Search

CN-121998040-A - Method, apparatus, device and computer program product for task planning

CN121998040ACN 121998040 ACN121998040 ACN 121998040ACN-121998040-A

Abstract

Embodiments of the present disclosure relate to methods, apparatuses, devices and computer program products for task planning. The method includes generating an heterographic representation of a plurality of objects related to a task based on a current state of the task, wherein the heterographic representation includes object nodes indicating the plurality of objects and a first edge indicating a relationship between the plurality of objects in the current state. In addition, the method includes generating task actions through the graph neural network based on the heterograph representation. Therefore, according to the embodiment of the disclosure, the task planning can be modeled through the heterogeneous graph data, task actions of the task planning are generated by utilizing the graph neural network, a manual design learning strategy is avoided, the complexity of the task planning is reduced, and the effect of the task planning can be improved.

Inventors

  • SUN ZHIGANG
  • JIANG HAO

Assignees

  • BSH家用电器有限公司

Dates

Publication Date
20260508
Application Date
20241031

Claims (15)

  1. 1. A method (200) for mission planning, comprising: Generating (202) an heterogram representation of a plurality of objects related to a task based on a current state of the task, wherein the heterogram representation includes object nodes indicating the plurality of objects and a first edge indicating a relationship between the plurality of objects in the current state; based on the heterographic representation, task actions are generated (204) through a graph neural network.
  2. 2. The method (200) of claim 1, wherein generating the task action through the graph neural network based on the heterograph representation comprises: predicting, based on the heterograph representation, a second edge between at least two object nodes in the heterograph representation through a graph neural network, wherein the second edge indicates a likelihood that a task action relationship exists between objects represented by the at least two object nodes And determining task actions in the current state based on the prediction result.
  3. 3. The method (200) of claim 1, further comprising: updating the current state of the task based on the generated task action; repeating the steps of executing the task based on the current state of the task, generating task actions, and updating the current state of the task based on the generated task actions until the task reaches an expected state.
  4. 4. The method (200) of claim 1, wherein generating the heterogeneous representation of the plurality of objects related to the task comprises: generating a isomorphic representation corresponding to the current state based on the current state of the task, the isomorphic representation including object information of the plurality of objects and connection information between the plurality of objects, and Based on the isomorphic representation, the heteromorphic representation is generated that includes the plurality of object nodes, wherein the plurality of object nodes have a plurality of object node types.
  5. 5. The method (200) of claim 4, wherein generating the heterogeneous graph representation including the plurality of object nodes comprises: The node type and initial node embedding of each object node and the connection type and initial connection embedding of each connection are saved by traversing the isomorphic representation; The heterographic representation is generated based on the node type and initial node embedding for each object node and the connection type and initial connection embedding for each connection.
  6. 6. The method (200) of claim 4, wherein generating the isomorphic representation comprises: Acquiring a text corresponding to the current state of the task; Generating an ordered triplet using a language model based on the text, the ordered triplet including a first object node, a connection of object nodes, and a second object node, and Based on the ordered triples, the isomorphic representation corresponding to the current state is generated.
  7. 7. The method (200) of claim 6, further comprising: switching the task from the current state to a next state corresponding to the task action being executed; based on the next state, updating a state diagram of the task, wherein the state diagram comprises a plurality of state nodes and a task scene diagram corresponding to each state node, each state node corresponds to one generated task action of the task, and the task scene diagram comprises isomorphic diagram representation of the task at the corresponding state node.
  8. 8. The method (200) of claim 7, wherein updating the state diagram of the task comprises: generating an ordered triplet of task actions using the language model based on action representation text corresponding to the task actions; generating an isomorphic representation of the next state by breadth-first search based on the ordered triples of task actions and the isomorphic representation of the current state, and Updating the state diagram of the task with a isomorphic representation of the next state.
  9. 9. The method (200) of claim 7, further comprising: Generating an expected isomorphic representation corresponding to an expected state of the task based on text corresponding to the expected state, and And determining that the task reaches the expected state in response to the latest state node in the state diagram of the task corresponding to the expected isomorphic representation.
  10. 10. The method (200) of claim 1, further comprising: And training the graph neural network through heterogeneous graph training data, wherein object nodes corresponding to the reference task actions in the heterogeneous graph training data are set with high-score truth labels.
  11. 11. The method (200) of claim 10, further comprising: Connections between object nodes in the heterogeneous graph training data are annotated with an annotation tool configured to visually annotate.
  12. 12. The method (200) of claim 2, wherein generating the task action comprises: generating node embeddings of object nodes in the heterograph representation using the graph neural network; Determining a second edge having a highest probability by taking the node-embedded vector inner product as the probability that the second edge exists between two object nodes, and And generating the task action based on the second edge with the highest probability and the corresponding object node.
  13. 13. An apparatus for task planning, comprising: A heterogeneous representation generation module configured to generate a heterogeneous composition representation of a plurality of objects related to a task based on a current state of the task, wherein the heterogeneous composition representation includes object nodes indicating the plurality of objects and a first edge indicating a relationship between the plurality of objects in the current state, an And the task action generating module is configured to generate task actions through the graph neural network based on the heterograph representation.
  14. 14. An electronic device, comprising: At least one processor, and A memory coupled to the at least one processor and having instructions stored thereon that, when executed by the at least one processor, cause the apparatus to perform the method of any of claims 1-12.
  15. 15. A computer program product tangibly stored on a non-transitory computer readable medium and comprising machine executable instructions for performing the method of any one of claims 1 to 12.

Description

Method, apparatus, device and computer program product for task planning Technical Field Embodiments of the present disclosure relate to the field of computers, and more particularly, to methods, apparatuses, devices, and computer program products for task planning. Background Task planning is a process of automatically generating a series of actions or decision sequences according to environment information, and relates to multiple steps of sensing environment, setting targets, planning paths, executing tasks and the like, so that efficient and accurate automatic operation is realized. The technology is widely applied to the fields of robots, industrial automation, unmanned operation and the like. With the development of technology, the importance of task planning is increasingly highlighted. The device not only improves the production efficiency and reduces the labor cost, but also can adapt to complex and changeable environments and execute finer and diversified tasks. Mission planning is becoming a key technology pushing the progress of many industries, and as its level of intelligence and autonomy increases, it will play a more important role in future society. Disclosure of Invention Embodiments of the present disclosure provide a method, apparatus, device, computer program product, and medium for task planning. According to a first aspect of the present disclosure, a method for task planning is provided. The method includes generating an heterogram representation of a plurality of objects related to a task based on a current state of the task, wherein the heterogram representation includes object nodes indicating the plurality of objects and a first edge indicating a relationship between the plurality of objects in the current state. Furthermore, the method includes generating a task action through the graph neural network based on the heterograph representation. According to a second aspect of the present disclosure, an apparatus for mission planning is provided. The apparatus includes a heterogeneous representation generation module configured to generate a heterographic representation of a plurality of objects related to a task based on a current state of the task, wherein the heterographic representation includes object nodes indicating the plurality of objects and a first edge indicating a relationship between the plurality of objects in the current state. In addition, the apparatus includes a task action generation module configured to generate a task action through the graph neural network based on the heterograph representation. According to a third aspect of the present disclosure, an electronic device is provided. The electronic device includes at least one processor, and a memory coupled to the at least one processor and having instructions stored thereon that, when executed by the at least one processor, cause the device to perform the steps of the method of the first aspect of the disclosure. According to a fourth aspect of the present disclosure, a computer program product is provided. The computer program product is tangibly stored on a non-transitory computer-readable medium and includes computer-executable instructions that, when executed, cause a computer to perform the steps of the method of the first aspect of the present disclosure. According to a fifth aspect of the present disclosure, a machine-readable storage medium is provided. The machine-readable storage medium has stored thereon machine-executable instructions which are executed by a processor to implement the steps of the method of the first aspect of the present disclosure. The summary is provided to introduce a selection of concepts in a simplified form that are further described below in the detailed description. This summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter. Drawings The foregoing and other objects, features and advantages of the disclosure will be apparent from the following more particular descriptions of exemplary embodiments of the disclosure as illustrated in the accompanying drawings wherein like reference numbers generally represent like parts throughout the exemplary embodiments of the disclosure. FIG. 1 illustrates a schematic diagram of an example environment in which devices and/or methods may be implemented, according to embodiments of the present disclosure; FIG. 2 illustrates a flow chart of a method for task planning in accordance with an embodiment of the present disclosure; FIG. 3 illustrates a flow chart of a process of task planning in accordance with an embodiment of the present disclosure; FIG. 4 illustrates a schematic diagram of a process of generating scene graph data utilizing an embodiment of the present disclosure; FIG. 5 illustrates a schematic diagram of a process of converting scene graph data into heterogeneous graph data, according to an embodiment