning Domain Definition Language).

[0003] The domain model built using PDDL mainly consists of two parts:

[0004] • domain.pddl: Domain experts predefine the predicates and actions in this domain using PDDL text for user reference.

[0005] problem.pddl: Based on the predicate and actions defined in domain.pddl, the user defines the objects, initial state, and goal state corresponding to the specific problem they want to solve.

[0006] Once the user represents the problem in the form of problem.pddl based on domain.pddl, both domain.pddl and problem.pddl can be submitted to the AI ​​Planning solver for solution. The solver will eventually return the solution to the problem in a user-readable format.

[0007] With the current surge in AI development, an increasing number of researchers in planning and scheduling fields need to utilize Programming Processing DL (PDDL) to represent and solve automated planning problems. Although PDDL now has officially defined specifications and numerous related tutorials, users still encounter some issues when using PDDL to represent planning problems. These can be summarized as follows:

[0008] 1. Currently, the mainstream method for editing and saving PDDL files is in text file format (.pddl). However, the hierarchical structure of PDDL files in text format is not intuitive and has poor readability. Furthermore, editing PDDL files in text format requires attention to formatting compliance, increasing the difficulty and cost for users.

[0009] 2. Currently, commonly used PDDL solvers require users to manually upload the problem.pddl and domain.pddl files to the solver before receiving the solution to the planning problem, making real-time solving difficult. If users need to continuously verify the solvability of the problem while editing it, this method wastes a significant amount of time.

[0010] 3. Most users may encounter some syntax errors when editing the problem.pddl file. At this time, users need to be able to accurately locate these errors and make targeted corrections. The text file format cannot provide users with the necessary error messages, which increases the difficulty for users to edit the problem.

[0011] 4. Different domains have different objects, predicates, and actions. For a completely new domain, users need some time to understand the meaning and function of the objects, predicates, and actions, which may make it difficult to get started.

[0012] Therefore, it is evident that the main problems users currently encounter are concentrated on editing the problem.pddl file in PDDL. Users need to be proficient in PDDL syntax and familiar with the relevant domain background in order to successfully edit an ideal problem.pddl file, which inadvertently raises the entry barrier for PDDL. Even for domain experts on related issues, editing PDDL-related files is a time-consuming and laborious task due to the limitations of PDDL's syntax structure and representation. Summary of the Invention

[0013] To address the problems and shortcomings of existing technologies, this invention provides a PDDL problem calculation method and system based on mind mapping, which can facilitate users in solving planning problems using PDDL.

[0014] The technical content of this invention includes:

[0015] A method for calculating PDDL (Problem-Based Daily Description) questions based on mind mapping, the method comprising:

[0016] Based on the PDDL question to be edited, a domain.pddl file is obtained, and according to the domain.pddl file, the mind map component information required to edit the PDDL question corresponding to the domain.pddl file is determined;

[0017] Using the information from the mind mapping component, the PDDL question can be edited into a mind map;

[0018] Parse the mind map to obtain the problem.pddl information;

[0019] Based on the problem.pddl information and the domain corresponding to the PDDL problem, generate the problem.pddl file;

[0020] Based on the domain.pddl file and the problem.pddl file, the solution to the PDDL problem is obtained.

[0021] Further, based on the PDDL question to be edited, a domain.pddl file is obtained, and according to the domain.pddl file, the mind map component information required to edit the PDDL question corresponding to the domain.pddl file is determined, including:

[0022] Based on the PDDL issue to be edited, select from the pre-integrated domain.pddl file;

[0023] Based on the selected domain.pddl file, obtain the component information required to edit the PDDL issue corresponding to the domain.pddl file;

[0024] or,

[0025] Based on the PDDL issue to be edited, the user uploads a custom domain.pddl file;

[0026] The information in the custom domain.pddl file is obtained, and the obtained information is represented in the form of a syntax tree to obtain tree-like information; wherein, the information in the custom domain.pddl file includes: predicates and verbs;

[0027] Based on the tree-like information, determine the mind map component information required to edit the PDDL question corresponding to the domain.pddl file.

[0028] Furthermore, the step of using the mind map component information to edit the PDDL question into a mind map includes:

[0029] The PDDL problems are organized into a tree structure, where the root node of the tree structure is the name of the PDDL problem;

[0030] The objects node, init node, and goal node are respectively designated as child nodes of the root node;

[0031] Set a type node according to the type of the object instance in the PDDL problem, and make the type node a child node of the objects node, and make the node corresponding to the object instance a child node of the type node;

[0032] An initial state predicate node is set according to the predicates that hold true in the initial state of the PDDL problem, and the initial state predicate node is used as a child node of the init node. The node corresponding to the predicate parameter is used as a child node of the initial state predicate node.

[0033] Based on the predicates that confirm the target state in the PDDL problem, a target state predicate node is set, and the target state predicate node is used as a child node of the goal node. The node corresponding to the predicate parameter is used as a child node of the target state predicate node.

[0034] Furthermore, the process of parsing the mind map to obtain the problem.pddl information also includes:

[0035] Perform a syntax check on the problem.pddl information;

[0036] If there is a syntax error, provide an error message;

[0037] If the syntax is correct, then proceed to the step of generating the corresponding problem.pddl file based on the problem.pddl information and the domain.pddl file.

[0038] Furthermore, the method also includes:

[0039] Save the mind map;

[0040] When faced with a new PDDL question to be edited, the mind map is edited based on the component information to obtain a new mind map.

[0041] Furthermore, the method also includes:

[0042] Upload the domain.pddl file;

[0043] The system parses the corresponding object type, predicate, and action from domain.pddl and automatically generates mind map editing buttons for users to use when editing questions.

[0044] A mind map-based PDDL problem calculation system includes:

[0045] The system front-end includes:

[0046] The domain selection module is used to jump to the corresponding domain introduction and mind map editing interface based on the domain to which the PDDL question to be edited belongs;

[0047] The autofit module is used to upload the domain.pddl file to the system backend so that the editing toolbar of the mind map can be rendered based on the returned component information;

[0048] The mind map editing module is used to edit the PDDL question into a mind map using the editing toolbar and send the mind map to the system backend;

[0049] The system backend includes:

[0050] The mind map component generation module is used to determine the mind map component information required to edit the PDDL question corresponding to the domain.pddl file based on the domain.pddl file.

[0051] The mind map parsing module is used to parse the mind map and obtain the problem.pddl information.

[0052] The problem.pddl generation module is used to generate a problem.pddl file based on the problem.pddl information and the domain corresponding to the PDDL problem;

[0053] A solver is used to obtain a solution to the PDDL problem based on the domain.pddl file and the problem.pddl file.

[0054] Furthermore, the system front end also includes:

[0055] The AI ​​Planning introduction module is used to introduce AI Planning and PDDL;

[0056] The problem.pddl file display module is used to display the problem.pddl file;

[0057] The solution result display module is used to display the solution to the PDDL problem.

[0058] Furthermore, it also includes:

[0059] The syntax checking module is used to check the syntactic correctness of the problem.pddl information.

[0060] Furthermore, the mind map component generation module includes:

[0061] The domain.pddl parsing unit is used to obtain information from the custom domain.pddl file, and represent the obtained information in the form of a syntax tree to obtain tree-like information; wherein, the information in the custom domain.pddl file includes: predicates and verbs;

[0062] Component generation unit, used for:

[0063] For the pre-integrated domain.pddl file, obtain the component information required to edit the PDDL issue corresponding to the domain.pddl file based on the domain.pddl file;

[0064] or,

[0065] For a custom domain.pddl file, based on the tree-like information, determine the component information required to edit the PDDL issue corresponding to the domain.pddl file.

[0066] Compared with the prior art, the beneficial effects of the present invention include:

[0067] ① This invention provides convenience for users to solve planning problems using PDDL: This invention solves the problems of poor readability, complex format maintenance, and high editing difficulty of text editing by converting the complex and abstract text representation of PDDL problems into a user-friendly and easy-to-understand mind map representation.

[0068] ②This invention provides users with support for solving planning problems represented by PDDL in real time, simplifies the process of solving PDDL problems, and solves the problems of time-consuming and cumbersome PDDL solutions.

[0069] ③ This invention provides users with support for parsing entirely new domains. By parsing domain.pddl into mind map editing components, it solves the problem of users having a high difficulty in getting started with entirely new domain.pddl. Attached Figure Description

[0070] Figure 1 This is the overall system architecture diagram of the present invention.

[0071] Figure 2 This is the front-end architecture diagram of the present invention.

[0072] Figure 3 This is a diagram of the backend architecture of the present invention.

[0073] Figure 4 This is a flowchart of the method of the present invention. Detailed Implementation

[0074] The present invention will be further illustrated by the following examples, but these examples are not intended to limit the scope of the invention in any way.

[0075] The overall architecture of the present invention is as follows Figure 1 As shown, the system mainly consists of three parts: the system frontend, the system backend, and the solver. The system frontend and backend interact through predefined APIs in Django, while the system backend interacts with the solver through its corresponding API. Depending on the solver's deployment location, it can be divided into two scenarios: server-side and local. The server-side solver is already deployed and can be accessed directly via a public URL; the local solver requires local configuration according to the instructions before use, and then can be accessed via a user-defined URL.

[0076] like Figure 2As shown, the system front-end mainly includes an AI Planning introduction module, a domain selection module, an auto-fit module, a mind map editing module, a problem.pddl file display module, and a solution result display module. The AI ​​Planning introduction module is the main interface of the system. This module includes an introduction to AI Planning and PDDL, giving users a preliminary understanding of the Graph Planner system. The domain selection module provides several domains for users to choose from. These domains are pre-integrated by the system and can be used to get started with Graph Planner, as well as serve as a demo during system introductions. The auto-fit module... The `fit` module sends the user-uploaded `domain.pddl` file to the system backend's `domain.pddl` parsing module for parsing. Based on the parsed information, it automatically generates corresponding mind map editing buttons for user editing. The mind map editing module provides a mind map editing page. The left side of this page features a toolbar including modules such as `object`, `predicate`, and `action`, while the right side displays the corresponding mind map. Users can save their edited mind maps locally in the system's specified format or upload them to the system for further modification. The `problem.pddl` file display module converts the user-edited mind map into a corresponding `problem.pddl` file and displays it, giving users a general understanding of the content. The solution result display module sends the user-edited mind map to the system backend, receives the solution results, and displays the solution in a user-friendly manner (syntax highlighting).

[0077] like Figure 3 As shown, the system backend mainly includes a mind map parsing module, a problem.pddl generation module, an AIPlanning solution module, a domain.pddl parsing module, a mind map component generation module, and a syntax checking module.

[0078] The mind map parsing module is responsible for receiving and parsing the mind map file returned by the front end in JSON format, and then saving the user-defined information about problem.pddl (objects, init, goal, etc.).

[0079] The problem.pddl generation module is responsible for generating the corresponding problem.pddl file according to the PDDL syntax format based on the information parsed from the mind map (objects, init, goal, etc.) and the domain corresponding to the problem, and then returning the corresponding file to the system front end for display.

[0080] The AI ​​Planning solver module is responsible for calling the API provided by the configured solver (server-side / local-side) to solve the problem and returning the solution to the system front-end for display.

[0081] The mind map component generation module includes a domain.pddl parsing unit and a component generation unit. The domain.pddl parsing unit is responsible for parsing the user-uploaded custom domain.pddl file. This module can extract predefined information such as predicates and actions from the domain.pddl file and save it as a tree-like structure. The component generation unit uses this tree-like structure to determine the component information needed to edit the problem corresponding to the domain, and then sends this information to the front end in the form of a JSON file for generating a customized interface.

[0082] The syntax checking module is used to detect problems that users may encounter while editing mind maps, analyze the processor's solution results, and provide corresponding prompts to the user.

[0083] An example scenario of the present invention is as follows: there is a robot named robot, two balls named ball1 and ball2, and two rooms named roomA and roomB. In the initial state, robot, ball1 and ball2 are all in roomA. Robot wants to move ball1 and ball2 to roomB and then return to roomA.

[0084] Based on the above, such as Figure 4 As shown, the implementation steps of the present invention are as follows:

[0085] Step 1: Open the system page

[0086] Users can open the main interface of the system of this invention by entering a URL (e.g., http: / / 127.0.0.1:8000) in their browser.

[0087] Step Two: Select Domain: The user selects the domain (carry_ball) corresponding to the problem to be edited, and the system will then redirect to the corresponding editing page. Users can use the components on this interface to edit the mind map. Additionally, this system provides mind map templates for several specific domains for user reference.

[0088] Step 3: Edit the mind map

[0089] Users can edit the questions they want to represent on the mind map editing interface. There are two editing methods.

[0090] Users can directly represent the problem they want to solve in the newly created editing interface. By adding or deleting mind map components and modifying the content (text) within the components, they can represent the various components (objects, init, goal) of the PDDL problem. In the mind map, the entire problem is organized into a tree structure. The problem's name is the root node of the tree. The objects node, the init node, and the goal node are the three child nodes of the name node, which are the root nodes of the three subtrees of the root, corresponding to the three main modules (objects, init, goal) in problem.pddl.

[0091] ① In the subtree corresponding to the objects node, the type of the object is a child node of the objects node. There are as many type nodes as there are types. The children of each type node are the object instances of that type in the problem. There are as many child nodes as there are instances of the type node.

[0092] ② In the subtree corresponding to the init node, each predicate node represents a subtree. The root node of this subtree is the name of the predicate, and the leaf nodes are the parameters of the predicate. Each parameter corresponds to a leaf node. In the initial state, all currently valid predicate nodes are children of the init node. The number of child nodes of the init node is equal to the number of valid predicates.

[0093] ③ In the subtree corresponding to the goal node, each predicate node represents a subtree. The root node of the subtree is the name of the predicate, and the leaf nodes of the subtree are the parameters of the predicate. Each parameter corresponds to a leaf node. In the goal state, all currently valid predicate nodes are child nodes of the goal node. The goal node has as many child nodes as there are valid predicates.

[0094] Taking this scenario as an example, when editing the mind map, the name of the PDDL problem, carry_ball, is used as the root node of the tree. The three main modules of the PDDL problem, objects, init, and goal, are used as the three child nodes of the root. This allows the corresponding content to be represented in three subtrees. In the objects subtree, the three object types bot, ball, and location correspond to the three child nodes of the objects node. Objects of the same type form a subtree. For example, robot is a child node of the subtree rooted in bot, ball1 and ball2 are two child nodes of the subtree rooted in ball, and roomA and roomB are two child nodes of the subtree rooted in location. In the init and goal subtrees, all currently valid predicates form the child nodes of the subtree root. For example, for the predicate (on ball1 roomA), on is a child node of init, and the parameters ball1 and roomA of the predicate form the child nodes of on. In the init state, there are 6 valid predicates, so the init node has 6 child nodes. This hierarchical representation allows PDDL problems to be represented in the form of a mind map.

[0095] Users upload previously edited mind map files, and the system automatically converts these files into mind maps for display. Users can then edit questions based on these mind maps.

[0096] Step 4: Generate problem.pddl

[0097] While editing the mind map, users can click the "problem.pddl generate" button. The system will then convert the mind map file into a corresponding problem.pddl file, which will be displayed on the interface at the bottom of the system. During the parsing process, since the mind map is organized in JSMID format, the system frontend will first send the user-edited JSMID mind map to the system backend as a JSON file. The backend will then parse the contents of the JSON file to obtain information related to the problem's objects, init, and goals, and then generate the corresponding problem.pddl file according to PDDL syntax.

[0098] Step 5: Problem Solving

[0099] After the user finishes editing the mind map and clicks the submit button, the system will parse the submitted mind map file. During the parsing process, a syntax check will be performed. If no syntax errors are found:

[0100] • If a syntax error is detected, an error message will be provided to the user.

[0101] If no syntax errors are found, the solver's API will be called to solve the problem, and the results will be displayed in the interface at the bottom of the system.

[0102] Step Six: Download the mind map file

[0103] After editing the question, users can save the mind map in a predefined format (.node_tree, .node_array, .freemind) for future use.

[0104] Step 7: Automatically resolve domain.pddl

[0105] If the domain corresponding to the question a user wants to edit is not integrated into the system, the user can upload the corresponding domain.pddl file to the system. The system can use the auto-fit function to parse the corresponding object type, predicate, and action in the domain.pddl file and automatically generate mind map editing buttons for the user to use when editing the question.

[0106] In summary, this invention addresses the problem of unclear hierarchical structure in text-based problem.pddl files by using a mind map to represent the problem.pddl file. Considering the tree-like structure of the mind map itself, when editing the problem, users can treat the three main parts—objects, init, and goal—as the three subtrees of the root node of the mind map tree, and then treat the content corresponding to each module as branches within the subtrees. This allows users to represent the problem to be solved in the form of a mind map according to the syntax of PDDL and the templates provided by this system. This makes the problem description clearer and more intuitive, reducing the time cost for users to edit the problem.

[0107] This invention addresses problems where users previously needed to manually upload domain.pddl and problem.pddl files for solution. Instead, it saves the uploaded domain.pddl file in a designated location, requiring only one upload. While the user is editing the problem, the system converts the edited mind map into a corresponding problem.pddl file in real time and displays it on the interface for user review. After the user has finished editing all the mind maps for the problem, they can click the submit button. The system then solves the problem using the solver's API along with the previously stored domain.pddl file, displaying the solution results on the corresponding interface. This allows users to solve problems with a single click while editing, significantly simplifying the solution process.

[0108] This invention addresses the problem of insufficient error messages in text editing methods, which hinders debugging. It integrates a syntax checking system that performs syntax checks during the parsing of the user-edited mind map and provides corresponding prompts. Simultaneously, this method processes the solver's results and provides hints for any issues, allowing users to promptly identify potential syntax problems during the editing process.

[0109] This invention addresses the challenge of getting started with a completely new domain. The method integrates a syntax analysis system that automatically parses the user-uploaded domain.pddl file, extracting domain-related information for the front-end to dynamically generate corresponding modular components. This allows users to use the generated components to edit mind maps, significantly reducing the difficulty of learning a new domain.

[0110] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Those skilled in the art can modify or make equivalent substitutions to the technical solutions of the present invention. The scope of protection of the present invention should be determined by the claims.

Claims

1. A method for calculating PDDL (Problem-Do-Check-Act) questions based on mind mapping, the method comprising: Based on the PDDL question to be edited, a domain.pddl file is obtained, and according to the domain.pddl file, the mind map component information required to edit the PDDL question corresponding to the domain.pddl file is determined; Using the information from the mind mapping component, the PDDL question can be edited into a mind map; Parse the mind map to obtain the problem.pddl information; Based on the problem.pddl information and the domain corresponding to the PDDL problem, generate the problem.pddl file; Based on the domain.pddl file and the problem.pddl file, the solution to the PDDL problem is obtained; Specifically, based on the PDDL question to be edited, a domain.pddl file is obtained, and according to the domain.pddl file, the mind map component information required to edit the PDDL question corresponding to the domain.pddl file is determined, including: Based on the PDDL issue to be edited, select from the pre-integrated domain.pddl file; Based on the selected domain.pddl file, obtain the component information required to edit the PDDL issue corresponding to the domain.pddl file; or, Based on the PDDL issue to be edited, the user uploads a custom domain.pddl file; The information in the custom domain.pddl file is obtained, and the obtained information is represented in the form of a syntax tree to obtain tree-like information; wherein, the information in the custom domain.pddl file includes: predicates and verbs; Based on the tree-like information, determine the mind map component information required to edit the PDDL question corresponding to the domain.pddl file; The process of editing a mind map for the PDDL question using the mind map component information includes: The PDDL problems are organized into a tree structure, where the root node of the tree structure is the name of the PDDL problem; The objects node, init node, and goal node are respectively designated as child nodes of the root node; Set a type node according to the type of the object instance in the PDDL problem, and make the type node a child node of the objects node, and make the node corresponding to the object instance a child node of the type node; An initial state predicate node is set according to the predicates that hold true in the initial state of the PDDL problem, and the initial state predicate node is used as a child node of the init node. The node corresponding to the predicate parameter is used as a child node of the initial state predicate node. Based on the predicates that confirm the target state in the PDDL problem, a target state predicate node is set, and the target state predicate node is used as a child node of the goal node. The node corresponding to the predicate parameter is used as a child node of the target state predicate node.

2. The method as described in claim 1, characterized in that, The process of parsing the mind map to obtain the problem.pddl information also includes: Perform a syntax check on the problem.pddl information; If there is a syntax error, provide an error message; If the syntax is correct, the program will jump to the domain corresponding to the problem and the PDDL problem, and generate the problem.pddl file.

3. The method as described in any one of claims 1-2, characterized in that, The method further includes: Save the mind map; When faced with a new PDDL question to be edited, the mind map is edited based on the component information to obtain a new mind map.

4. The method as described in any one of claims 1-2, characterized in that, The method further includes: Upload the domain.pddl file; The system parses the corresponding object type, predicate, and action from domain.pddl and automatically generates mind map editing buttons for users to use when editing questions.

5. A PDDL problem calculation system based on mind mapping, comprising: The system front-end includes: The domain selection module is used to jump to the corresponding domain introduction and mind map editing interface based on the domain to which the PDDL question to be edited belongs; The autofit module is used to upload the domain.pddl file to the system backend so that the editing toolbar of the mind map can be rendered based on the returned component information; The mind map editing module is used to edit the PDDL question into a mind map using the editing toolbar and send the mind map to the system backend; The system backend includes: The mind map component generation module is used to determine the mind map component information required to edit the PDDL question corresponding to the domain.pddl file based on the domain.pddl file. The mind map parsing module is used to parse the mind map and obtain the problem.pddl information. The problem.pddl generation module is used to generate a problem.pddl file based on the problem.pddl information and the domain corresponding to the PDDL problem; A solver is used to obtain a solution to the PDDL problem based on the domain.pddl file and the problem.pddl file; The generation process of the domain.pddl file includes: selecting from the pre-integrated domain.pddl file based on the PDDL issue to be edited, or uploading a custom domain.pddl file based on the PDDL issue to be edited; When the domain.pddl file is selected from the pre-integrated domain.pddl files, the step of determining the mind map component information required to edit the PDDL question corresponding to the domain.pddl file based on the domain.pddl file includes: Based on the selected domain.pddl file, obtain the component information required to edit the PDDL issue corresponding to the domain.pddl file; When the domain.pddl file is a custom domain.pddl file, the step of determining the mind map component information required to edit the PDDL question corresponding to the domain.pddl file based on the domain.pddl file includes: The information in the custom domain.pddl file is obtained, and the obtained information is represented in the form of a syntax tree to obtain tree-like information; wherein, the information in the custom domain.pddl file includes: predicates and verbs; Based on the tree-like information, determine the mind map component information required to edit the PDDL question corresponding to the domain.pddl file; The process of creating a mind map for the PDDL question includes: The PDDL problems are organized into a tree structure, where the root node of the tree structure is the name of the PDDL problem; The objects node, init node, and goal node are respectively designated as child nodes of the root node; Set a type node according to the type of the object instance in the PDDL problem, and make the type node a child node of the objects node, and make the node corresponding to the object instance a child node of the type node; An initial state predicate node is set according to the predicates that hold true in the initial state of the PDDL problem, and the initial state predicate node is used as a child node of the init node. The node corresponding to the predicate parameter is used as a child node of the initial state predicate node. Based on the predicates that confirm the target state in the PDDL problem, a target state predicate node is set, and the target state predicate node is used as a child node of the goal node. The node corresponding to the predicate parameter is used as a child node of the target state predicate node.

6. The system as described in claim 5, characterized in that, The system front-end also includes: The AI ​​Planning introduction module is used to introduce AI Planning and PDDL; The problem.pddl file display module is used to display the problem.pddl file; The solution result display module is used to display the solution to the PDDL problem.

7. The system as described in claim 5, characterized in that, The system backend also includes: The syntax checking module is used to check the syntactic correctness of the problem.pddl information.

Images