WO2023164841A1

WO2023164841A1 - Workflow execution method and apparatus, and storage medium and program product

Info

Publication number: WO2023164841A1
Application number: PCT/CN2022/078844
Authority: WO
Inventors: 周振华; 汪艇
Original assignee: 西门子股份公司; 西门子（中国）有限公司
Priority date: 2022-03-02
Filing date: 2022-03-02
Publication date: 2023-09-07

Abstract

Disclosed in the embodiments of the present application are a workflow execution method and apparatus, and a storage medium and a program product. The method comprises: acquiring a workflow in a markup language format, wherein the workflow in the markup language format is generated on the basis of a behavior tree construction operation that is performed by a user on a graphical user interface; converting the workflow in the markup language format into a workflow topology structure that is realized by means of node link assembling; and executing the workflow topology structure. In the embodiments of the present application, the workflow logic execution is realized, and the execution of a circular logic of jumping forward and a conditional logic of jumping backward is facilitated. Automatically setting a blank node ensures that there is at least one function block node between a conditional jump node and a logic AND node, thereby avoiding logical errors.

Description

Workflow execution method, device, storage medium and program product

technical field

The embodiments of the present application relate to the field of industrial technology, and in particular, to a workflow execution method, device, storage medium, and program product.

Background technique

Workflow can be simply defined as a description of a series of operations. Workflow is widely used in automation systems, artificial intelligence, robotics and other fields. For example, the workflow of a product sorting line in an automated system can be simply described as starting, taking pictures, sorting and moving the products to the target location. The model deployment workflow in the field of artificial intelligence can be described as data collection, data labeling, model training, and model deployment.

But at present, these workflows only have text descriptions. If users want to execute such workflows, they need to follow the text descriptions and may use a variety of engineering tools. However, these tools are almost irrelevant and provide completely different user operation behaviors. It is not only a challenge to users, but also greatly increases costs, reduces efficiency, and limits flexibility due to the long development cycle. For example, in the field of artificial intelligence, users need to use a tool for data collection, manually or use other tools for data labeling, write python scripts for model training, and deploy tools for deployment.

For this reason, those skilled in the art are still working on finding other workflow solutions.

Contents of the invention

Embodiments of the present application provide a workflow execution method, device, storage medium, and program product.

In the first aspect, the embodiment of the present invention proposes a workflow execution method. The method includes:

Acquiring a workflow in a markup language format, wherein the workflow in a markup language format is generated based on a behavior tree construction operation performed by a user on a graphical user interface;

converting the workflow in the markup language format into a workflow topology implemented in Node Link Assembly;

Execute the workflow topology.

It can be seen that the embodiment of the present invention logically converts the behavior tree represented by the workflow in the language format into the workflow topology structure processed by node link compilation, so as to facilitate the execution of the workflow by the device.

In one embodiment, converting the workflow in the markup language format into a workflow topology implemented in node link assembly includes:

determining a sequential execution structure and a loop execution structure included in the workflow in the markup language format;

determining a sequential execution part of the workflow topology based on the sequential execution structure, the sequential execution part including function block nodes in the sequential execution structure and links between the function block nodes;

In the sequential execution part, at the first position corresponding to the position of the loop execution structure in the workflow in the markup language format, a first conditional jump node is added, and the first conditional jump node is based on As determined by the loop execution structure, the first conditional jump node contains a first preset condition and is linked to the first function block node before the first position, and the first conditional jump node is also based on the The execution sequence of the sequential execution part is linked to the second function block node, wherein at the first condition jump node: when the first preset condition is satisfied, jump to the first function block node; when When the first preset condition is not satisfied, execute the second function block node based on the execution sequence.

Therefore, in the embodiment of the present invention, the reverse logic flow can be realized through the conditional jump node, which facilitates the realization of various forward jumping loop logics.

In one embodiment, the conversion of the workflow in the markup language format into a workflow topology implemented in node link assembly includes:

determining a sequential execution structure and a conditional execution structure to be included in the workflow in the markup language format;

In the sequential execution part, at a second position corresponding to the position of the conditional execution structure in the workflow in the markup language format, a second conditional jump node is added, the second conditional jump node is based on As determined by the conditional execution structure, the second conditional jump node contains a second preset condition and is linked to a third function block node after the second position, and the second conditional jump node is also based on the The execution order of the sequential execution part is linked to a fourth function block node; wherein at the second condition jump node: when the second preset condition is satisfied, the fourth function block is executed based on the execution order A node; when the preset condition is not satisfied, jump to the third function block node.

Therefore, in the embodiment of the present invention, the forward logic flow can be realized through the conditional jump node, which facilitates the realization of various backward jumping conditional logics.

determining a sequential execution structure, a loop execution structure, and a conditional execution structure included in the workflow in the markup language format;

In the sequential execution part, at the first position corresponding to the position of the loop execution structure in the workflow in the markup language format, a first conditional jump node is added, and the first conditional jump node is based on As determined by the loop execution structure, the first conditional jump node contains a first preset condition and is linked to the first function block node before the first position, and the first conditional jump node is also based on the The execution sequence of the sequential execution part is linked to the second function block node, wherein at the first condition jump node: when the first preset condition is satisfied, jump to the first function block node; when When the first preset condition is not satisfied, execute the second function block node based on the execution sequence;

Therefore, in the embodiment of the present invention, the forward logic flow and the reverse logic flow are realized respectively through a plurality of conditional jump nodes, which facilitates the realization of various forward jumping loop logics and various backward jumping conditional logics.

determining a parallel execution structure to be included in the workflow in the markup language format;

At a third position in the sequential execution part corresponding to the position of the parallel execution structure in the workflow in the markup language format, at least one parallel execution part is added, the at least one parallel execution part converges to the On the parallel sink node of the sequential execution part, the parallel sink node is also linked to the fifth function block node based on the execution order of the sequential execution part; the parallel sink node is a logical AND node or a logical OR node, wherein when converged to Execute the fifth function block node when all the parallel execution parts of the logical AND node are executed; when at least one parallel execution part of all the parallel execution parts converged to the logical OR node is executed, execute the fifth function block node Describe the fifth function block node.

It can be seen that the embodiment of the present invention also implements node link assembly processing for parallel logic.

In one embodiment, also include:

When the first conditional jump node or the second conditional jump node jumps to a logical AND node, an empty node is added before the logical AND node, and the conditional jump node is jumped to the empty node, Wherein the empty node is used to indicate not to perform an operation, and the empty node is linked to the logical AND node based on the execution order of the sequential execution part.

Therefore, by automatically setting empty nodes, it is ensured that there is at least one function block node between the conditional jump node and the logic AND node, thereby avoiding logic errors.

In one embodiment, further comprising: associating and storing the workflow in the markup language format and the workflow topology; receiving a reverse interpretation request adapted to the workflow requesting the workflow topology; responding to the The reverse interpretation request is provided, and the workflow in the markup language format is provided based on the associated storage.

It can be seen that the embodiment of the present invention can quickly provide a workflow in markup language format, realize a reverse interpretation, and solve the technical problem that the workflow topology after node link assembly processing is difficult to restore to the workflow in markup language format.

In the second aspect, the embodiment of the present invention provides a workflow execution device. The unit includes:

An acquisition module, configured to acquire a workflow in a markup language format, wherein the workflow in a markup language format is generated based on a behavior tree construction operation performed by a user on a graphical user interface;

A conversion module, configured to convert the workflow in the markup language format into a workflow topology implemented in node link assembly;

An execution module, configured to execute the workflow topology.

It can be seen that the embodiment of the present invention converts the behavior tree logic represented by the workflow in the language format into a workflow topology structure after node link assembly processing, so as to facilitate the execution of the workflow by the underlying devices.

In one embodiment, the conversion module is configured to determine the sequential execution structure and the loop execution structure contained in the workflow in the markup language format; determine the sequential execution of the workflow topology based on the sequential execution structure part, the sequential execution part includes the function block nodes in the sequential execution structure and the links between the function block nodes; in the sequential execution part, corresponding to the loop execution structure in the markup language format At the first position of the workflow position, add a first conditional jump node, the first conditional jump node is determined based on the loop execution structure, and the first conditional jump node contains the first preset conditional and linked to the first function block node before the first position, the first conditional jump node is also linked to a second function block node based on the execution order of the sequential execution part, wherein in the first conditional At the jump node: when the first preset condition is satisfied, jump to the first function block node; when the first preset condition is not satisfied, execute the first function block node based on the execution order Two function block nodes.

In one embodiment, the conversion module is configured to determine the sequential execution structure and the conditional execution structure contained in the workflow in the markup language format; determine the sequential execution of the workflow topology based on the sequential execution structure part, the sequential execution part includes the function block nodes in the sequential execution structure and the links between the function block nodes; in the sequential execution part, corresponding to the conditional execution structure in the markup language format At the second position of the position of the workflow, add a second conditional jump node, the second conditional jump node is determined based on the conditional execution structure, and the second conditional jump node contains a second preset condition and linked to the third function block node after the second position, the second condition jump node is also linked to the fourth function block node based on the execution order of the sequential execution part; wherein in the second condition At the jump node: when the second preset condition is satisfied, execute the fourth function block node based on the execution sequence; when the preset condition is not satisfied, jump to the third function block node.

In one embodiment, the conversion module is configured to determine the sequential execution structure, loop execution structure and conditional execution structure contained in the workflow in the markup language format; determine the workflow topology based on the sequential execution structure The sequential execution part of the structure, the sequential execution part includes the function block nodes in the sequential execution structure and the links between the function block nodes; in the sequential execution part, corresponding to the loop execution structure in the At the first position of the workflow in the markup language format, add a first conditional jump node, the first conditional jump node is determined based on the loop execution structure, and the first conditional jump node includes The first preset condition is linked to the first function block node before the first position, and the first conditional jump node is also linked to the second function block node based on the execution order of the sequential execution part, wherein at the At the first condition jump node: when the first preset condition is met, jump to the first function block node; when the first preset condition is not met, based on the execution sequence Executing the second function block node; in the sequential execution part, at a second position corresponding to the position of the conditional execution structure in the workflow in the markup language format, adding a second conditional jump node, the The second conditional jump node is determined based on the conditional execution structure, the second conditional jump node contains a second preset condition and is linked to a third function block node after the second position, the first The second conditional jump node is also linked to the fourth function block node based on the execution sequence of the sequential execution part; wherein at the second conditional jump node: when the second preset condition is satisfied, based on the The execution sequence executes the fourth function block node; when the preset condition is not satisfied, jump to the third function block node.

In one embodiment, the conversion module is configured to determine the parallel execution structure contained in the workflow in the markup language format; in the sequential execution part, the parallel execution structure in the markup language At the third position of the position of the workflow in the format, add at least one parallel execution part, the at least one parallel execution part converges to the parallel sink node of the sequential execution part, and the parallel sink node also executes based on the sequence The execution order of the part is linked to the fifth function block node; the parallel sink node is a logical AND node or a logical OR node, wherein when all the parallel execution parts converged to the logical AND node are executed, the fifth function block node is executed A function block node: when at least one parallel execution part of all parallel execution parts converged to the logical OR node has been executed, execute the fifth function block node.

In one embodiment, the conversion module is configured to add an empty node before the logical AND node when the first conditional jump node or the second conditional jump node jumps to the logical AND node, and convert the The conditional jump node jumps to the empty node, wherein the empty node is used to indicate that no operation is performed, and the empty node is linked to the logical AND node based on the execution sequence of the sequential execution part.

In one embodiment, also include:

The reverse interpretation module is used to associate and store the workflow in the markup language format and the workflow topology; receive a reverse interpretation request adapted to the workflow requesting the workflow topology; respond to the reverse A workflow in the markup language format is provided based on the associative store to an interpretation request.

In a third aspect, the embodiment of the present invention proposes a workflow execution device, including: at least one memory configured to store computer-readable codes; at least one processor configured to invoke the computer-readable codes to execute the above-mentioned A step in the workflow execution method.

In a fourth aspect, the embodiment of the present invention provides a computer-readable medium, wherein computer-readable instructions are stored on the computer-readable medium, and when the computer-readable instructions are executed by a processor, the The processor executes the steps in any one of the workflow execution methods described above.

In the fifth aspect, the embodiment of the present invention proposes a computer program product, the computer program product is tangibly stored on a computer-readable medium and includes computer-readable instructions, and when executed, the computer-readable instructions cause at least one The processor executes the steps in any one of the workflow execution methods described above.

Description of drawings

FIG. 1A is an exemplary flow chart of workflow creation methods provided by various embodiments of the present application.

FIG. 1B is an example diagram of a resource knowledge graph in an example of the present application.

FIG. 1C is an example diagram in which resource nodes are associated with function block nodes in an example of the present application.

FIG. 2A is a schematic diagram of a behavior tree created in an example of the present application.

FIG. 2B to FIG. 2S are respectively schematic diagrams of partial behavior trees created in an example of the present application.

FIG. 2T is an example of building a behavior tree based on the function block nodes of the function block type diagram in an example of the present application.

2U to 2X are schematic diagrams of hiding or displaying data input ports or data output ports in an example of the present application, respectively.

FIG. 2Y is a schematic diagram of adding data blocks to four function block nodes in the behavior tree and displaying corresponding data in each data block in an example of the present application.

Fig. 3 is an exemplary structural diagram of a workflow creation system provided by an embodiment of the present application.

4A-4D are schematic diagrams of workflow creation systems provided by various embodiments of the present application.

FIG. 4E is a diagram of an application scenario of the OT domain low-code development platform 100 in the field of industrial automation.

FIG. 5 is a schematic diagram of hardware implementation of the workflow creation system provided by the embodiment of the present application.

FIG. 6 is an exemplary flowchart of a workflow control method provided by various embodiments of the present application.

FIG. 7A is a first schematic diagram of a low-code example of an FBTD logic synthesizer provided by various embodiments of the present application.

FIG. 7B is a second schematic diagram of a low-code example of the FBTD logic synthesizer provided by various embodiments of the present application.

FIG. 7C is a schematic diagram of a low-code example of a folded FBTD logic synthesizer provided by various embodiments of the present application.

FIG. 8 is an exemplary flowchart of a workflow execution method provided by an embodiment of the present application.

FIG. 9A is a first schematic diagram of a workflow topology implemented by node link assembly provided by an embodiment of the present application.

FIG. 9B is a second schematic diagram of a workflow topology implemented by node link assembly provided by the embodiment of the present application.

FIG. 9C is a third schematic diagram of the workflow topology implemented by node link assembly provided by the embodiment of the present application.

FIG. 9D is a topological structure diagram of a workflow with a parallel execution mechanism implemented by node link assembly provided by the embodiment of the present application.

FIG. 10 is an exemplary structural diagram of a workflow execution device provided by various embodiments of the present application.

FIG. 11 is an exemplary structural diagram of a workflow execution device with a memory-processor architecture provided by various embodiments of the present application.

Wherein, the reference signs are as follows:

Detailed ways

The subject matter described herein will now be discussed with reference to example implementations. It should be understood that the discussion of these implementations is only to enable those skilled in the art to better understand and realize the subject matter described herein, and is not intended to limit the protection scope, applicability or examples set forth in the claims. Changes may be made to the function and arrangement of the discussed elements without departing from the protection scope of the content of the embodiments of the present application. Various examples may omit, substitute, or add various procedures or components as needed. For example, the methods described may be performed in an order different from that described, and various steps may be added, omitted, or combined. In addition, features described with respect to some examples may also be combined in other examples.

As used herein, the term "comprising" and its variants represent open terms meaning "including but not limited to". The term "based on" means "based at least in part on". The terms "one embodiment" and "an embodiment" mean "at least one embodiment." The term "another embodiment" means "at least one other embodiment." The terms "first", "second", etc. may refer to different or the same object. The following may include other definitions, either express or implied. Unless the context clearly indicates otherwise, the definition of a term is consistent throughout the specification.

Embodiments of the present application will be described in detail below in conjunction with the accompanying drawings.

FIG. 1A shows an exemplary flow chart of a method for creating a workflow provided by an embodiment of the present application. As shown in Figure 1A, the method may include the following steps:

Step S11, receiving a behavior tree construction operation performed by the user on the graphical user interface based on the preset behavior tree nodes.

In this embodiment, a behavior tree is used to represent a workflow, and the workflow is used to define operations to be performed by a work unit, for example, it may represent a distributed process within a work unit. During specific implementation, the workflow here can also be subdivided into main workflow and sub-workflow. The main workflow is used to define the start, end, and other flow controls that trigger the flow of the entire unit of work. The main workflow is the entry point of the entire process, which is linked to at least one sub-workflow. Sub-workflows are usually main workflows, and each sub-workflow corresponds to a sub-process for implementing specific business operations.

Wherein, a work cell (Workcell) may be a combination of resources such as systems or devices capable of implementing a relatively complete and independent control process and operation. In this embodiment, the workflow is created with the work unit as the basic unit, which is more in line with the characteristics of industrial control, can improve the integration degree of development, and reduce the complexity of development. For example, taking the field of industrial technology as an example, the work unit can be defined according to the actual industrial scene, for example: you can define a process corresponding to a work unit, or you can define a work station in the process as a work unit, or you can also define A station in the workstation corresponds to a work unit, etc. Different work units have different technological processes.

Behavior trees are a formalized graphical modeling approach. Behavior trees are widely used for intelligent decision-making in various artificial types. Since most of the logic is set according to the rules, the judgment of behavior is like a tree with many judgment branches. The behavior of the behavior tree is on the leaf node, which is the behavior logic to be executed. Execute each judgment branch from top to bottom, execute to the leaf node, and return the final judgment to the successful behavior node for logical execution, so as to realize the decision. This is the basic principle of the behavior tree. Through the behavior tree approach, well-defined symbols can be used to clearly express the relevant requirements of the software integration system. The structure of the behavior tree is organized in the form of a tree, each node has a corresponding node type, and carries different functions with related parameters. You can use the component editor to provide users with an intuitive and visual behavior tree editor. With the Behavior Tree Editor, users can quickly edit behavior trees. For example, after starting the Behavior Tree Editor, you can create a new Behavior Tree, and then select the Behavior Tree node type to enter the Behavior Tree, where you can also edit the properties of the Behavior Tree nodes that make up the Behavior Tree.

In this embodiment, the behavior tree nodes may include: flow control nodes, function block nodes, and so on. Each of them will be described in detail below.

1. Flow control node

Flow control nodes are used to implement logical control in workflows, which are usually independent of specific business operations in work units. Through flow control nodes, users can create various workflows according to their needs.

During specific implementation, the flow control node may include: a main control node, a compositor node (also called an aggregator node or a logic control node), and a condition node. Alternatively, the flow control node may also include: one or any combination of a main control node, a synthesizer node and a condition node. They are briefly described below.

1. Master control node

The main control node may include: some or all of a Start node, an End node, a Goto node, an Anchor node, a Stop node, and an Abort node.

The main control nodes in this embodiment are not standard behavior tree elements, but they can be used to control the main flow of the workflow in this embodiment and can be linked to the state machine of the workflow. Wherein, the start node is mandatory, and one of the end node or the go-to node can also be mandatory. The main control node is mainly used to define the start and end of the workflow. In addition, other element nodes can control the state machine (such as abort and stop) or mark the key process steps that can be jumped (such as critical nodes).

2. Synthesizer node

Compositor nodes are used to implement logic control in workflows. The synthesizer node can include: sequence (Se, Sequence) node, reactive sequence (RSe, Reactive Sequence) node, parallel (Pa, Parallel) node, process quality control (IPQ, In-Process QC) node, priority ( Pr, Priority (Fallback)) node, reactive priority (RPr, Reactive Priority (Fallback)) node, whether conditional judgment (ITE, If-Then-Else) node, and multi-branch selection (Sw, Switch) node some or all.

Compositor nodes define how branches are executed in the Behavior Tree, are used to implement branching logic control in workflows, and more. For example, here's a brief description of a typical synthesizer node:

1) Sequence node, which can have one or more sub-nodes, and will trigger routing to each sub-node in sequence according to the workflow sequence.

2) Reactive sequence node, which has the same function as the sequence node, but will continuously check the trigger conditions.

3) Parallel node, which can have one or more sub-nodes, starts up sequentially from top to bottom, and executes all sub-nodes in multi-process or multi-thread simultaneously.

4) The quality control node in the process is that the quality engineer performs quality inspection on a part of the process steps. If the inspection fails, the exception handling process will be executed, and if it succeeds, it will continue.

5) Priority node: Execute each sub-branch in order according to the priority. If the previous execution fails, the next one will be executed, and any branch will pass if the execution is successful.

6) Reactive priority node: The function is the same as the priority node, but the trigger condition will be continuously checked.

7) Whether condition judgment node: check the trigger expression, if it is true, execute the true branch, if it is false, execute the false branch.

8) Multi-branch selection node: check the trigger expression, and execute different branches according to different conditions, all of which are not satisfied to execute the default (default) branch.

Typically, sequential and parallel nodes drive most of the logic in a workflow.

3. Condition (C, Condition) node

The condition node is usually the basic logical element of the check expression in the behavior tree, which is used to execute the condition judgment and return the judgment result. It returns success or failure depending on whether the condition is true. Conditional nodes never return a running state.

In addition, in other implementation manners, a conditional node may also be included in a function block node. Alternatively, conditional nodes can also be used as a single type of node.

2. Function block (FB, Function Block) node

Function block nodes are used to execute commands and implement business operations in the workflow. Typically, success is returned if the operation completed correctly, and failure is returned if the operation failed. Returns running while the operation is in progress.

In this embodiment, the function block nodes include logic nodes, and may also include some specific types of function block nodes, such as manual (Manual) nodes, dynamic (Dynamic) nodes, delay (Delay) nodes, and idle (Empty ( idle)) some or all of the nodes. Among them, dynamic node is used to dynamically inject node instance at runtime. The manual node represents a manual step, stop at the current node before obtaining the confirmation signal, and exit after obtaining the confirmation signal. Delay node means to exit the current node after a specified time delay. An idle node means no operation is performed, a placeholder, and can be replaced by any function block node. Wherein, each logical node may correspond to an operation template, and each operation template predefines operations that at least one type of resource (such as a collaborative robot or a device such as a PLC) can perform. For example, the operations may include: actions, methods or skills.

During specific implementation, each operation template may consist of an interface part and an implementation part. The implementation part can be an application (for example, a containerized application) that contains functional code and runtime dependencies. The application program can run independently and be exposed to the outside world through a specific network communication interface. The interface part may be a logical node presented as a graphical element, that is, it can be dragged and dropped, connected and configured in a graphical user interface like other behavior tree nodes. During specific implementation, each logical node may have a parameter panel for configuring parameters of the logical node, such as input and output parameters. Of course, these input and output parameters can also be preset with default values.

Each logical node can be configured and executed individually. When executing a logical node in the Behavior Tree, the user-configured input for the logical node is read and transferred to the implementation part of the action template, the corresponding application. After completing a specific operation such as model transformation, the result of the operation such as the transformed model will be converted back to the output of the logical node.

In this embodiment, the interface part and implementation part of each operation template can be stored separately. For example, only the interface part, that is, the logical node, may be stored in the node library, and its implementation part may be stored in a node service module, which may be called a runtime (Runtime) in this embodiment. The node service module can be located in a server or locally.

Optionally, the above-mentioned logical nodes follow an information model for runtime interaction with the main controller. In this way, the standardization of communication between the OT domain workflow and the main controllers of various OT devices is realized.

In addition, considering that the business operations corresponding to a function block node may be performed by different subjects, such as a specific physical device, a person, or other virtualized noun resources, for the convenience of description , these physical devices, personnel, and virtualized noun resources are collectively referred to as resources in this article, and these resources usually refer to resources that can execute workflows on site as the main body of each operation.

In the specific implementation, the resource as the main body of operation can be used as a common configuration parameter of the function block node to configure the corresponding resource for the required function block node when creating the behavior tree; or, when creating the function block node, set the function Block nodes are configured with resources as the subject of operation execution, so that when creating a behavior tree, it is not necessary to configure resources for the required function block nodes. Or, in order to facilitate the management of resources, these resources can also be represented in the form of resource nodes, and these resource nodes can be stored in the form of resource knowledge graphs. The resource knowledge graph includes: various resource nodes, and connections representing relationships among resource nodes. For example, an example diagram of a resource knowledge graph in an example is shown in FIG. 1B . As shown in FIG. 1B , the resource knowledge graph is a factory resource knowledge graph, which describes the real system configuration of a factory. The factory (F, Factory) node has an industrial computer (IPC) node, and the industrial computer (IPC) node has a collaborative robot (CR, Cooperative Robot) node, a PLC node and a bar code scanner (BCS, Bar Code Scanner) node. Among them, the collaborative robot (CR, Cooperative Robot) node has a gripper (CJ, Clamping Jaw) node, a torque wrench (TW, Torque Wrench) node and a camera (CA, Camera) node; the PLC node has a button (B, Button) node and LED warning light nodes. Of course, in some applications, there may be more than one type of equipment such as a collaborative robot, and each equipment of this type of equipment can be distinguished by a label or model, which will not be repeated here.

Each function block node can be associated with a resource node, and the function block node can be instantiated as an operation of the corresponding resource. For example, a logical node can be instantiated as an operation of a corresponding device by associating a device resource node. During specific implementation, a resource header (Header) may be set for the function block node, and resources associated with the function block node will be displayed in the resource header.

The specific association process may be implemented in many different manners. For example, each resource node can be associated with a corresponding function block node in advance, so that when creating a behavior tree, the function block node associated with the corresponding resource can be directly pulled, and there is no need for temporary configuration. For example, FIG. 1C shows an example diagram in which each resource node is associated with a function block node. As shown in FIG. 1C , for the resources shown in the resource knowledge graph shown in FIG. 1B , corresponding function block nodes are respectively associated. Among them, the industrial computer (IPC) node is associated with a button (PB, Press Button) node and a dialog box (DDB, Display Dialog Box on Screen) node; the collaborative robot (CR, Cooperative Robot) node is associated with a linear movement (LM, Linear Move) nodes and joint movement (SM, Shutdown Mobile) nodes; PLC nodes are associated with read I/O (RIO, Read I/O) nodes and write I/O (WIO, Write I/O) nodes; The bar code scanner (BCS, Bar Code Scanner) node is associated with the scanning bar code (SBC, Scan Bar Code) node; the clamping jaw (CJ, Clamping Jaw) node is associated with the open (O, Open) node and the grabbing (Gr, Grab) node Node; torque wrench (TW, Torque Wrench) node is associated with twist (T, Twist) node; camera (CA, Camera) node is associated with registration (R, Register) node, calibration (Cb, Calibration) node, camera (TP, Take Photo) node, object recognition (OR, Object Recognition) node; button (B, Button) node is associated with button (PB, Press Button) node; LED alarm light node is associated with switch open (SO, Switch On) node and Switch off (SF, Switch Off) node.

Alternatively, resource nodes may not be pre-associated with function block nodes, but corresponding resource nodes may be associated with required function block nodes when the behavior tree is created.

Alternatively, there may also be function block nodes associated with resource nodes (which may be called dedicated function block nodes) and function block nodes not associated with resource nodes (which may be called general function block nodes). Because although the common function block node is not associated with on-site resources, it does not affect the simulation of the workflow corresponding to the behavior tree including the function block node. For example, a collaborative robot may not be purchased yet. In order to verify the feasibility, the corresponding general function block node can be used for simulation in advance. When it is confirmed that it is feasible, it can be purchased.

In addition, this embodiment may further include the following decorator nodes:

3. Decorator node

Decorator nodes are mainly used to decorate function block nodes driven by compositor nodes, for example, can be used to decide whether a branch or even a single node in a behavior tree can be executed. Can include: repeat (Rp, Repeat) node, retry (Rt, Retry) node, one-time (OS, One-Shot) node, timeout (TO, Timeout) node, timer (Tm, Timer) node, negation (Iv, Inverter) node, forced operation (FR, Force Run) node, forced success (FO, Force OK) node, forced failure (FF, Force Failed) node, and some or all of monitoring (G, Guard) nodes . The following is a brief description of some of the decorator nodes:

1) Negative node, which can have a child node, used to negate the child node. If the child node fails, it returns success; if the child node succeeds, it returns failure.

2) Mandatory success node, which can have a child node (such as a function block node), no matter whether its child node is successful or not, the node will always return success.

3) A mandatory failure node, which can have a child node (such as a function block node), regardless of whether its child node is successful or not, the node will always return failure.

4) A repeated node, which may have a child node (such as a function block node), can repeatedly execute its child node for a fixed number of times.

5) A retry node, which can have a child node (such as a function block node), can trigger its child node up to N times, if its child node returns failure, retry and the number of times will be reduced by one, when the number of retries is zero , returns success; if the child node returns success, break the loop and also return success. Wherein, N is a positive integer.

6) A one-time node, which may have a child node, means that its child node is only executed once in the workflow, and will not be executed until the workflow is restarted.

7) A timeout node, which may have a child node, which is used to time the execution time of its child nodes (such as function block nodes), and exit execution when the specified time exceeds (even if the execution is not completed).

8) A timer node, which may have a child node, executes its child node (such as a function block node) after the specified time is reached.

9) Forcibly run the node, which can have a child node, regardless of whether its child node (such as a function block node) is executed or not, it will be forced to return to the running state.

10) A monitoring node, which may have at least one child node, and is used to monitor the status of all its child nodes. When any child node executes incorrectly, an error will be reported. When all child nodes are normal, it returns normal.

In other implementations, a decorator node may also be contained within a flow control node. That is, flow control nodes may include: all or part of main control nodes, synthesizer nodes, and decorator nodes.

In this embodiment, each behavior tree node can be listed on the graphical user interface in the form of an icon, and the user can determine the node needed to create a workflow by selecting and dragging the icon to add to the canvas. The node performs necessary parameter configuration, such as resource configuration and/or input and output parameter configuration. Among them, if there is more than one operation to be performed by a work unit, that is, more than one operation defined in the required workflow, the behavior tree corresponding to the workflow can include multiple function block nodes. For the corresponding flow control node, the behavior tree of the corresponding workflow is finally generated by performing corresponding discharge connections on the dragged nodes. That is, the behavior tree construction operation includes adding and connecting nodes of the behavior tree. Further, an operation of associating resources for the added function block nodes may also be included. In addition, it may also include: configuration operations on the input and output parameters of the behavior tree nodes.

Step S12, in response to the construction operation of the behavior tree, generate a behavior tree corresponding to a workflow, and the logical nodes in the behavior tree are instantiated as operations of corresponding equipment.

In this embodiment, each behavior tree node may be instantiated in response to the construction operation of the behavior tree, and a connection relationship between each instantiated behavior tree node may be established. For example, by performing this step, the added logical node can be instantiated as an operation of the corresponding device. Then, based on the connection relationship between the instantiated behavior tree nodes, a behavior tree corresponding to a workflow is generated.

During specific implementation, the aforementioned behavior tree nodes may be stored in a node library. In addition, for similar application scenarios, in order to avoid the waste of manpower and time for repeatedly building behavior trees, it is best for the user to build a behavior tree corresponding to a workflow or a corresponding sub-workflow that has been debugged or successfully run (or not The instantiated behavior tree framework) is stored as a workflow node or sub-workflow node. Correspondingly, the node library may further include a workflow (WF, WorkFlow) node and a sub-workflow (SWF, SubWorkFlow) node. When the user needs to build a similar behavior tree or build a behavior tree including the workflow or sub-workflow, he can select the corresponding workflow node or sub-workflow node and perform necessary configurations to obtain the required A behavior tree for a workflow.

FIG. 2A shows a schematic diagram of constructing a behavior tree of a work unit of a quality inspection production line in an example. FIG. 2B to FIG. 2S respectively show schematic diagrams of partial behavior trees constructed in an example. In the process of constructing the behavior tree representing the workflow described above, the function block nodes can be understood as presented in the form of a label graph, and the construction of this behavior tree based on the function block label graph requires the participation of flow control nodes and even decorator nodes.

In addition, an embodiment of the present invention also provides a method for constructing a behavior tree based on a function block type diagram. In the method, the function block nodes are presented in the form of a type diagram. In practical applications, these two behavior tree construction methods (that is, the behavior tree construction method based on the function block type diagram and the behavior tree construction method that presents the function block nodes in the label diagram) can coexist, and one of them can be used When building behavior trees, another method is also built synchronously. For example, when two function block nodes are sequentially connected based on the function block type graph, when synchronously constructing a behavior tree based on the function block label graph, a sequence node will be automatically added, and the sequence node will be added from top to bottom. A function block node.

The behavior tree construction method based on the function block type diagram may include the following steps (1) and (2). in:

Step (1): receiving the addition and connection operations of function block nodes performed by the user on the graphical user interface based on the function block type diagram. In this application, the type diagram and the label diagram can be understood as two presentation modes of the same function block node, and their functions are both used to realize a corresponding business operation.

FIG. 2T shows an example of constructing a behavior tree based on the function block nodes of the function block type graph in an example. In this example, two function block nodes F1, F2 are shown. As shown in Figure 2T, the functional block type diagram may include:

1) A function block name 201 used to indicate the type of business operation. For example, screw fastening, image capture, video recording, visual guidance, etc. In specific implementation, when dragging a function block node from the node library to the canvas, the default function block name will be generated by adding the default name of the function block definition plus the instance number. In Figure 2U, the logical node in the function block node is taken as an example , the names are represented by "logical node 1" and "logical node 2". Function block names are editable, but should be unique within the current global workflow.

2) The function block header 202 used to indicate the resource for executing the business operation. For example, the resource for tightening screws is a torque wrench, the resource for image collection is a monitor, the resource for video recording is a monitor, and the resource for visual guidance is a camera. Normally, the functional block header 202 of a logical node is generally a physical resource, such as "physical node 1" and "physical node 2" as shown in FIG. 2U. As mentioned above, the resource indicated by the function block header 202 may be pre-associated, or the corresponding resource may be associated through resource configuration after the function block node is added.

3) A link input port 203 and a link output port 204 for triggering link connections. In this embodiment, the connection operation of the function block nodes may include: a connection operation between the link output port 204 and the link input port 203 between the two function block nodes F1 and F2. Through this wiring operation, a link connection 205 is established between the two function block nodes F1 and F2. The link connection 205 in this example is a one-way connection used to indicate the running process of the workflow. For example, it can indicate the execution order of the two function block nodes, that is, the function block node F1 on the left is executed first, and then the function block node on the right is executed. F2.

During specific implementation, in order to avoid click errors and improve the sensitivity of the connection of function block nodes, there is a sensitive area 206 within the setting range of the link input port 203, which is called the first sensitive area here, and is used to When the user's click and connection operations are received in the sensitive area, the connection endpoint is positioned on the link input port 203 . Similarly, there may also be a sensitive area (not shown in the figure) within the setting range of the link output port 204, which is referred to as the second sensitive area herein, and is used to receive the user's point in the second sensitive area. During the selection and connection operation, the connection endpoint is positioned on the link output port 204.

In addition, in this embodiment, in response to the connection operation between the link output port 204 and the link input port 203 between the two function block nodes F1 and F2, the interconnected function block nodes F1 and F2 can be further An instruction label 207 for indicating the execution order of each function block node is generated, and the instruction label 207 is marked on the function block type diagram of the function block nodes F1 and F2. 1 and 2 as shown in the picture. In addition, the instruction tag 205 can also be used as an index of jump instructions and as a chapter index of the documentation.

4) Input data chunk 208 for representing the set of all data input ports and output data chunk 209 for representing the set of all data output ports. In this embodiment, the number of data input ports of the function block node can be marked on the input data block 208, such as 5 on the input data block 208 of the function block node F1 and the input data block 208 of the function block node F2 2 above, if the number of data input ports is zero, the input data chunk 208 can be hidden. Similarly, the number of data output ports of the function block node can be marked on the output data block 209, such as 3 on the output data block 209 of the function block node F1 and 1 on the output data block 209 of the function block node F2 , if the number of data output ports is zero, the output data chunk 209 can be hidden. In this embodiment, as shown in FIGS. 2U to 2X , each data input port of the function block node can be expanded or hidden by clicking on the input data block 208 . Each data output port of the function block node can be expanded or hidden by clicking on the output data block 209 . Among them, 2U shows a schematic diagram of hiding the data input port of logical node 2 by clicking, 2V shows a schematic diagram of hiding the data output port of logical node 1 by clicking, and FIG. 2W shows that each data input port is displayed simultaneously. and a schematic diagram of each data output port, FIG. 2X shows a schematic diagram of simultaneously hiding each data input port and each data output port.

5) Data input port 210 and data output port 211 for triggering data transmission. In this embodiment, the connection operation of the function block nodes may also include: a connection operation between the data output port 211 and the data input port 210 between the corresponding data of the two function block nodes F1 and F2. Through this wiring operation, a data connection 212 is established between the two function block nodes F1 and F2. Data connection 212 in this example is used to indicate data transmission between two function block nodes F1 and F2. As shown in Figure 2U, a data connection 212 is established between the data output port 211 of the output data 1 of the function block node F1 and the data input port 210 of the input data 2 of the function block node F2, which means that the output of the function block node F1 Data 1 is used as input data 2 of the function block node F2. During specific implementation, similar to the link input/output port, in order to avoid point selection errors and improve the sensitivity of the connection of function block nodes, a sensitive area (not shown in the figure) can also be provided within the set range of the data input port 210 output), which is referred to as the third sensitive area here, and is used to locate the connection end point on the data input port 210 when the user's click and connection operations are received in the third sensitive area. Similarly, there may also be a sensitive area (not shown in the figure) within the setting range of the data output port 211, which is referred to as the fourth sensitive area here, and is used to receive the user's click in the fourth sensitive area. And during the connection operation, the connection endpoint is positioned on the data output port 211 .

6) The function block icon 213, specifically, the function block icon 213 may be a vector icon 213, which is used to visually represent the service operation of the function block node. Of course, in other implementation manners, the function block icons 213 may also be omitted.

7) Function block main body 214 for carrying the above components. In this embodiment, the adding operation of the function block node may include: dragging and dropping the function block main body 214 .

Step (2): In response to the addition and connection operations of the function block nodes, construct a behavior tree corresponding to a workflow.

In step (2), in response to the addition of the function block node and the connection operation, the behavior tree including the flow control node and the function block node based on the function block label graph in S12A can be constructed synchronously; similarly, in step S12A , when constructing the behavior tree including the flow control node and the function block node based on the function block label diagram, the behavior tree based on the function block type diagram in S12B may also be constructed synchronously. And the two behavior tree construction interfaces can be switched according to the user's choice. For example, according to the user's click operation on the behavior tree based on the function block type diagram or the behavior tree based on the function block label diagram, the behavior tree based on the function block type diagram or the function block label diagram based The Behavior Tree toggles the display.

Further, in other embodiments, no matter whether the behavior tree is created based on the function block node in the form of label diagram or the function block node in the form of type diagram, at least one of the function block nodes in the behavior tree can be further Add and connect at least one data block to each function block node. Wherein, each data block is used to present the corresponding data in the business operation of the function block node connected to it. The types of data blocks can include some or all of data pairs, data tables, images, videos, charts, etc.

FIG. 2Y shows a schematic diagram of adding data blocks to four function block nodes in the behavior tree and displaying corresponding data in each data block in an example of the present application. In this example, take the behavior tree creation interface based on the function block type diagram as an example, as shown in Figure 2Y, in this example, the function block node associated with the monitor to record the production video is added with a type of real-time video data (Live Data -Video) data block; the function block node associated with the screw fastening of the torque wrench is added with a data block of type Live Data-Text and a data block of type Live Data-Chart ; The function block node of the robot image acquisition associated with the monitor is added with a data block of type Live Data-Image; the function block node of the machine vision guide associated with a camera is added with a type of Live Video Data (Live Data-Video) data block.

As shown in FIG. 2Y, each data block may include a data block label 215 for indicating the type of data block and a data block body 216 having a display area for presenting specific data. Wherein, the data block label 215 can be a draggable label, for example, can be moved to any position in the canvas. The size of the display area of the data block 216 is adjustable, and is used to display different types of data from the data layer in real time. A monitoring link 217 is established between each data block and the corresponding function block node, and one function block can be mapped to multiple data blocks. When the workflow is executed, for example, at runtime, the monitoring and output data corresponding to the function block node will be transferred to the corresponding data block for real-time display.

The data block in this embodiment is a low-code data block, which is different from other SCADA and dashboard systems in that the low-code data block in this embodiment is also a low-code element, which can be used as a part of the behavior tree, and all of its Properties can all be managed in low-code logic. Among them, the data source in the low-code data block comes from the data layer, which can obtain data through the interface provided by the data layer at runtime or in the cloud execution engine. The data source can be a time series database, RDBMS or NoSQL. Low-code data block is a flexible, scalable and adaptable system that can be applied to any functional block node that can be associated with physical devices. In short, the data blocks used to implement data monitoring can be considered as the visual interface of the data layer.

As shown in FIG. 2Y, each data block may include a data block label 215 for indicating the type of data block and a data block body 216 having a display area for presenting specific data. Wherein, the data block label 215 can be a draggable label, for example, can be moved to any position in the canvas. The size of the display area of the data block 216 is adjustable, and is used to display different types of data from the data layer in real time. A monitoring link 217 is established between each data block and the corresponding function block node, and one function block can be mapped to multiple data blocks. When the workflow is executed, for example, at runtime, the monitoring and output data corresponding to the function block nodes will be transmitted to the corresponding data blocks for real-time display.

In addition, in other implementation manners, this embodiment may further include the following step S13 as shown by the dotted line in FIG. 1 .

Step S13, analyzing the behavior tree, and deploying the workflow corresponding to the behavior tree to the runtime of the corresponding work unit, so that each resource in the work unit performs operations according to the workflow.

In a specific implementation, the work unit may have a main controller, and in this case, the runtime may be located on the work unit's main controller. Correspondingly, the equipment resources in the resources can be connected to the main controller, and the main controller can control the equipment resources connected to it to perform corresponding operations according to the runtime workflow; the human resources in the resources can be directly based on the running The workflow at the time prompts you to perform the appropriate action.

In this embodiment, the behavior tree can be stored in a Markup markup language such as XML (Extensible Markup Language), and can be verified by an XML Schema (XSD) prototype to verify that the XML format of the behavior tree is correct. After parsing the behavior tree, a workflow expressed in the form of node link assembly can be obtained, and then the workflow can be compiled and downloaded to the runtime of the main controller of the corresponding working unit.

In addition, according to Gartner's definition, the OT domain usually refers to operational technology (Operational Technology, OT), which integrates hardware and software, and detects or triggers processes in the enterprise by directly monitoring and/or controlling physical devices (called OT devices). changes or events. OT utilizes computers to monitor or change physical conditions such as industrial control systems (Industrial Control System, ICS). Among them, the industrial control system is based on computer-implemented facilities, systems, and equipment for remote monitoring and/or control of key industrial processes to achieve physical functions. The word "OT" is used to distinguish industrial control systems from traditional information technology (Information Technology, IT) systems in terms of technical implementation and functions.

At present, there are many IT low-code development tools or platforms on the market. Among them, some tools are aimed at the use scenarios of the Internet of Things and are aimed at experienced IT engineers, while OT engineers and junior IT engineers have difficulty understanding their paradigms. Some tools are more suitable for low-code development scenarios in the IT domain, but not well suited for the OT domain.

The above workflow creation method in this embodiment can be used in this OT domain as a low-code development method in the OT domain. Specifically, the workflow creation method shown in FIG. 1A can be implemented in an OT domain, such as an OT domain low-code development platform. Correspondingly, the workflow can be an OT domain workflow; the work unit can be an OT domain workflow. A working unit; the device may be an OT device. Here, OT devices may include, but are not limited to: Internet of Things (IoT) devices, Programmable Logic Controllers (PLC), Robotics, Manual Process, Industrial Computers (Industrial Personal Computer, IPC) and so on.

In addition, in the embodiment of the present application, it is considered that the integration of the IT domain and the OT domain is becoming increasingly important in the process of enterprise digital transformation. In order to integrate the IT domain and the OT domain into an ITOT system, an urgent problem to be solved at present is how to collect data in the OT domain and control the process of the OT domain in a way that is easy to understand instead of programming in the IT domain.

The above-mentioned workflow creation method in this embodiment can be used in this ITOT system as a low-code development method in the OT domain that can be integrated with the IT domain. Similarly, the workflow creation method shown in Figure 1A can be implemented on an OT domain such as an OT domain low-code development platform, and correspondingly, the workflow can be an OT domain workflow; the work unit can be a work in the OT domain unit.

In addition, in order to realize the integration of the IT domain and the OT domain, on the basis of the workflow creation method shown in Figure 1A, it may further include: generating a microservice based on the behavior tree, so that an IT device can call the microservice The runtime execution of the master controller of the work unit is triggered to execute the OT domain workflow. During specific implementation, the IT equipment can call the microservice directly or through a knowledge center.

When generating a microservice based on the behavior tree, an API of the microservice may be generated based on the behavior tree. Wherein, the processing procedure in the API includes each operation in the OT domain workflow, the input parameter of the API is the parameter obtained from the input port of the OT domain workflow, and the output parameter of the API is the Parameters output by the output port of the OT domain workflow.

To realize that the IT domain can call the microservice, the IT domain needs to be able to obtain the information of the microservice. The specific implementation methods include but are not limited to the following two:

method one,

In method 1, the code developers in the OT domain can notify the code developers in the IT domain of the names and IP addresses of the generated microservices. In this way, the code developers in the IT domain can directly assign the The information is written into the code, so that the IT equipment can call the microservice. Method 1 is more suitable for scenarios with a small number of microservices.

Method two,

In mode 2, registration and discovery mechanisms can be used. That is, each microservice is registered on the knowledge center platform, so that an IT domain code development tool realizes that IT equipment discovers the connected microservices through the knowledge center platform. In the specific implementation, an IT domain code development tool can be used to realize the connected microservices discovered by the IT domain equipment through the knowledge center through code development. The device that completes microservice registration can be an OT domain microservice generator or a third-party device. The third-party device can be regarded as a part of the OT domain low-code development platform, or implemented in the knowledge center platform. Method 2 is more suitable for scenarios with a large number of microservices.

In this embodiment, IT equipment may include, but not limited to: Manufacturing Operation Management (MOM) system, Manufacturing Execution System (MES), Enterprise Resource Planning (Enterprise Resource Planning, ERP) system, enterprise service Bus (Enterprise Service Bus, ERP), Product Lifecycle Management (Product Lifecycle Management, PLM) system, etc.

In this embodiment, because the IT domain code development tool can be programmed to realize that IT equipment invokes microservices through a knowledge platform to trigger the runtime execution of the OT domain workflow of the main controller of the work unit, thereby realizing the code development of the IT domain The platform controls the process of the OT domain, which realizes the integration of the IT domain and the OT domain. Here, the microservice is automatically generated by the OT domain microservice generator based on the OT domain behavior tree. It is not necessary for the code development tools of the IT domain to understand the details of the OT domain workflow. It only needs to obtain the identification (such as: name) and IP address of the microservice That is, developers in the IT domain do not need to understand OT domain devices and control processes, and are easy to implement and understand.

The applicable fields of the embodiments of the present application include but are not limited to: Industrial Automation, Logistics, Laboratory, Maritime, Smart Grid, Electric Vehicle Infrastructure Vehicle Infrastructure), Electric Vehicle, Building Automation, Smart City, Water Treatment, Garbage Recycling and Smart Farm, etc.

The workflow creation method in the embodiment of the present application has been described in detail above, and the workflow creation system in the embodiment of the present application will be described in detail below. The workflow creation system in the embodiment of the present application can be used to implement the workflow creation method in the embodiment of the present application. For the details not disclosed in the system embodiment of the present invention, please refer to the corresponding description in the method embodiment of the present invention. Here I won't repeat them one by one.

Fig. 3 shows a schematic structural diagram of a workflow creation system in the embodiment of the present application. As shown in FIG. 3 , the system may include: a node library 110 , a graphical interface module 120 and an editing processing module 130 .

Wherein, the node library 110 is provided with behavior tree nodes for constructing behavior trees; the behavior tree nodes may include: flow control nodes and function block nodes. Wherein, a behavior tree is used to represent a workflow, and the workflow is used to define an operation to be performed by a work unit. The flow control node is used to realize the logic control in the workflow; the function block node is used to realize the business operation in the workflow, and the function block node may include: logic nodes, each logic node corresponds to an operation template, each The operation template predefines at least one type of operations that can be performed by resources such as equipment, and the operations include actions, methods or skills.

In one embodiment, the resources are represented in the form of resource nodes, and all resource nodes are associated and stored in the form of a resource knowledge graph; the resource knowledge graph includes: each resource node, and a connection representing the relationship between the resource nodes. Correspondingly, this embodiment may further include: a resource library 150, configured to store various resources in the form of a resource knowledge graph, and each resource can perform at least one business operation.

In one embodiment, the flow control node may include: some or all of the main control node, logical control node and condition node; the main control node may include: a start node, an end node, a go node, a key node , stop node, abort some or all of the nodes. The logic control nodes include: some or all of sequence nodes, reactive sequence nodes, parallel nodes, intra-process quality control nodes, priority nodes, reactive priority nodes, condition judgment nodes, and multi-branch selection nodes.

In one embodiment, the function block nodes may further include: some or all of manual nodes, dynamic nodes, delay nodes, and idle nodes.

In one embodiment, the behavior tree node further includes: a decorator node, which may include: a repeat node, a retry node, a one-off node, a timeout node, a timer node, a negate node, a forced run node, and a forced success Some or all of nodes, force-fail nodes, and monitor nodes.

In one embodiment, some or all of the function block nodes in the node library 110 are respectively bound with resources for executing the service operations corresponding to the function block nodes.

The graphical interface module 120 is configured to provide a graphical user interface GUI for the user to construct a behavior tree based on the behavior tree nodes in the node library.

Wherein, each behavior tree node can be listed on the graphical user interface GUI in the form of an icon.

The editing processing module 130 is configured to generate a behavior tree corresponding to a workflow in response to the construction operation of the behavior tree, and the logical nodes in the behavior tree are instantiated as operations corresponding to resources such as equipment. During specific implementation, the editing processing module 130 can respond to the construction operation of the behavior tree, instantiate each behavior tree node, and establish the connection relationship between each instantiated behavior tree node; based on the instantiated behavior tree The connection relationship between nodes generates a behavior tree corresponding to a workflow. Wherein, some or all of the instantiated function block nodes are associated with resources for executing corresponding business operations. For example, through this operation, a logical node is instantiated as an operation corresponding to a resource such as a device.

In this embodiment, each behavior tree node can be listed on the graphical user interface in the form of an icon, and the user can determine the node needed to create a workflow by selecting and dragging the icon onto the canvas, and further, the node can also be Perform necessary parameter configuration, such as resource configuration and/or input/output parameter configuration. Among them, if there is more than one operation to be performed by a work unit, that is, more than one operation is defined in the required workflow, the behavior tree corresponding to the workflow can include multiple logical nodes, and the corresponding The flow control node, through the corresponding discharge connection of the dragged nodes, finally generates the behavior tree corresponding to the workflow.

In one embodiment, the construction operation may include: an operation of adding a function block node and an operation of associating resources with the added function block node.

Corresponding to the method shown in FIG. 1A, the workflow creation system in this embodiment may further include: a parsing and deployment module 140 configured to parse the behavior tree, and deploy the workflow corresponding to the behavior tree to the corresponding The running time of the main controller of the work unit, so that each resource in the work unit connected to the main controller performs operations according to the workflow.

FIG. 4A shows an OT domain low-code development platform 100 provided by an embodiment of the present application, and the platform 100 can be used to implement the workflow creation system shown in FIG. 3 .

As mentioned earlier, there is currently no single low-code development tool and platform suitable for the OT domain. In the platform 100 shown in FIG. 4A , since the operation templates that various types of OT devices can perform operations are pre-defined (which can be regarded as providing the capability of the OT domain), and the corresponding logical nodes are constructed based on the operation templates. It is used to construct other behavior tree nodes of OT domain workflow, so it can conveniently and quickly create a behavior tree corresponding to OT domain workflow, realizing low-code development suitable for OT domain. Development engineers can realize OT domain development without having a deep understanding of various OT devices.

Specifically, as shown in FIG. 4A, the OT domain low-code development platform 100 may include:

An OT domain low-code development tool 10, the OT domain low-code development tool 10 can be configured to implement the graphical interface module 120 and the editing processing module 130 in the workflow creation system shown in FIG. The parse deployment module 140 in the workflow creation system is shown. The node library 110 in the workflow creation system shown in FIG. 3 can be stored in a memory.

Further, as shown in FIG. 4B , the OT domain low-code development platform 100 may also include the runtime 30 of the main controller of the above-mentioned working unit. The OT domain low-code development tool 10 can deploy the OT domain workflow corresponding to the generated behavior tree to the runtime 30 of the main controller of the work unit, so that each OT in the work unit connected to the main controller The device performs operations according to the OT domain workflow. The runtime 30 in this embodiment is an interpreter-driven runtime system for executing workflow and managing necessary related processes. It's based on an interpreter with a good open source community, such as the Python and Google V8 interpreters, so the ecosystem is easy to build and can be easily extended to use other interpreters. For the behavior tree connected with the data block, the runtime may further provide corresponding data obtained during the execution of the business operation to the corresponding data block for display. During specific implementation, the runtime 30 may directly provide the corresponding data obtained during the execution of the service operation to the corresponding data block for display, or provide the corresponding data block for display through a third-party device. In this embodiment, the standard field bus and device protocols can be used in advance to access the field bus and devices during the runtime. In a specific implementation, the working unit may have a main controller, and at this time, the runtime 30 may be located on the main controller of the working unit. The equipment resources in the resources can be connected to the main controller, and the main controller can control the equipment resources connected to it to perform corresponding operations according to the runtime workflow; the human resources in the resources can be directly based on the runtime work The stream prompts to perform the corresponding action.

The composition of the OT domain low-code development platform 100 shown in FIG. 4A and FIG. 4B only involves the OT domain. However, the integration of the IT domain and the OT domain has become increasingly important for the digital transformation of enterprises. What needs to be realized is how the enterprise can control the process of the OT domain in an understandable and non-IT programming way. The OT domain low-code development platform 100 shown in FIG. 4C solves how to control the flow of the OT domain through the code development platform 300 of the IT domain. As shown in FIG. 4C , on the basis of the structure shown in FIG. 4B , the OT domain low-code development platform 100 may further include an OT domain microservice generator 20 , which can generate microservices 40 based on the OT domain behavior tree. In this way, the IT domain code development tool 301 can be programmed to enable the IT device to call the microservice 40 through a knowledge center 200 to trigger the runtime 30 of the main controller of the work unit to execute the OT domain workflow. In this way, the code development platform 300 of the IT domain can control the process of the OT domain, that is, the integration of the IT domain and the OT domain is realized. Here, the microservice 40 is automatically generated by the OT domain microservice generator 20 based on the OT domain behavior tree. It is not necessary for the code development tool 301 in the IT domain to understand the details of the OT domain workflow. It only needs to obtain the identification of the microservice 40 (for example: name ) and IP address, it is not necessary for IT domain developers to understand OT domain devices and control processes, and it is easy to implement and understand.

To realize that the IT domain can call the microservice 40 , it is required that the IT domain can obtain the information of the microservice 40 . Corresponding to the method shown in Figure 1A, its possible implementation methods include but are not limited to the following two:

method one,

In mode 1, the code developers in the OT domain can notify the code developers in the IT domain of the names and IP addresses of the generated microservices 40, so that the code developers in the IT domain can directly assign the microservices 40 to The information of 40 is written into the code, so as to implement the call of the IT equipment to the microservice 40. Method 1 is more suitable for scenarios with a small number of microservices.

Method two,

In mode 2, registration and discovery mechanisms can be used. Among them, each microservice 40 can be registered on the knowledge center 200 , so that the IT domain code development tool 301 can realize the connected microservice 40 through the knowledge center 200 by the IT domain equipment through code development. The device for completing the registration of the microservice 40 may be the OT domain microservice generator 20 or a third-party device 50 as shown in FIG. 4D . The third-party device 50 can be regarded as a part of the low-code development platform 100 in the OT domain, or implemented in the knowledge center 200 . Method 2 is more suitable for scenarios with a large number of microservices. By registering microservices on the knowledge center platform, it is more effective to realize the calling of microservices by IT equipment, and strengthen the integration of OT domain and IT domain.

Optionally, the OT domain microservice generator 20 can generate the API of the microservice 40 based on the OT domain behavior tree, wherein, the processing procedure in the API can include the operation of each functional block in the OT domain workflow, and the input parameter of the API is OT The parameters obtained by the input port of the domain workflow, and the output parameters of the API are the parameters output by the output port of the OT domain workflow.

An application scenario of the OT domain low-code development platform 100 provided in the embodiment of the present application in the field of industrial automation is shown in FIG. 4E . The low-code development tool 10 generates a behavior tree corresponding to the OT domain workflow under the operation of the user, and the OT domain workflow defines the operations to be performed by the production line as a work unit shown on the right side of FIG. 4E . Based on the behavior tree, the corresponding workflow is generated and published to the runtime 30, so that the runtime 30 controls the completion of the production line operation of the work unit; at the same time, the corresponding microservice can be generated by the microservice generator 20 based on the behavior tree and registered in The knowledge center 200, so that the code development tool 301 in the IT domain can call the corresponding microservice through the knowledge center 200. As shown in the GUI in the lower left corner of Figure 4E, the user can edit each node including the function block node by dragging and dropping to edit the behavior tree of the OT domain. Data (for example: workpiece processing parameters), control the operation of the entire work unit. The working unit here is a production line, which includes machines, conveyor belts, robotic arms, people, PLC, AGB, etc. During specific implementation, the code development tool 301 of the IT domain can also be located on the same hardware device as the low-code development tool 10 , for example, on the same computer.

FIG. 5 shows a schematic structural diagram of another workflow creation system provided by the embodiment of the present application. As shown in Fig. 5, the system can be used to implement the method shown in Fig. 1A, or realize the workflow creation system shown in Fig. 3, or realize the workflow creation system described in any one of Fig. 4A to Fig. 4D , that is, the low-code development platform 100 in the OT domain. Among them, the aforementioned OT domain low-code development tool 10, OT domain microservice generator 20, runtime 30, and third-party device 60 can all be implemented as individual hardware devices, such as servers, workstations, single-chip microcomputers or processing chips. Alternatively, these devices are implemented on the same hardware device, which is stored in at least one memory as a software program, and is called by at least one processor to implement the aforementioned OT domain low-code development method. Wherein, the node library 110 and each generated microservice 40 may be stored in at least one memory.

As shown in FIG. 5 , the system may include: at least one memory 51 , at least one processor 52 and at least one display 53 . In addition, some other components may also be included, such as a communication port (not shown in FIG. 5 ) and the like. These components communicate via bus 54 .

Among them, at least one memory 51 is used to store computer programs. At least one memory 51 may include computer readable media such as random access memory (RAM). In addition, at least one memory 51 can also store an operating system and the like. The operating system includes but is not limited to: Android operating system, Symbian operating system, Windows operating system, Linux operating system and so on. The above computer stored program may include the following program modules: node library 110, graphical interface module 120, editing and processing module 130, parsing and deployment module 140, optionally, may also include OT domain microservice generator 20, runtime 30, third-party device 50.

At least one processor 52 is configured to invoke a computer program stored in at least one memory 51 to execute the workflow creation method described in the embodiment of the present application. At least one processor 52 may be a microprocessor, an application specific integrated circuit (ASIC), a digital signal processor (DSP), a central processing unit (CPU), a graphics processing unit (GPU), a state machine, or the like. It can receive and send data through the communication port.

At least one display 53 is used to display a graphical user interface.

Specifically, at least one processor 52 is configured to invoke a computer program stored in at least one memory 51 to enable the system to execute the operations in the workflow creation method in any of the foregoing implementation manners. In addition, the communication interface is used to implement communication with other devices, such as communication with the knowledge center 200 .

In the embodiment of the present application, the OT domain low-code development tool 10 used to realize the graphical interface module 120, the editing processing module 130 and the parsing and deployment module 140 can be a lightweight web-based application program, which can be used in industrial sites (such as edge device or local server), or in the cloud (public cloud such as AWS or private cloud such as OpenStack). Its visual engineering paradigm is derived from Function Block Typed Diagram (FBTD). The OT domain microservice generator 20 can use modern translation programming languages to generate standard APIs such as RESTful or RPC. Runtime 30 can easily implement OT domain workflow and provide openness based on the ecosystem of open source communities (such as Python). The runtime 30 can be deployed on an embedded IoT device such as a single board computer (Single Board Computer, SBC).

It should be mentioned that the embodiment of the present application may include an apparatus having an architecture different from that shown in FIG. 5 . The above architecture is only exemplary, and is used to explain the workflow construction method provided by the embodiment of the present application.

In addition, the embodiment of the present application also provides an IT domain OT domain integration system, that is, an ITOT system, which may include IT equipment and the workflow creation system in any implementation manner of the present application. In addition, it may further include: a code development platform 300 in the IT domain as shown in FIG. 4C and FIG. 4D .

The embodiment of the present invention also proposes a technical solution of using a synthesizer node with a low-code paradigm (FBTD, Function Block Typed Diagram) to control the workflow. FIG. 6 is an exemplary flowchart of a workflow control method provided by various embodiments of the present application.

As shown in FIG. 6, the method 600 includes:

Step 601: Determine the type of synthesizer node based on the user's operation on the graphical user interface. The synthesizer node includes a start block adapted to start executing the synthesizer node, an end block adapted to end the execution of the synthesizer node, and arranged in A plurality of working links between the start block and the end block.

Step 602: Based on the type of the combiner node, determine the target working link from the plurality of working links.

Step 603: Based on the purpose work link, control the logic of the workflow, wherein the workflow is generated based on the behavior tree including the combiner node.

Synthesizer nodes are always paired with a start block and an end block. The start block has an optional input port to provide a synthesizer expression so that the target working link can be selected based on the synthesizer expression. For example, the target working link is selected according to the value of the combiner expression, or the target working link is selected based on the working link selection rule defined by the combiner expression. The end block has an optional input port called the end logical value, which is used by parallel type synthesizer nodes to determine whether it is a logical AND (AND) or a logical OR (OR). A set of synthesizer pair blocks can contain two or more working links (ie branches). Optionally, a Data Value Block (DVB, Data Value Block) is attached before each working link in the synthesizer pair block, representing the trigger value of the synthesizer expression of the corresponding working link. In one embodiment, the data value block may be a value generated by the corresponding working link based on a synthesizer expression and used to select the corresponding working link. In another embodiment, the block of data values may also be a condition for selecting a corresponding working link, wherein the condition is related to a value generated based on a synthesizer expression.

In one embodiment, at least one working link in the plurality of working links includes a function block node, and the function block node is used to realize the service operation in the workflow.

In one embodiment, the determining the type of the synthesizer node based on the user's operation on the graphical user interface includes: determining the type of the synthesizer node based on the user's selection operation on the graphical user interface containing the node library, The node library includes: a synthesizer node that identifies types in a semantic manner or a synthesizer node that identifies types in a presentation style.

In the synthesizer node that identifies the type in a semantic manner, the type of the synthesizer node is directly identified in a textual manner in the icon of the synthesizer node that has a visual effect.

In the synthesizer node whose type is identified by the presentation style, different presentation styles are used to identify the type of the synthesizer node accordingly, for example, in the icon of the synthesizer node, different line types (for example, Solid line, dotted line, spaced line, etc.) to identify the type of synthesizer node accordingly.

In one embodiment, it also includes: receiving the user's construction operation of the behavior tree on the graphical user interface; wherein, the construction operation includes the addition and connection operation of the behavior tree node, and the behavior tree node Including the synthesizer node and the function block node; the behavior tree is used to characterize the workflow, and the workflow is used to define the operation to be performed by a work unit; the function block node includes: a logic node, each Each logical node corresponds to an operation template, and each operation template predefines at least one type of operation that can be performed by a device, and the operation includes: action, method or skill; in response to the construction operation of the behavior tree, generate a corresponding workflow The behavior tree, wherein the logical nodes in the behavior tree are instantiated as the operation of the corresponding device; the behavior tree is parsed to obtain the workflow; the workflow is deployed to the corresponding work unit At runtime, each device in the work unit performs operations according to the workflow; wherein the workflow is an OT domain workflow; and the device is an OT device.

In one embodiment, the method further includes: generating a microservice based on the behavior tree, so that the IT device triggers the runtime of the main controller of the work unit to execute the OT domain workflow by invoking the microservice .

The following demonstrates the implementation of flexible control of the workflow by using different types of synthesizer nodes.

FIG. 7A is a first schematic diagram of a low-code example of an FBTD logic synthesizer provided by various embodiments of the present application. In FIG. 7A , the synthesizer node 70 includes a start block 71 , an

end block

72 and 3 working links arranged between the start block 71 and the end block 72 . The first working link includes the executive function block node 73 , the second working link includes the executive function block node 74 , and the third working link includes the executive function block node 75 . The start block 71 may contain an input for providing a composite expression 76 ; the end block 72 may contain an input for providing an end logical value 77 . Each function block node 73 , function block node 74 and function block node 75 has its own data value block. The data value block 78 of the function block node 73, the data value block 79 of the function block node 74 and the data value block 80 of the function block node 75 may be arranged before the respective corresponding function block nodes.

Depending on the different types of combiner nodes 70, different ways of determining the destination working link from the plurality of working links can be implemented, thus realizing flexible control for the workflow.

In one embodiment, the type of combiner node 70 is parallel selection (Pa, Parallel); Determining the destination working link from the plurality of working links includes: combining each working link in the plurality of working links All roads are determined as the target working link, and start to execute the target working link; the method also includes: receiving the end logic value through the input terminal of the end block; wherein when the end logic value is logic and, the plurality of working links When all executions are completed, the execution of the synthesizer node is terminated through the end block; when the end logic value is logical or, when at least one of the multiple working links appears to be executed, the execution of the synthesis is terminated through the end block server node.

Specifically, in FIG. 7A , when the type of synthesizer node 70 is parallel selection, there is no need to provide the synthesizer node 70 with a synthesized expression 76, correspondingly there is no need to generate synthesized data value blocks 78-80, and it is necessary for the synthesizer node 70 provides an end logical value 77 . At this time, the first working link, the second working link and the third working link are all determined as the target working links. Moreover, the first working link, the second working link and the third working link are executed respectively (i.e. synchronously executing function block node 73, function block node 74 and function block node 75), wherein based on the end logic value The value of 77 determines when to stop executing the compositor node 70. Wherein: when the value of the end logical value 77 provided is logic AND, when the first working link, the second working link and the third working link are all executed, the execution of the synthesis is ended via the ending block 72 Device node 70; when the value of the end logic value 77 is logical or, at least one working link in the first working link, the second working link and the third working link occurs when the execution is completed, via the end Block 72 ends execution of combiner node 70 .

In one embodiment, the type of the synthesizer node is multi-branch selection (SW, Switch), and the method further includes: receiving a preset synthesis expression through the input terminal of the start block; Determining the target working link in the working link includes: determining the data value block when each working link is selected; based on the calculation result of the composite expression, selecting the corresponding data value from the plurality of working links The destination working link of the value block.

Specifically, in FIG. 7A , when the type of the combiner node 70 is multi-branch selection, a combined expression 76 needs to be provided, and correspondingly, the combined data value blocks 78-80 need to be generated or manually provided based on the combined expression 76, but there is no need to An end logical value of 77 is provided. For example, assume that the data value block is a condition for selecting a corresponding working link, wherein the data value block 78 is: "the temperature is greater than 30 degrees and less than 50 degrees"; the data value block 79 is: "the temperature is greater than 50 degrees"; the data The value block 80 is: "the temperature is 30 degrees lower". Then, based on the specific temperature generated by the composite expression, the specific branch corresponding to the numerical block can be selected. For example, when the generated temperature is 20 degrees, the third working link corresponding to the data value block 80 is selected.

In one embodiment, the type of the synthesizer node is a conditional judgment (ITE, If-Then-Else), the method further includes: receiving a preset synthesis expression via the input terminal of the start block, the synthesis The result of the expression contains a logical true value or a logical false value; the determining the target working link from the plurality of working links includes: based on the result of the composite expression being a logical true value or a logical false value, from A corresponding destination working link is determined among the plurality of working links.

Here, based on the logical value generated by the composite expression being true or false, a preset working link corresponding to the logical value is selected.

Specifically, in FIG. 7A, when the type of combiner node 70 is a conditional judgment, it is necessary to provide a combination expression 76 that can generate a logical value of True (Ture) or False (False), assuming that the first working link corresponds to true and the second working link corresponds to false. Then, when the value of composite expression 76 is true, then select and execute the first working link as the destination working link; when the value of composite expression 76 is false, then select and execute as the destination working link The second working link of . In one embodiment, the type of the synthesizer node is priority (Pr, Priority (Fallback)), the method further includes: receiving a preset synthesis expression via the input terminal of the start block; Determining the target working link among the plurality of working links includes: calculating the data value block of each working link based on the composite expression; based on the sorting result of the data value block of each working link, determining the A priority order of the plurality of working links; based on the priority order, the target working link is determined from the plurality of working links. The data value block may be a value generated by the corresponding working link based on the synthesizer expression and used to select the corresponding working link. That is, the data value block 78, data value block 79, and data value block 80 generated based on the synthesizer expression may be different (for example, the parameter difference of each execution link), at this time based on the data of each working link According to the sorting result of the value block, the priority order of the plurality of working links is determined; based on the priority order, the target working link is determined from the plurality of working links, wherein based on the preset, the data The sorting result of the value block can be sorted from largest to smallest or from smallest to largest, and so on. For example, in FIG. 7A , when the type of the combiner node 70 is priority, a combination expression 76 needs to be provided. Assume that the data value block generated by the first link based on the composite expression 76 is 15; assume that the data value block 78 generated by the first link based on the composite expression 76 is 18, and the second link is based on the composite expression The data value block 79 generated by the expression 76 is 20; the data value block 80 generated by the third link based on the composite expression 76 is 25. Determine the priority sequence of the working link according to the size order of the data value block, it can be seen that the third link has the highest priority (because the data value block 80 of the third link is the largest in the data value block), thereby determining the first Three links are the destination working links, and the third link as the destination working link is executed.

In one embodiment, the type of the synthesizer node is Reactive Priority (RPr, Reactive Priority (Fallback)), the method further includes: receiving the first synthesized expression and the second synthesized expression via the input of the start block Composite expression; determining the target working link from the plurality of working links includes: calculating the first data value block of each working link based on the first composite expression; based on the second composite The expression calculates the second data value block of each working link; based on the sorting result of the first data value block of the plurality of working links, the priority order of the plurality of working links is determined; based on the priority determining the target working link from the plurality of working links according to the level order and the second data value block of each working link.

In Fig. 7A, in the icon with visual effects of the synthesizer node, the type of the synthesizer node can be directly identified by text (for example, in the box of the start block 71 and the end block 72, the type of the synthesizer node can be identified in text form type). FIG. 7B is a second schematic diagram of a low-code example of the FBTD logic synthesizer provided by various embodiments of the present application. In FIG. 7B , the start block 71 and the end block 72 are briefly identified by spaced lines with specific shapes, so that the types of synthesizer nodes can be visually identified and the display resources of the canvas can be saved.

The synthesizer node described above is easy to understand, but only works with a limited number of branches. Therefore, the embodiment of the present invention also introduces a new synthesizer node paradigm, which can support a large number of branches. The folded synthesizer has similar functionality to the normal synthesizer, but supports folding.

In one embodiment, the synthesizer node is displayed in a folded manner in a collapsible box on the graphical user interface, wherein the collapsible frame includes a first display area adapted to display the The currently displayed working links of the above-mentioned synthesizer nodes and hide the working links except the currently displayed working links. In one embodiment, the folding box further includes a second display area, a third display area, and a switching control; wherein the second display area is adapted to display the data value blocks of the currently displayed working link; the The third display area is adapted to display the label of the currently displayed working link; the switching control is adapted to switch the currently displayed working link among the plurality of working links.

FIG. 7C is a schematic diagram of a low-code example of a synthesizer with folded FBTD logic provided by various embodiments of the present application.

In FIG. 7C, the first display area 88, the second display area 83, and the third display area 81 are included in the folding frame 90, wherein the first display area 88 displays a current display working link, and the current display working link includes Function block node 85 and function block node 86 . The data value blocks of the currently displayed working link are displayed in the second display area 83 . The label of the currently displayed working link is displayed in the third display area 81 . The folding box 90 also includes a switching control 84 (for example, in the shape of an arrow) for switching the currently displayed working link among the plurality of working links. The first display area 88 is preferably stretchable based on the length of the currently displayed active link contained therein. As you can see, folding synthesizers are another type of low-code paradigm for FBTD synthesizers. Data value blocks for synthesizer expression results can be in a new column in the header row. Also, there are left and right arrow buttons to switch between different branches. The data value block for the current branch is displayed in the Data Value Block column. At the same time, the current label of the current branch of the compositor will be displayed in the work order label index block. The size of the folded synthesizer can be stretched for different numbers of functional blocks, and this paradigm provides a convenient way to display complex logic in a limited area of the low-code canvas.

The embodiment of the present invention also proposes a workflow execution method. After parsing the workflow in markup language format, the workflow expressed in the form of Node Link Assembly (NLA) is obtained, and then the workflow is compiled. And download to the runtime of the main controller of the corresponding work unit, and then the workflow can be executed. Node-linked assembly is implemented as a general-purpose, runtime execution engine logic assembly language, preferably for executing event-based low-code workflows.

Basic elements of a node link assembly can include:

(1), node link structure

A node-link structure is the basic type of cyclic directed graph data structure. Most event-based low-codes are based on node-link structures, such as the IEC61499 standard. Embodiments of the present invention preferably use an event-based behavior tree, so interpreting the behavior tree into a node-link structure facilitates execution. The node-link structure can be seen as a low-code assembly language.

(2), conditional jump node

A conditional jump is a single node added by the NodeLinkAssembly's interpreter to convert an event-based behavior tree into a workflow expressed in NodeLinkAssembly. The function of the conditional jump node is: when the condition (usually an expression) is met, jump to any place specified by the link. In assembly language, it is a node needed for flow control and introduced as an execution element to support low-code interpretation.

(3), reverse logical flow and backward vector

Reverse logic flow is a basic generic logic that can support behavior trees or all reverse logic in other programming languages. Loop (Repeat) or Retry in the behavior tree low code is a reverse logic flow, which can be converted into a conditional jump node with reverse event jump. The skipped function blocks are part of the reverse vector. Reverse logic flow and backward vectors provide a metamethod to describe logic as a node-linked assembly.

(4), forward logic flow and forward vector

Forward logic flow is a basic general logic that can support behavior trees or all forward logic in other programming languages. The condition/timeout in the behavior tree low code is a forward logic flow, which can be converted into a conditional jump node with a forward event jump. The skipped functional blocks are part of the forward vector. Forward Logic Flow and Forward Vector provide a metamethod to describe logic as a node-link assembly.

(5), nesting and jump (Goto) logic

Backward vectors and forward vectors are logically complete and can represent all logic, such as compound logic (such as If-Then-Else and Switch-Case-Default) and other behavior tree logic (such as Fallback or Guard). Logic assembled by node linking can be part of a low-code workflow that also includes sequential, parallel, and flow control nodes (such as start, end, and goto).

(6), empty node

An empty node is an empty function block with no actual function. It is automatically generated by the interpreter of node link assembly for internalization and jump logic for jump targets.

FIG. 8 is an exemplary flowchart of a workflow execution method provided by an embodiment of the present application. As shown in Figure 8, the method 800 includes:

Step 801: Obtain the workflow in markup language format, wherein the workflow in markup language format is generated based on the behavior tree construction operation performed by the user on the graphical user interface.

Here, the workflow in the markup language format can be obtained by parsing the behavior tree constructed by presenting function block nodes in the form of a label graph, as shown in Figure 2B to Figure 2S , or it can be obtained by FIG. 2T-FIG. 2Y or FIG. 7A-FIG. 7C are examples obtained by analyzing the behavior tree constructed by presenting function block nodes in the form of a type diagram.

Step 802: Convert the workflow in markup language format into a workflow topology implemented in node-link assembly.

Here, the workflow topology implemented by node-link assembly can be characterized as the sequence of machine instructions converted from the workflow in language format by the assembly process of node-link assembly.

Step 803: Execute the workflow topology.

For example, the sequence of machine instructions is executed in the device's runtime to execute the workflow.

In one embodiment, in step 802, converting the workflow in markup language format into a workflow topology structure realized by node link assembly includes: determining the sequential execution structure and the loop execution structure included in the workflow in markup language format; The sequential execution structure determines the sequential execution part of the workflow topology, and the sequential execution part includes the function block nodes in the sequential execution structure and the links between the function block nodes; in the sequential execution part, corresponding to the loop execution structure in the markup language format At the first position of the workflow position, add the first conditional jump node. The first conditional jump node is determined based on the loop execution structure. The first conditional jump node contains the first preset condition and is linked to the first position Before the first function block node, the first condition jump node is also linked to the second function block node based on the execution sequence of the sequential execution part, wherein at the first condition jump node: when the first preset condition is satisfied, Jump to the first function block node; when the first preset condition is not satisfied, execute the second function block node based on the execution order. It can be seen that in the embodiment of the present invention, the reverse logic flow can be realized through the conditional jump node, which facilitates the realization of various loop logics that jump forward (for example, while, Repeat, or Retry, etc. can be realized).

In one embodiment, in step 802, converting the workflow in markup language format into a workflow topology implemented in node link assembly includes: determining the sequential execution structure and conditional execution structure included in the workflow in markup language format; The sequential execution structure determines the sequential execution part of the workflow topology, and the sequential execution part includes the function block nodes in the sequential execution structure and the links between the function block nodes; in the sequential execution part, corresponding to the conditional execution structure in the markup language format At the second position of the workflow position, add a second conditional jump node, the second conditional jump node is determined based on the conditional execution structure, the second conditional jump node contains the second preset condition and is linked to the second position After the third function block node, the second condition jump node is also linked to the fourth function block node based on the execution sequence of the sequential execution part; wherein at the second condition jump node: when the second preset condition is satisfied, Executing the fourth function block node based on the execution order; jumping to the third function block node when the preset condition is not satisfied. Therefore, in the embodiment of the present invention, the forward logic flow can be realized through the conditional jump node, which facilitates the realization of various conditional logics for backward jumping (for example, Condition or timeout, etc.).

In one embodiment, converting the workflow in markup language format into a workflow topology implemented in node link assembly in step 802 includes: converting the workflow in markup language format into a workflow topology implemented in node link assembly includes : Determine the sequential execution structure, loop execution structure and conditional execution structure contained in the workflow in markup language format; determine the sequential execution part of the workflow topology based on the sequential execution structure, and the sequential execution part includes the function block nodes in the sequential execution structure and links between function block nodes; in the sequential execution part, at the first position corresponding to the position of the loop execution structure in the workflow in markup language format, add a first conditional jump node, the first conditional jump node is Determined based on the loop execution structure, the first conditional jump node contains the first preset condition and is linked to the first function block node before the first position, and the first conditional jump node is also linked to the first function block node based on the execution sequence of the sequential execution part Two function block nodes, where at the first condition jump node: when the first preset condition is met, jump to the first function block node; when the first preset condition is not met, execute the first function block node based on the execution sequence Two function block nodes; in the sequential execution part, at the second position corresponding to the position of the conditional execution structure in the workflow in markup language format, add a second conditional jump node, the second conditional jump node is based on the conditional execution structure Determined, the second conditional jump node contains the second preset condition and is linked to the third function block node after the second position, and the second conditional jump node is also linked to the fourth function block node based on the execution sequence of the sequential execution part ; Wherein at the second condition jump node: when the second preset condition is met, execute the fourth function block node based on the execution order; when the preset condition is not met, jump to the third function block node. Therefore, in the embodiment of the present invention, the forward logic flow and the reverse logic flow are realized respectively through a plurality of conditional jump nodes, which facilitates the realization of various forward jumping loop logics and various backward jumping conditional logics.

In one embodiment, in step 802, converting the workflow in markup language format into a workflow topology implemented in node link assembly includes: determining the parallel execution structure included in the workflow in markup language format; , corresponding to the third position of the workflow in the markup language format of the parallel execution structure, add at least one parallel execution part, at least one parallel execution part converges to the parallel sink node of the sequential execution part, and the parallel sink node is also based on the sequence The execution order of the execution part is linked to the fifth function block node; the parallel sink node is a logic AND node or a logic OR node, wherein when all parallel execution parts converged to the logic AND node are executed, the fifth function block node is executed; when The fifth function block node is executed when at least one parallel execution part of all the parallel execution parts converged to the logical OR node has been executed. It can be seen that the embodiment of the present invention also implements node link assembly processing for parallel logic.

In one embodiment, the method 800 further includes: when the first conditional jump node or the second conditional jump node jumps to a logical AND node, adding an empty node before the logical AND node, and jumping the conditional jump node Go to an empty node, where an empty node is used to indicate no operation, linked to a logical AND node based on the execution order of the sequential execution part. Therefore, by automatically setting empty nodes, it is ensured that there is at least one function block node between the conditional jump node and the logical AND node, thereby avoiding logic errors.

Currently, the main problems with node link assembly processing include: It is difficult to revert to event-based behavior trees. To solve this problem, when the node link assembly interpreter is invoked, the original behavior tree-based workflow information is further stored and used as an input for reverse interpretation and uploading. In one embodiment, the method 800 further includes: associating and storing the workflow in markup language format and the workflow topology; receiving a reverse interpretation request of a workflow adapted to the requested workflow topology; responding to the reverse interpretation request , which provides workflows in markup language format based on associative storage. It can be seen that the embodiment of the present invention can quickly provide a workflow in markup language format, realize a reverse interpretation, and solve the technical problem that the workflow topology after node link assembly processing is difficult to restore to the workflow in markup language format.

In FIG. 9A , in the sequential execution structure indicated in the workflow in the markup language format, the sequentially executed function block FB1 , function block FB2 , function block FB3 and function block FB4 are included. Furthermore, a loop execution structure or a conditional execution structure may be provided between the function block FB2 and the function block FB3.

After node link compilation, the converted workflow topology includes function block FB1, function block FB2, function block FB3, function block FB4 and the links between each function block determined based on the execution sequence. Moreover, between the function block FB2 and the function block FB3 , the loop execution structure or the conditional execution structure is also characterized by a conditional jump node (JMPC), which has a predetermined condition 903 .

Alternatively, JMPC may be linked to function block FB2 via reverse logic flow 902 , or to function block FB4 via forward logic flow 901 . in:

(1), when JMPC is linked to function block FB2 via reverse logic flow 902: after running to JMPC, if judgment predetermined condition 903 does not get satisfied, then execute to function block FB3 from JMPC sequence, if judgment predetermined condition 903 When it is satisfied, return to function block FB2 from JMPC.

(2), when JMPC is linked to function block FB4 via forward logical flow 901: after running to JMPC, if judge predetermined condition 903 to be satisfied, then execute to function block FB3 from JMPC order, if judge predetermined condition 903 must not When it is satisfied, jump from JMPC to function block FB4.

In the embodiment of the present invention, the JMPC may perform a jump when the predetermined condition 903 is satisfied, and may also perform a jump when the predetermined condition 903 is not satisfied.

The converted workflow topology structure includes the start node S1 and the end node S2, and also includes the nodes between the start node S1 and the end node S2, which are executed sequentially based on the direction of the arrow (including multiple function blocks and the first conditional jump node J1 , the second conditional jump node J2). The first conditional jump node J1 is not only linked to the function block FB5 based on the order, but also coupled to the function block FB2 based on the reverse logic. The second conditional jump node J2 is not only linked to the function block FB7 based on sequence, but also coupled to the function block FB9 based on forward logic.

When the first conditional jump node J1 is executed based on the sequence, if the condition 92 of the first conditional jump node J1 is satisfied, jump to the function block FB2 linked to the first conditional jump node J1, if the first conditional jump node J1 is linked to the function block FB2, if the first conditional jump node J1 is When the condition 92 of the transfer node J1 is not satisfied, the function block FB5 is executed based on the execution sequence shown by the arrow.

After the second conditional jump node J2 is executed based on the sequence, if the condition 93 of the second conditional jump node J2 is satisfied, the function block FB7 is executed based on the execution sequence shown by the arrow, and if the condition of the second conditional jump node J1 93 If not satisfied, jump to function block FB9.

The converted workflow topology includes the start node S1 and the jump node S3, and also includes the nodes between the start node S1 and the jump node S3, which are executed in sequence based on the direction of the arrow (including multiple function blocks and the first conditional jump node J1, the second conditional jump node J2). The jump node S3 is used to unconditionally jump to the function block FB1. The first conditional jump node J1 is not only linked to the function block FB5 based on the order, but also coupled to the function block FB2 based on the reverse logic. The second conditional jump node J2 is not only linked to the function block FB7 based on sequence, but also coupled to the function block FB9 based on forward logic.

After the first conditional jump node J1 is executed based on the order, if the condition 92 of the first conditional jump node J1 is met, jump to the function block FB2, if the first conditional jump node J1’s condition 92 is not satisfied , the function block FB5 is executed based on the execution sequence shown by the arrow.

After the second conditional jump node J2 is executed based on the sequence, if the condition 93 of the second conditional jump node J2 is satisfied, the function block FB7 is executed based on the execution sequence shown by the arrow, and if the condition of the second conditional jump node J1 When 93 is not satisfied, jump to function block FB9.

When the sequence-based execution reaches the jump node S3, it jumps to the function block FB1.

It can be seen that the loop execution workflow logic is realized based on the jump node S3.

The converted workflow topology includes start node S1 and jump node S3. The workflow topology also includes nodes between the start node S1 and the jump node S3, which are executed sequentially based on the direction of the arrow (including multiple function blocks and conditional jump nodes J3, conditional jump nodes J4, parallel sink nodes A1, parallel sink node A2 and empty node L). The jump node S3 is used to unconditionally jump to the function block 98 .

After executing function block 98, in addition to including the trunk road of conditional jump node J3 (execution direction shown by the top row of arrows in the figure), there is also a first parallel execution structure from function block 94 to function block 95 . After the first parallel execution structure executes the function block 94, it further has a second parallel execution structure starting from the function block 96 and ending with the function block 97. The first parallel execution structure, the second parallel execution structure and the main road finally converge at the parallel sink node A1.

When the parallel sink node A1 is implemented as a logical AND node: the first parallel execution structure, the second parallel execution structure and the main road (the main road is also regarded as a parallel execution structure) are all executed, then the parallel sink node A1 is executed. next sequential node. When the parallel sink node A1 is implemented as a logical OR node: when one of the first parallel execution structure, the second parallel execution structure, and the trunk road (the trunk road is also regarded as a parallel execution structure) is executed, the parallel sink node is executed The next sequential node of A1.

When the conditional jump node J4 jumps to the parallel sink node A2 implemented as a logical AND node, the node link assembly interpreter automatically adds an empty node L1 at the nearest adjacent position before the parallel sink node A2, and transfers the conditional jump node J4 Jump to the empty node L1, where the empty node L1 is used to indicate not to execute the operation, and the empty node L1 is linked to the parallel sink node A2 based on the execution sequence of the sequential execution part.

It can be seen that by automatically setting empty nodes, it is ensured that there is at least one function block node between the conditional jump node and the logic AND node, thereby avoiding logic errors.

FIG. 10 is an exemplary structural diagram of a workflow execution device provided by various embodiments of the present application. The workflow execution device 700 includes: an acquisition module 701, configured to acquire a workflow in a markup language format, wherein the workflow in a markup language format is generated based on a behavior tree construction operation performed by a user on a graphical user interface; a conversion module 702, configured to convert the workflow in the markup language format into a workflow topology implemented in node link assembly; an execution module 703, configured to execute the workflow topology.

In one embodiment, the conversion module 702 is configured to determine the sequence execution structure and the loop execution structure contained in the workflow in the markup language format; determine the sequence execution part of the workflow topology based on the sequence execution structure, and the sequence execution part includes the sequence The function block nodes in the execution structure and the links between the function block nodes; in the sequential execution part, corresponding to the first position of the loop execution structure in the workflow in markup language format, add a first conditional jump node, The first conditional jump node is determined based on the loop execution structure. The first conditional jump node contains the first preset condition and is linked to the first function block node before the first position. The first conditional jump node is also based on sequential execution Part of the execution sequence is linked to the second function block node, wherein at the first condition jump node: when the first preset condition is met, jump to the first function block node; when the first preset condition is not met , the second function block node is executed based on the execution order.

In one embodiment, the conversion module 702 is used to determine the sequential execution structure and the conditional execution structure contained in the workflow in the markup language format; determine the sequential execution part of the workflow topology based on the sequential execution structure, and the sequential execution part includes the sequence Execute the function block nodes in the structure and the links between the function block nodes; in the sequential execution part, at the second position corresponding to the position of the conditional execution structure in the workflow in markup language format, add a second conditional jump node, The second conditional jump node is determined based on the conditional execution structure, the second conditional jump node contains the second preset condition and is linked to the third function block node after the second position, and the second conditional jump node is also based on sequential execution The execution order of the part is linked to the fourth function block node; wherein at the second condition jump node: when the second preset condition is met, execute the fourth function block node based on the execution order; when the preset condition is not met , jump to the third function block node.

In one embodiment, the conversion module 702 is used to determine the sequential execution structure, loop execution structure and conditional execution structure contained in the workflow in the markup language format; determine the sequential execution part of the workflow topology based on the sequential execution structure, the sequence The execution part includes the function block nodes in the sequential execution structure and the links between the function block nodes; in the sequential execution part, at the first position corresponding to the position of the loop execution structure in the workflow in markup language format, add the first condition Jump node, the first condition jump node is determined based on the loop execution structure, the first condition jump node contains the first preset condition and is linked to the first function block node before the first position, the first condition jump node Also link to the second function block node based on the execution sequence of the sequential execution part, wherein at the first condition jump node: when the first preset condition is satisfied, jump to the first function block node; when the first preset When the condition is not met, the second function block node is executed based on the execution order; in the sequential execution part, at the second position corresponding to the position of the conditional execution structure in the workflow in markup language format, a second conditional jump node is added, The second conditional jump node is determined based on the conditional execution structure, the second conditional jump node contains the second preset condition and is linked to the third function block node after the second position, and the second conditional jump node is also based on sequential execution The execution order of the part is linked to the fourth function block node; wherein at the second condition jump node: when the second preset condition is met, execute the fourth function block node based on the execution order; when the preset condition is not met , jump to the third function block node.

In one embodiment, the conversion module 702 is configured to determine the parallel execution structure included in the workflow in the markup language format; in the sequential execution part, the third corresponding to the position of the parallel execution structure in the workflow in the markup language format position, add at least one parallel execution part, at least one parallel execution part converges to the parallel sink node of the sequential execution part, and the parallel sink node is also linked to the fifth function block node based on the execution sequence of the sequential execution part; the parallel sink node is logic AND node or logical OR node, wherein when all the parallel execution parts converged to the logical AND node are executed, the fifth function block node is executed; when at least one parallel execution part is executed in all the parallel execution parts converged to the logical OR node When finished, the fifth function block node is executed.

In one embodiment, the conversion module 702 is configured to add an empty node before the logical AND node when the first conditional jump node or the second conditional jump node jumps to the logical AND node, and jump the conditional jump node Go to an empty node, where an empty node is used to indicate no operation, linked to a logical AND node based on the execution order of the sequential execution part.

In one embodiment, it also includes a reverse interpretation module 704, which is used for associating and storing the workflow in markup language format and the workflow topology; receiving the reverse interpretation request of the workflow adapted to the requested workflow topology; responding Inversely interpret requests and provide workflows in markup language format based on associative storage.

In addition, the embodiment of the present application also provides a computer-readable storage medium, on which computer-readable codes are stored, and when the computer-readable codes are executed by the processor, the processor executes the aforementioned workflow execution method. In addition, an embodiment of the present application further provides a computer program product, the computer program product is tangibly stored on a computer-readable medium and includes computer-readable instructions, and when executed, the computer-readable instructions cause at least one processing The controller executes the steps in the workflow execution method in the embodiment of the present application.

Fig. 11 is an exemplary structural diagram of a workflow execution device with a memory-processor architecture provided by various embodiments of the present application. The workflow execution apparatus 500 includes: at least one memory 501 and at least one processor 502 . At least one processor 502 is configured to invoke a computer program stored in at least one memory 501 to execute the workflow execution method described in the embodiment of the present application.

Specifically, a system or device equipped with a storage medium may be provided, on which computer-readable codes for realizing the functions of any implementation manner in the above-mentioned embodiments are stored, and the computer (or CPU or The MPU) reads and executes the computer readable codes stored in the storage medium. In addition, some or all of the actual operations can also be completed by an operating system or the like operating on the computer through instructions based on computer readable codes. It is also possible to write the computer-readable code read from the storage medium into a memory provided in an expansion board inserted into the computer or in a memory provided in an expansion unit connected to the computer, and then based on the computer-readable code The instruction causes the CPU installed on the expansion board or the expansion unit to perform some or all of the actual operations, so as to realize the functions of any one of the above-mentioned embodiments. In this embodiment, examples of computer-readable media include, but are not limited to, floppy disks, CD-ROMs, magnetic disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW , DVD+RW), memory chips, ROM, RAM, ASIC, configured processor, all-optical media, all tape or other magnetic media, or any other media from which a computer processor can read instructions. In addition, various other forms of computer-readable media can transmit or carry instructions to the computer, including routers, private or public networks, or other wired and wireless transmission devices or channels, such as downloading from a server computer or cloud by a communication network computer readable instructions. Instructions may include code in any computer programming language, including C, C++, C++, Visual Basic, java, and JavaScript.

It should be noted that not all the steps and modules in the above processes and system structure diagrams are necessary, and some steps or modules can be ignored according to actual needs. The execution order of each step is not fixed and can be adjusted as required. The system structures described in the above embodiments may be physical structures or logical structures, that is, some modules may be realized by the same physical entity, or some modules may be realized by multiple physical entities, or may be realized by multiple Certain components in individual devices are implemented together.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the scope of the present invention. within the scope of protection.

Claims

The workflow execution method (800), is characterized by comprising:

Obtaining a workflow in a markup language format, wherein the workflow in a markup language format is generated based on a behavior tree construction operation performed by a user on a graphical user interface (801);

Converting the workflow in the markup language format into a workflow topology implemented in Node Link Assembly (802);

The workflow topology is executed (803).
The workflow execution method (800) according to claim 1, characterized in that said converting the workflow in markup language format into a workflow topology (802) realized by node link assembly comprises:

determining a sequential execution structure and a loop execution structure included in the workflow in the markup language format;

determining a sequential execution part of the workflow topology based on the sequential execution structure, the sequential execution part including function block nodes in the sequential execution structure and links between the function block nodes;

In the sequential execution part, at the first position corresponding to the position of the loop execution structure in the workflow in the markup language format, a first conditional jump node is added, and the first conditional jump node is based on As determined by the loop execution structure, the first conditional jump node contains a first preset condition and is linked to the first function block node before the first position, and the first conditional jump node is also based on the The execution sequence of the sequential execution part is linked to the second function block node, wherein at the first condition jump node: when the first preset condition is satisfied, jump to the first function block node; when When the first preset condition is not satisfied, execute the second function block node based on the execution sequence.
The workflow execution method (800) according to claim 1, characterized in that said converting the workflow in markup language format into a workflow topology (802) realized by node link assembly comprises:

determining a sequential execution structure and a conditional execution structure to be included in the workflow in the markup language format;

determining a sequential execution part of the workflow topology based on the sequential execution structure, the sequential execution part including function block nodes in the sequential execution structure and links between the function block nodes;

In the sequential execution part, at a second position corresponding to the position of the conditional execution structure in the workflow in the markup language format, a second conditional jump node is added, the second conditional jump node is based on As determined by the conditional execution structure, the second conditional jump node contains a second preset condition and is linked to a third function block node after the second position, and the second conditional jump node is also based on the The execution order of the sequential execution part is linked to a fourth function block node; wherein at the second condition jump node: when the second preset condition is satisfied, the fourth function block is executed based on the execution order A node; when the preset condition is not satisfied, jump to the third function block node.
The workflow execution method (800) according to claim 1, characterized in that said converting the workflow in markup language format into a workflow topology (802) realized by node link assembly comprises:

determining a sequential execution structure, a loop execution structure, and a conditional execution structure included in the workflow in the markup language format;

determining a sequential execution part of the workflow topology based on the sequential execution structure, the sequential execution part including function block nodes in the sequential execution structure and links between the function block nodes;

In the sequential execution part, at the first position corresponding to the position of the loop execution structure in the workflow in the markup language format, a first conditional jump node is added, and the first conditional jump node is based on As determined by the loop execution structure, the first conditional jump node contains a first preset condition and is linked to the first function block node before the first position, and the first conditional jump node is also based on the The execution sequence of the sequential execution part is linked to the second function block node, wherein at the first condition jump node: when the first preset condition is satisfied, jump to the first function block node; when When the first preset condition is not satisfied, execute the second function block node based on the execution sequence;

In the sequential execution part, at a second position corresponding to the position of the conditional execution structure in the workflow in the markup language format, a second conditional jump node is added, the second conditional jump node is based on As determined by the conditional execution structure, the second conditional jump node contains a second preset condition and is linked to a third function block node after the second position, and the second conditional jump node is also based on the The execution order of the sequential execution part is linked to a fourth function block node; wherein at the second condition jump node: when the second preset condition is satisfied, the fourth function block is executed based on the execution order A node; when the preset condition is not satisfied, jump to the third function block node.
The workflow execution method (800) according to any one of claims 1-4, characterized in that,

The converting the workflow in the markup language format into a workflow topology implemented in node link assembly (802) includes:

determining a parallel execution structure to be included in the workflow in the markup language format;

At a third position in the sequential execution part corresponding to the position of the parallel execution structure in the workflow in the markup language format, at least one parallel execution part is added, the at least one parallel execution part converges to the On the parallel sink node of the sequential execution part, the parallel sink node is also linked to the fifth function block node based on the execution order of the sequential execution part; the parallel sink node is a logical AND node or a logical OR node, wherein when converged to Execute the fifth function block node when all the parallel execution parts of the logical AND node are executed; when at least one parallel execution part of all the parallel execution parts converged to the logical OR node is executed, execute the fifth function block node Describe the fifth function block node.
The workflow execution method (800) according to claim 5, further comprising:

When the first conditional jump node or the second conditional jump node jumps to a logical AND node, an empty node is added before the logical AND node, and the conditional jump node is jumped to the empty node, Wherein the empty node is used to indicate not to perform an operation, and the empty node is linked to the logical AND node based on the execution order of the sequential execution part.
The workflow execution method (800) according to any one of claims 1-4, further comprising:

associatively storing the workflow in markup language format and the workflow topology;

receiving a reverse interpretation request adapted to a workflow requesting said workflow topology;

In response to the reverse interpretation request, the workflow in markup language format is provided based on the associative store.
The workflow execution device (700), is characterized in that it includes:

An acquisition module (701), configured to acquire a workflow in a markup language format, wherein the workflow in a markup language format is generated based on a behavior tree construction operation performed by a user on a graphical user interface;

A conversion module (702), configured to convert the workflow in markup language format into a workflow topology implemented in node link assembly;

An execution module (703), configured to execute the workflow topology.
The workflow execution device (700) according to claim 8, characterized in that,

The conversion module (702), configured to determine the sequence execution structure and the loop execution structure contained in the workflow in the markup language format; determine the sequence execution part of the workflow topology based on the sequence execution structure, the The sequential execution part includes the function block nodes in the sequential execution structure and the links between the function block nodes; in the sequential execution part, the workflow corresponding to the loop execution structure in the markup language format At the first position of the position, add a first conditional jump node, the first conditional jump node is determined based on the loop execution structure, the first conditional jump node contains the first preset condition and link to a first function block node before the first location, the first conditional jump node is also linked to a second function block node based on the execution order of the sequential execution part, wherein at the first conditional jump node Where: when the first preset condition is satisfied, jump to the first function block node; when the first preset condition is not satisfied, execute the second function block based on the execution sequence node.
The workflow execution device (700) according to claim 8, characterized in that,

The conversion module (702), configured to determine the sequential execution structure and the conditional execution structure contained in the workflow in the markup language format; determine the sequential execution part of the workflow topology based on the sequential execution structure, the The sequential execution part includes the function block nodes in the sequential execution structure and the links between the function block nodes; in the sequential execution part, the workflow corresponding to the conditional execution structure in the markup language format At the second position of the position, add a second conditional jump node, the second conditional jump node is determined based on the conditional execution structure, the second conditional jump node contains the second preset condition and links To the third function block node after the second position, the second conditional jump node is also linked to the fourth function block node based on the execution order of the sequential execution part; wherein at the second conditional jump node Where: when the second preset condition is satisfied, execute the fourth function block node based on the execution order; when the preset condition is not satisfied, jump to the third function block node.
The workflow execution device (700) according to claim 8, characterized in that,

The conversion module (702), configured to determine the sequential execution structure, loop execution structure and conditional execution structure contained in the workflow in the markup language format; determine the order of the workflow topology based on the sequential execution structure Execution part, the sequential execution part includes function block nodes in the sequential execution structure and links between the function block nodes; in the sequential execution part, corresponding to the loop execution structure in the markup language At the first position of the workflow in the format, a first conditional jump node is added, the first conditional jump node is determined based on the loop execution structure, and the first conditional jump node contains the first preset conditional and linked to a first function block node before the first position, the first conditional jump node is also linked to a second function block node based on the execution order of the sequential execution part, wherein at the first At the condition jump node: when the first preset condition is satisfied, jump to the first function block node; when the first preset condition is not satisfied, execute the A second function block node; in the sequential execution part, at a second position corresponding to the position of the conditional execution structure in the workflow in the markup language format, add a second conditional jump node, the second A conditional jump node is determined based on the conditional execution structure, the second conditional jump node contains a second preset condition and is linked to a third function block node after the second position, and the second conditional jump node The jump node is also linked to the fourth function block node based on the execution sequence of the sequential execution part; wherein at the second condition jump node: when the second preset condition is satisfied, the execution sequence is executed based on the execution sequence The fourth function block node; when the preset condition is not satisfied, jump to the third function block node.
The workflow execution device (700) according to any one of claims 8-11, characterized in that,

The conversion module (702), configured to determine the parallel execution structure contained in the workflow in the markup language format; in the sequential execution part, the work corresponding to the parallel execution structure in the markup language format At a third position of the position of the flow, at least one parallel execution part is added, the at least one parallel execution part converges on the parallel sink node of the sequential execution part, and the parallel sink node is also based on the execution of the sequential execution part Sequentially linked to a fifth function block node; the parallel sink node is a logic AND node or a logic OR node, wherein when all parallel execution parts converged to the logic AND node are executed, the fifth function block node is executed ; Execute the fifth function block node when at least one parallel execution part of all parallel execution parts converged to the logic or node has been executed.
The workflow execution device (700) according to claim 12, characterized in that,

The conversion module (702), configured to add an empty node before the logical AND node when the first conditional jump node or the second conditional jump node jumps to the logical AND node, and jump the conditional node The node jumps to the empty node, wherein the empty node is used to indicate that no operation is performed, and the empty node is linked to the logical AND node based on the execution sequence of the sequential execution part.
The workflow execution device (700) according to any one of claims 8-11, further comprising:

A reverse interpretation module (704), configured to associate and store the workflow in the markup language format and the workflow topology; receive a reverse interpretation request adapted to the workflow requesting the workflow topology; respond to The reverse interpretation request provides the workflow in markup language format based on the associated storage.
A workflow execution device (500), characterized by comprising:

at least one memory (501) configured to store computer readable code;

At least one processor (502) configured to call the computer-readable code to execute the steps in the workflow execution method (800) according to any one of claims 1-7.
A computer-readable medium, characterized in that computer-readable instructions are stored on the computer-readable medium, and when the computer-readable instructions are executed by a processor, the processor executes the method according to claims 1-7. The workflow execution method (800) of any one of the steps in.
A computer program product tangibly stored on a computer-readable medium and comprising computer-readable instructions which, when executed, cause at least one processor to perform the The steps in the workflow execution method (800) according to any one of claims 1-7.