Task Processing Method, Device, and Storage Medium Based on an Evaluation Hierarchical Task Network

Abstract

The present invention discloses a task processing method, device and storage medium based on an evaluation hierarchical task network, which relates to the technical field of computer-aided intelligent information processing, and includes: generating an initial evaluation hierarchical task network according to a target task, where the initial evaluation hierarchical task network includes a hierarchical task network; when the target node finishes execution, determining the evaluation information of the target node; determining the pre-order task sequence of the associated nodes of the target node according to the evaluation information; determining an optimized evaluation hierarchical task network according to the pre-order task sequence, and completing the target task according to the optimized evaluation hierarchical task network. Based on the hierarchical task network, the target node in the initial evaluation hierarchical task network is evaluated to obtain evaluation information, the pre-order task sequence of the associated nodes of the target node is adjusted according to the evaluation information, and then the initial evaluation hierarchical task network is optimized according to the evaluation information of the target node to obtain an optimized evaluation hierarchical task network, thereby improving the execution efficiency of the target task.

Description

Technical Field

[0001] The present invention relates to the technical field of computer-aided intelligent information processing, and particularly to a task processing method, device and storage medium based on an evaluation hierarchical task network. Background Art

[0002] With the continuous enhancement of device intelligence, devices are gradually able to complete more and more complex tasks. Currently, when a device executes a complex task, computer-aided technologies such as a Hierarchical Task Network (HTN) are required to provide specific parameters for the tasks and actions of the device. However, currently, the execution efficiency of tasks when using a hierarchical task network to execute tasks is low. Summary of the Invention

[0003] The present invention provides a task processing method, device and storage medium based on an evaluation hierarchical task network to solve the problem of low task execution efficiency when a current device executes a task through a hierarchical task network.

[0004] According to one aspect of the present invention, there is provided a task processing method based on an evaluation hierarchical task network, including:

[0005] generating an initial evaluation hierarchical task network according to a target task, the initial evaluation hierarchical task network including a hierarchical task network;

[0006] when the target node is executed, determining evaluation information of the target node, the target node being any task node in the evaluation hierarchical task network;

[0007] determining a pre-order task sequence of associated nodes of the target node according to the evaluation information; determining an optimized evaluation hierarchical task network according to the pre-order task sequence, and completing the target task according to the optimized evaluation hierarchical task network.

[0008] According to another aspect of the present invention, there is provided a task processing device based on an evaluation hierarchical task network, including:

[0009] an initialization module for generating an initial evaluation hierarchical task network according to a target task, the initial evaluation hierarchical task network including a hierarchical task network;

[0010] an evaluation module for determining evaluation information of the target node when the target node is executed, the target node being any task node in the evaluation hierarchical task network;

[0011] An optimization module, configured to determine a pre-task sequence of associated nodes of the target node according to the evaluation information; determine an optimized evaluation hierarchical task network according to the pre-task sequence, and complete the target task according to the optimized evaluation hierarchical task network.

[0012] According to another aspect of the present invention, there is provided an electronic device, including:

[0013] At least one processor; and

[0014] A memory communicatively connected to the at least one processor; wherein,

[0015] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the task processing method based on the evaluation hierarchical task network according to any embodiment of the present invention.

[0016] According to another aspect of the present invention, there is provided a computer-readable storage medium storing computer instructions for implementing the task processing method based on the evaluation hierarchical task network according to any embodiment of the present invention when executed by a processor.

[0017] The technical solution of the embodiment of the present invention generates an initial evaluation hierarchical task network according to a target task, and the initial evaluation hierarchical task network includes a hierarchical task network; when the target node finishes execution, determine the evaluation information of the target node, where the target node is any task node in the evaluation hierarchical task network; determine a pre-task sequence of associated nodes of the target node according to the evaluation information; determine an optimized evaluation hierarchical task network according to the pre-task sequence, and complete the target task according to the optimized evaluation hierarchical task network. Based on the hierarchical task network, evaluate the target node in the initial evaluation hierarchical task network to obtain evaluation information, adjust the pre-task sequence of the associated nodes of the target node according to the evaluation information, and then optimize the initial evaluation hierarchical task network according to the evaluation information of the target node to obtain an optimized evaluation hierarchical task network, thereby improving the execution efficiency of the target task.

[0018] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present invention, nor is it used to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0020] Figure 1 It is a schematic flowchart of a task processing method based on an evaluation hierarchical task network provided by an embodiment of the present invention;

[0021] Figure 2 It is a schematic diagram of process data related to weight scores provided by an embodiment of the present invention;

[0022] Figure 3 It is a schematic flowchart of another task processing method based on an evaluation hierarchical task network provided by an embodiment of the present invention;

[0023] Figure 4 It is a schematic structural diagram of a task processing device based on an evaluation hierarchical task network provided by an embodiment of the present invention;

[0024] Figure 5 It is a schematic structural diagram of an electronic device for implementing the task processing method based on an evaluation hierarchical task network in the embodiment of the present invention. Detailed implementation manners

[0025] In order to enable those skilled in the art to better understand the solution of the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0026] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to describe a specific order or sequence. It should be understood that such used data can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those clearly listed steps or units, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0027] As the intelligence of devices continues to strengthen, devices are gradually able to complete more and more complex tasks. Currently, when a device executes a complex task, computer-aided technologies such as a hierarchical task network need to provide specific parameters for the task and actions of the device. However, currently, when using a hierarchical task network to execute a task, the execution efficiency of the task is low. Specifically, currently, the hierarchical task network is separated from task evaluation, that is, after the target task is completed based on the hierarchical task network, the overall completion of the target task is evaluated. And the hierarchical task network cannot be optimized according to this evaluation, resulting in low execution efficiency of the target task.

[0028] The task processing method based on an evaluation hierarchical task network provided by an embodiment of the present invention can, during the execution of a target task, perform an evaluation based on the execution result of any task node, that is, the target node, to obtain evaluation information. Optimize the initial evaluation hierarchical task network according to the evaluation information, so that during the execution of the target task, according to the execution result of a certain node during the execution process, adjust the subsequent execution process of the target task, that is, adjust the pre-task sequence of the associated nodes of the target node. So that the target task can dynamically adjust the task flow according to the evaluation information, improving the execution efficiency of the target task.

[0029] Figure 1 FIG. is a schematic flowchart of a task processing method based on an evaluation hierarchical task network provided by an embodiment of the present invention. This embodiment is applicable to the situation of task processing through a computer-aided device. The computer-aided device is used to provide specific actions for an intelligent device to execute. The intelligent device includes, but is not limited to, an unmanned aerial vehicle, an intelligent robot, etc. Among them, the intelligent robot can be a machine with a robotic arm or equipment required to execute tasks. Such as a floor-sweeping robot with a rolling brush, etc. This method can be executed by a task processing device based on an evaluation hierarchical task network. The task processing device based on an evaluation hierarchical task network can be implemented in the form of hardware and / or software. The task processing device based on an evaluation hierarchical task network can be configured in an electronic device such as an intelligent device, a personal computer, or a server. As Figure 1 shown, the method includes:

[0030] Step S101, generate an initial evaluation hierarchical task network according to the target task, and the initial evaluation hierarchical task network includes a hierarchical task network.

[0031] The target task is a task that the device needs to complete. By splitting the target task, the subtasks required to complete the target task and the actions included in the subtasks are obtained. The target task can be divided into multiple subtasks, multiple actions, or a combination of multiple actions and subtasks. An action can be the smallest task unit that the device can execute. A subtask can be composed of multiple actions. The subtasks can be in a parent-child relationship or a sibling relationship.

[0032] Exemplarily, when the device is a floor cleaning robot, the target task can be to clean the floors of four rooms. The target task includes four subtasks corresponding to the floor cleaning of each room respectively. Each subtask further includes its own subtasks for generating targeted subtasks or actions according to the different terrains of each room.

[0033] The application scenarios cover the work of an agent or multiple agents, such as remote sensing photography, drone shows, intelligent delivery, Automated Guided Vehicle (AGV) scheduling, etc. Specifically, when the agent is a drone swarm, the target task can be to complete the entire show, and each drone has its own subtasks, such as flying along the designed route and avoiding obstacles.

[0034] Construct an initial evaluation hierarchical task network according to the subtasks included in the target task. The initial evaluation hierarchical task network includes a hierarchical task network. The hierarchical task network can be constructed according to the construction method of the hierarchical task network, and the nodes of the hierarchical task network include hierarchical task network parameters, such as , where represents the task information of task node , represents the output information of task node , represents the input information of task node .

[0035] Different from the hierarchical task network in the prior art, the initial evaluation hierarchical task network provided by the embodiments of the present invention further evaluates relevant parameters in each node on the basis of the hierarchical task network. The target node is any task node in the evaluation hierarchical task network. The target node of the evaluation hierarchical task network includes hierarchical task network parameters and evaluation relevant parameters. The evaluation relevant parameters include one or a combination of more than one of the following parameters: task subject, evaluation information, previous task, operation target, evaluation mode, task time, or task feedback information.

[0036] Exemplarily, the node structure of the initial evaluation hierarchical task network provided by the embodiments of the present invention can be , and on the basis of the hierarchical task network node, it further includes evaluation relevant parameters, specifically: task subject , evaluation information , previous task , operation target , evaluation mode , task time , task feedback information .

[0037] Among them, the task subject Represents the entity performing the task, which can be a single user, a group, or an individual user belonging to a group. The task entity can be a natural person or a robot executing the target node. Evaluation information Represents the weight of the task of the target node in the overall task effectiveness evaluation. Predecessor task Represents the task that needs to be completed before this task. Operation target Represents the goal of the task execution of the target node (the last node) or the goal for effectiveness evaluation. If the target node is the last node of the execution target task, then the operation target Is the ultimate goal that the target task needs to achieve. Otherwise, if the target node is not the last node of the execution target task, then the operation target Is the goal of effectiveness evaluation. Evaluation mode Represents the evaluation method applicable to the target node. The evaluation mode is divided into task-level evaluation and action-level evaluation, or a combination of both. Task time Is an optional item. For some tasks strongly related to time, it is necessary to record the task execution time for evaluation purposes. Task feedback information Includes two types. One is the result directly fed back to the operator in real time, and the other is the score used for evaluating the overall effectiveness, and the final score is calculated after the task is completed.

[0038] Step S102. When the target node finishes execution, determine the evaluation information of the target node, where the target node is any task node in the evaluation hierarchy task network.

[0039] After the target node finishes execution, evaluate the target node to obtain evaluation information. The target node can be evaluated in the following ways.

[0040] Optionally, determining the evaluation information of the target node can be implemented in the following ways:

[0041] If the target node is a subtask node, determine the evaluation information according to one or a combination of the following evaluation methods: azimuth distance evaluation, indirect feedback evaluation, numerical comparison evaluation, or affiliation evaluation.

[0042] Azimuth distance evaluation is used to evaluate whether the operation target reaches a specified location or a specified direction according to the operation target of the target node. It can be converted into any value between 0 and 1 within the specified accuracy according to business requirements, and the conversion matrix can be linearly or non-linearly transformed according to business requirements.

[0043] Indirect feedback evaluation is used in scenarios where the evaluation result of the operation target cannot be directly obtained, and it can be evaluated by other evaluation entities according to business requirements. Indirect feedback is used to evaluate the target node based on the information provided by other entities.

[0044] The numerical comparison evaluation is used for scenarios with high numerical requirements. If the numerical value meets the requirements within the error range, it is 1; otherwise, it is 0. There is no other evaluation status in this process.

[0045] The affiliation evaluation is used to determine the status between two task objectives. If the two task objectives reach this status, it is 1; otherwise, it is 0. There is no other evaluation status in this process.

[0046] The above-mentioned implementation manner can evaluate the subtask nodes, select one or a combination of multiple evaluation methods according to requirements, and obtain the evaluation information of the subtask nodes.

[0047] Optionally, to determine the evaluation information of the target node, it can be implemented through the following method:

[0048] If the target node is an action-level node, determine the evaluation information according to one or a combination of the following evaluation methods: dynamic rule evaluation, time series evaluation, or state transformation evaluation.

[0049] The action-level evaluation mainly focuses on the early warning of abnormal states during task execution. Since the failure of each task starts from each small error, to complete the abnormal evaluation at the task level, it is necessary to start from the action-level evaluation. The action-level evaluation methods are mainly divided into three types: dynamic rule evaluation, time series evaluation, and state transformation evaluation. The corresponding outputs of the above three action-level evaluation methods are no abnormality, no impact on the failure, and unable to work properly. The specific contents corresponding to the dynamic rule evaluation, time series evaluation, and state transformation evaluation can be configured according to requirements.

[0050] The above-mentioned implementation manner can evaluate the action-level nodes, select one or a combination of multiple evaluation methods according to requirements, and obtain the evaluation information of the action-level nodes.

[0051] Optionally, to determine the evaluation information of the target node, it can also be implemented through the following method:

[0052] Step 1: Determine the input features according to the graph density of the initial evaluation hierarchical task network.

[0053] Specifically, determining the input features according to the graph density of the initial evaluation hierarchical task network can be implemented as:

[0054] Determine the graph structure according to the initial evaluation level task network; determine the graph intimacy matrix of the graph structure; obtain the context nodes of each task node; determine the original features according to the context nodes; determine the absolute role features according to the position information of the target node in the graph structure and the graph embedding information of the context nodes; determine the relative position features of the target node according to the position information of the target node in the graph structure; determine the input features according to the original features, the absolute role features and the relative position features.

[0055] Exemplarily, step S1: Convert the task model into a graph structure G.

[0056] Figure 2 Schematic diagram of process data related to the weight score provided by the embodiment of the present invention Figure 2 The red node in the figure is the target node.

[0057] Step S2: Calculate the graph intimacy matrix of the graph structure G , where is The hyperparameter between usually takes (0.15), represents the column-normalized adjacency matrix, is the adjacency matrix of the graph, is the corresponding diagonal matrix. represents the identity matrix.

[0058] Step S3: Obtain the context nodes of each node.

[0059] Assume the target node is , and the context nodes of the target node are represented as , .

[0060] Among them, the context nodes are nodes whose intimacy determined according to the graph intimacy matrix is greater than the intimacy threshold, and the intimacy threshold is .

[0061] Step S4: Calculate the original features of the node .

[0062] Among them, represents the graph embedding algorithm.

[0063] Step S5: Calculate the absolute role features of the node

[0064]

[0065] Among them is the graph kernel vector extraction method.

[0066] Is the position information of the target node in the graph structure. Represents the dimension of the node features, Represents the length of the graph.

[0067] Step S6: Calculate the relative position features of the nodes: .

[0068] Is the context node of the target node.

[0069] Step S7: Aggregate the original features, absolute role features, and relative position features to obtain the input of the LSTM .

[0070] Step Two: Input the input features into a time series neural network to obtain output information.

[0071] The optional time series neural network can be a Long Short-Term Memory (LSTM) or a Recurrent Neural Network (GRU).

[0072] Optionally, the aggregated node features Are sequentially input into the LSTM model according to the task order to obtain . For Perform convolution and normalization to obtain scores.

[0073] Step Three: Normalize the output information of multiple target nodes to obtain the weight scores of each target node, and use the weight scores as evaluation information.

[0074] The above implementation can calculate the weight scores of target nodes as evaluation information through a time series neural network such as LSTM, improving the accuracy of the evaluation information.

[0075] Step S103: Determine the pre-order task sequence of the associated nodes of the target node according to the evaluation information; determine an optimized evaluation hierarchical task network according to the pre-order task sequence, and complete the target task according to the optimized evaluation hierarchical task network.

[0076] Optionally, determining the pre-order task sequence of the associated nodes of the target node according to the evaluation information can be implemented in the following manner:

[0077] If the initial evaluation hierarchical task network is a task model in a fixed state, determine the pre-order task sequence of the associated nodes of the target node according to the evaluation information and the rule base, where the associated nodes are the child nodes or sibling nodes of the target node;

[0078] If the initial evaluation hierarchical task network is a task model with a non-fixed state, determine the key nodes according to the previous tasks of the associated nodes of the target node, and start re-verification based on the evaluation information according to the key nodes.

[0079] After task evaluation, a task sequence can be regenerated and tasks can be carried out in a rule library manner or in a manner of re-verification based on key nodes based on the task evaluation results. Among them, the rule library method is mainly for task models with a fixed state in task evaluation. As shown in the following table, the main inputs of the rule library are the states of relevant subtasks and the states of previous tasks, and the output is the tasks to be carried out; the method of re-verification based on key nodes is mainly for tasks with a non-fixed state (represented by scores) in task evaluation. It is necessary to set a task failure threshold (the task is considered failed if the score is lower than the threshold) and a key node score threshold (the task with a task weight higher than the threshold is a key node task) before the task starts. During the task execution, subtasks with task failures can be found according to the threshold, and key node tasks in the previous tasks of the task can be found in the task hierarchical network, and re-verification can start from the key node tasks.

[0080] Exemplarily, the previous task states are shown in Table 1. The current task state can be determined according to the evaluation information of the target node. Determine the evaluation information of the associated nodes according to the previous tasks of the target node, and then determine the previous task states. Determine the tasks to be carried out according to the current task state, the previous task states and the subtask states, and then adjust the previous tasks of the target node.

[0081] Table 1

[0082]

[0083] The task processing method based on the evaluation hierarchical task network provided by the embodiments of the present invention generates an initial evaluation hierarchical task network according to the target task, and the initial evaluation hierarchical task network includes a hierarchical task network; when the target node finishes execution, determine the evaluation information of the target node, where the target node is any task node in the evaluation hierarchical task network; determine the previous task sequence of the associated nodes of the target node according to the evaluation information; determine the optimized evaluation hierarchical task network according to the previous task sequence, and complete the target task according to the optimized evaluation hierarchical task network. On the basis of the hierarchical task network, evaluate the target node in the initial evaluation hierarchical task network to obtain evaluation information, adjust the previous task sequence of the associated nodes of the target node according to the evaluation information, and then optimize the initial evaluation hierarchical task network according to the evaluation information of the target node to obtain an optimized evaluation hierarchical task network, thereby improving the execution efficiency of the target task.

[0084] Figure 3Schematic diagram of a task processing method based on an evaluation hierarchical task network provided by an embodiment of the present invention, including:

[0085] Step S201, generate an initial evaluation hierarchical task network according to the target task.

[0086] Among them, the initial evaluation hierarchical task network includes a hierarchical task network. The target node of the evaluation hierarchical task network includes hierarchical task network parameters and evaluation-related parameters. The evaluation-related parameters include one or a combination of more than one of the following parameters: task subject, evaluation information, previous task, operation target, evaluation mode, task time, or task feedback information.

[0087] Step S202, when the execution of the target node is completed, determine the evaluation information of the target node, where the target node is any task node in the evaluation hierarchical task network.

[0088] Optionally, if the target node is a subtask node, determine the evaluation information according to one or a combination of more than one of the following evaluation methods: azimuth distance evaluation, indirect feedback evaluation, numerical comparison evaluation, or affiliation evaluation.

[0089] Optionally, if the target node is an action-level node, determine the evaluation information according to one or a combination of more than one of the following evaluation methods: dynamic rule evaluation, time series evaluation, or state transformation evaluation.

[0090] Optionally, calculate the evaluation information of the target node through the following method:

[0091] Step 1, determine the graph structure according to the initial evaluation hierarchical task network;

[0092] Step 2, determine the graph affinity matrix of the graph structure;

[0093] Step 3, obtain the context nodes of each task node;

[0094] Step 4, determine the original features according to the context nodes;

[0095] Step 5, determine the absolute role features according to the position information of the target node in the graph structure and the graph embedding information of the context nodes;

[0096] Step 6, determine the relative position features of the target node according to the position information of the target node in the graph structure;

[0097] Step 7, determine the input features according to the original features, the absolute role features, and the relative position features;

[0098] Step 8, input the input features into a time series neural network to obtain output information;

[0099] Step 9: Normalize the output information of multiple target nodes to obtain the weight score of each target node, and use the weight score as the evaluation information.

[0100] Step S203: If the initial evaluation hierarchical task network is a task model in a fixed state, determine the pre-task sequence of the associated nodes of the target node according to the evaluation information and the rule base, where the associated nodes are the child nodes or sibling nodes of the target node.

[0101] Step S204: If the initial evaluation hierarchical task network is a task model in a non-fixed state, determine the key nodes according to the pre-tasks of the associated nodes of the target node, and start re-verification based on the evaluation information according to the key nodes.

[0102] Step S205: Determine the optimized evaluation hierarchical task network according to the pre-task sequence, and complete the target task according to the optimized evaluation hierarchical task network.

[0103] Figure 4 It is a schematic structural diagram of a task processing device based on an evaluation hierarchical task network provided by an embodiment of the present invention. This embodiment is applicable to the situation of task processing through computer-aided devices. The computer-aided device is used to provide specific actions for intelligent devices to execute. The intelligent devices include but are not limited to unmanned aerial vehicles, intelligent robots, etc. Among them, the intelligent robot can be a machine with a robotic arm or devices required for task execution. The task processing device based on the evaluation hierarchical task network can be implemented in the form of hardware and / or software, and the task processing device based on the evaluation hierarchical task network can be configured in electronic devices such as intelligent devices, personal computers, or servers. As Figure 4 shown, the device includes: an initialization module 31, an evaluation module 32, and an optimization module 33.

[0104] The initialization module 31 is used to generate an initial evaluation hierarchical task network according to the target task, and the initial evaluation hierarchical task network includes a hierarchical task network;

[0105] The evaluation module 32 is used to determine the evaluation information of the target node when the target node finishes execution, where the target node is any task node in the evaluation hierarchical task network;

[0106] The optimization module 33 is used to determine the pre-task sequence of the associated nodes of the target node according to the evaluation information; determine the optimized evaluation hierarchical task network according to the pre-task sequence, and complete the target task according to the optimized evaluation hierarchical task network.

[0107] Based on the above embodiments, optionally, the target nodes of the evaluation hierarchical task network include hierarchical task network parameters and evaluation-related parameters;

[0108] The evaluation-related parameters include one or a combination of more than one of the following parameters: task subject, evaluation information, previous task, operation target, evaluation mode, task time, or task feedback information.

[0109] Based on the above embodiments, optionally, the evaluation module 32 is used for:

[0110] If the target node is a subtask node, determine the evaluation information according to one or a combination of more than one of the following evaluation methods:

[0111] Azimuth distance evaluation, indirect feedback evaluation, numerical comparison evaluation, or affiliation evaluation.

[0112] Based on the above embodiments, optionally, the evaluation module 32 is used for:

[0113] If the target node is an action-level node, determine the evaluation information according to one or a combination of more than one of the following evaluation methods:

[0114] Dynamic rule evaluation, time series evaluation, or state transformation evaluation.

[0115] Based on the above embodiments, optionally, the evaluation module 32 is used for:

[0116] Determine the input features according to the graph density of the initial evaluation hierarchical task network;

[0117] Input the input features into a time series neural network to obtain output information;

[0118] Normalize the output information of multiple target nodes to obtain the weight score of each target node, and use the weight score as the evaluation information.

[0119] Based on the above embodiments, optionally, the evaluation module 32 is used to determine the input features according to the graph density of the initial evaluation hierarchical task network, including:

[0120] Determine the graph structure according to the initial evaluation hierarchical task network;

[0121] Determine the graph affinity matrix of the graph structure;

[0122] Obtain the context nodes of each task node;

[0123] Determine the original features according to the context nodes;

[0124] Determine the absolute role feature based on the position information of the target node in the graph structure and the graph embedding information of the context node;

[0125] Determine the relative position feature of the target node according to the position information of the target node in the graph structure;

[0126] Determine the input feature according to the original feature, the absolute role feature and the relative position feature.

[0127] Based on the above embodiments, optionally, the optimization module 33 is used for:

[0128] If the initial evaluation hierarchical task network is a task model in a fixed state, determine the pre-order task sequence of the associated node of the target node according to the evaluation information and the rule base, where the associated node is a child node or a sibling node of the target node;

[0129] If the initial evaluation hierarchical task network is a task model in a non-fixed state, determine the key node according to the pre-order task of the associated node of the target node, and start re-verification based on the evaluation information according to the key node.

[0130] The task processing device based on the evaluation hierarchical task network provided by the embodiment of the present invention includes an initialization module 31, which is used to generate an initial evaluation hierarchical task network according to the target task, and the initial evaluation hierarchical task network includes a hierarchical task network; an evaluation module 32, which is used to determine the evaluation information of the target node when the target node finishes execution, and the target node is any task node in the evaluation hierarchical task network; an optimization module 33, which is used to determine the pre-order task sequence of the associated node of the target node according to the evaluation information; determine the optimized evaluation hierarchical task network according to the pre-order task sequence, and complete the target task according to the optimized evaluation hierarchical task network. Based on the hierarchical task network, evaluate the target node in the initial evaluation hierarchical task network to obtain evaluation information, adjust the pre-order task sequence of the associated node of the target node according to the evaluation information, and then optimize the initial evaluation hierarchical task network according to the evaluation information of the target node to obtain the optimized evaluation hierarchical task network, thereby improving the execution efficiency of the target task.

[0131] The task processing device based on the evaluation hierarchical task network provided by the embodiment of the present invention can execute the task processing method based on the evaluation hierarchical task network provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0132] Figure 5This is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. The electronic device 10 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described herein and / or claimed.

[0133] As Figure 5 shown, the electronic device 10 includes at least one processor 11, and a memory communicatively connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc. Among them, the memory stores a computer program executable by the at least one processor. The processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 into the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. The input / output (I / O) interface 15 is also connected to the bus 14.

[0134] Multiple components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0135] The processor 11 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the processor 11 include but are not limited to a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the task processing method based on the evaluation hierarchical task network.

[0136] In some embodiments, the task processing method based on the evaluation hierarchical task network can be implemented as a computer program, which is tangibly included in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the task processing method based on the evaluation hierarchical task network described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to execute the task processing method based on the evaluation hierarchical task network in any other suitable manner (e.g., by means of firmware).

[0137] The various embodiments of the systems and techniques described above in this document can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGA), application specific integrated circuits (ASIC), application specific standard products (ASSP), system-on-chip systems (SOC), complex programmable logic devices (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs, which can be executed and / or interpreted on a programmable system including at least one programmable processor, the programmable processor can be a dedicated or general-purpose programmable processor, and can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0138] The computer program for implementing the task processing method based on the evaluation hierarchical task network of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to the processor of a general-purpose computer, a dedicated computer, or other programmable data processing devices, such that when the computer program is executed by the processor, the functions / operations specified in the flowchart and / or block diagram are implemented. The computer program can be executed entirely on the machine, partially on the machine, executed partially on the machine and partially on a remote machine as an independent software package, or executed entirely on a remote machine or server.

[0139] The invention embodiments also provide a computer-readable storage medium, which stores computer instructions for causing a processor to execute a task processing method based on an evaluation hierarchical task network, and the method includes:

[0140] generating an initial evaluation hierarchical task network according to a target task, the initial evaluation hierarchical task network including a hierarchical task network;

[0141] When the target node finishes execution, determine the evaluation information of the target node, where the target node is any task node in the evaluation hierarchical task network;

[0142] Determine the pre-task sequence of the associated nodes of the target node according to the evaluation information; determine the optimized evaluation hierarchical task network according to the pre-task sequence, and complete the target task according to the optimized evaluation hierarchical task network.

[0143] Based on the above embodiments, optionally, the target nodes of the evaluation hierarchical task network include hierarchical task network parameters and evaluation-related parameters;

[0144] The evaluation-related parameters include one or a combination of the following parameters: task subject, evaluation information, pre-task, operation target, evaluation mode, task time, or task feedback information.

[0145] Based on the above embodiments, optionally, determining the evaluation information of the target node includes:

[0146] If the target node is a subtask node, determine the evaluation information according to one or a combination of the following evaluation methods:

[0147] Azimuth distance evaluation, indirect feedback evaluation, numerical comparison evaluation, or affiliation evaluation.

[0148] Based on the above embodiments, optionally, determining the evaluation information of the target node includes:

[0149] If the target node is an action-level node, determine the evaluation information according to one or a combination of the following evaluation methods:

[0150] Dynamic rule evaluation, time series evaluation, or state transformation evaluation.

[0151] Based on the above embodiments, optionally, determining the evaluation information of the target node includes:

[0152] Determine the input features according to the graph density of the initial evaluation hierarchical task network;

[0153] Input the input features into a time series neural network to obtain output information;

[0154] Normalize the output information of multiple target nodes to obtain the weight score of each target node, and use the weight score as the evaluation information.

[0155] Based on the above embodiments, optionally, determining the input features according to the graph density of the initial evaluation hierarchical task network includes:

[0156] Determine the graph structure according to the initial evaluation hierarchical task network;

[0157] Determine the graph intimacy matrix of the graph structure;

[0158] Obtain the context nodes of each task node;

[0159] Determine the original features according to the context nodes;

[0160] Determine the absolute role features according to the position information of the target node in the graph structure and the graph embedding information of the context nodes;

[0161] Determine the relative position features of the target node according to the position information of the target node in the graph structure;

[0162] Determine the input features according to the original features, the absolute role features and the relative position features.

[0163] Based on the above embodiments, optionally, determining the pre-task sequence of the associated nodes of the target node according to the evaluation information includes:

[0164] If the initial evaluation hierarchical task network is a task model in a fixed state, determine the pre-task sequence of the associated nodes of the target node according to the evaluation information and the rule base, where the associated nodes are the child nodes or sibling nodes of the target node;

[0165] If the initial evaluation hierarchical task network is a task model in a non-fixed state, determine the key nodes according to the pre-tasks of the associated nodes of the target node, and start re-verification based on the evaluation information according to the key nodes.

[0166] In the context of the present invention, a computer-readable storage medium may be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, the computer-readable storage medium may be a machine-readable signal medium. More specific examples of the machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0167] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and a pointing device (e.g., a mouse or a trackball) through which the user can provide input to the electronic device. Other kinds of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0168] The systems and techniques described herein can be implemented in a computing system including backend components (e.g., as a data server), or a computing system including middleware components (e.g., an application server), or a computing system including frontend components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system including any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected to each other by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local area network (LAN), wide area network (WAN), blockchain network, and the Internet.

[0169] A computing system can include a client and a server. The client and the server are generally remote from each other and typically interact through a communication network. The relationship between the client and the server is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system and solves the defects of difficult management and weak business scalability existing in traditional physical hosts and VPS services.

[0170] It should be understood that various forms of the processes shown above can be used, with steps reordered, added, or deleted. For example, the steps recited in the present invention can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved, and no limitation is made herein.

[0171] The above specific embodiments do not constitute a limitation on the protection scope of the present invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A task processing method based on an evaluation hierarchical task network, characterized in that: The method is applied to perform task processing by computer-assisted equipment, wherein the computer-assisted equipment is used to assist intelligent equipment in performing task actions, wherein the intelligent equipment includes a drone or an intelligent robot, and the method includes: Generate an initial evaluation hierarchical task network according to the target task, wherein the initial evaluation hierarchical task network includes a hierarchical task network, and the target task is a task that the smart device needs to complete; When the target node is executed, determining the evaluation information of the target node, including: if the target node is a subtask node, determining the evaluation information according to a combination of one or more of the following evaluation methods: azimuth distance evaluation, indirect feedback evaluation, numerical comparison evaluation or affiliation evaluation; the target node is any task node in the evaluation hierarchical task network; Determine the evaluation information of the target node, including: determine the graph structure according to the initial evaluation hierarchical task network; determine the graph intimacy matrix of the graph structure; obtain the context node of each task node; determine the original feature according to the context node; determine the absolute role feature according to the position information of the target node in the graph structure and the graph embedding information of the context node; determine the relative position feature of the target node according to the position information of the target node in the graph structure; determine the input feature according to the original feature, the absolute role feature and the relative position feature; input the input feature into the time series neural network to obtain output information; normalize the output information of multiple target nodes to obtain the weight score of each target node, and use the weight score as the evaluation information; Determine the preceding task sequence of the associated nodes of the target node according to the evaluation information; determine the optimized evaluation hierarchical task network according to the preceding task sequence, and complete the target task according to the optimized evaluation hierarchical task network.

2. The method according to claim 1, characterized in that: The target node of the evaluation hierarchical task network includes hierarchical task network parameters and evaluation related parameters; The evaluation-related parameters include a combination of one or more of the following parameters: task subject, evaluation information, preceding task, operation target, evaluation mode, task time or task feedback information.

3. The method according to claim 1, characterized in that Determine the evaluation information of the target node, including: If the target node is an action-level node, the evaluation information is determined according to a combination of one or more of the following evaluation methods: Dynamic rule evaluation, time series evaluation, or state change evaluation.

4. The method according to claim 1, characterized in that: Determining a preceding task sequence of a node associated with the target node according to the evaluation information includes: If the initial evaluation hierarchical task network is a fixed-state task model, determining a preceding task sequence of an associated node of the target node according to the evaluation information and a rule base, wherein the associated node is a child node or a sibling node of the target node; If the initial evaluation hierarchical task network is a task model in a non-fixed state, a key node is determined according to a preceding task of a node associated with the target node, and re-verification based on the evaluation information is started according to the key node.

5. A task processing device based on an evaluation hierarchical task network, characterized in that: Applicable to the case where a computer assists a device to perform task processing, the computer assists an intelligent device to perform task actions, the intelligent device includes a drone or an intelligent robot, and the method includes: An initialization module, used to generate an initial evaluation hierarchical task network according to a target task, wherein the initial evaluation hierarchical task network includes a hierarchical task network, and the target task is a task that needs to be completed by the intelligent device; An evaluation module, used for determining evaluation information of the target node when the target node is executed, including: if the target node is a subtask node, determining the evaluation information according to a combination of one or more of the following evaluation methods: azimuth distance evaluation, indirect feedback evaluation, numerical comparison evaluation or affiliation evaluation; the target node is any task node in the evaluation hierarchical task network; Determine the evaluation information of the target node, including: determine the graph structure according to the initial evaluation hierarchical task network; determine the graph intimacy matrix of the graph structure; obtain the context node of each task node; determine the original feature according to the context node; determine the absolute role feature according to the position information of the target node in the graph structure and the graph embedding information of the context node; determine the relative position feature of the target node according to the position information of the target node in the graph structure; determine the input feature according to the original feature, the absolute role feature and the relative position feature; input the input feature into the time series neural network to obtain output information; normalize the output information of multiple target nodes to obtain the weight score of each target node, and use the weight score as the evaluation information; The optimization module is used to determine the predecessor task sequence of the associated nodes of the target node according to the evaluation information; determine the optimized evaluation hierarchical task network according to the predecessor task sequence, and complete the target task according to the optimized evaluation hierarchical task network.

6. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the task processing method based on the evaluation hierarchical task network according to any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the task processing method based on an evaluation hierarchical task network according to any one of claims 1 to 4 when executed.

Images