Metascene-based food industry equipment flow construction method and system

Abstract

The invention provides a meta-scene-based food industry equipment flow construction method and system, and the method comprises the steps: constructing a plurality of different meta-scenes based on the motion type and processing mode of a target object in a target scene; wherein the various different meta-scenes comprise a moving target + in-situ processing meta-scene, a moving target + workbench processing meta-scene, a static target + in-situ processing meta-scene and a static target + workbench processing meta-scene; according to the method, a process framework and a node mode are built for various different meta-scenes, and different nodes are linked based on the process framework, so that equipment works in sequence according to the process framework and the node mode.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of process building, in particular to a food industry equipment process building method and system based on meta-scenarios. BACKGROUND

[0002] In the automatic production process of the food industry, equipment needs to cope with different work scenarios of raw material processing, processing and manufacturing, packaging and storage, quality inspection, etc. Due to the variety of food, the complexity of processing technology, and the different needs of customers' factories for production processes, the traditional equipment workflow building method usually develops for a single specific scenario. When facing new production tasks or changes in scenarios, a large amount of code needs to be rewritten to adjust the overall workflow of the equipment, which not only has a long development cycle and high cost, but also has poor flexibility and universality. For example, when dealing with moving targets and stationary targets, and using different modes such as in-place processing and workbench processing, traditional equipment has difficulty in quickly switching workflows and cannot efficiently adapt to diversified production needs. With the increasing demand for production efficiency, flexibility and intelligence in the food industry, there is an urgent need for a device workflow building technology that can be quickly built, flexibly configured, and suitable for multiple scenarios to meet the growing diversified production needs of the industry. SUMMARY

[0003] In view of the defects in the prior art, the purpose of the present application is to provide a food industry equipment process building method and system based on meta-scenarios.

[0004] According to the food industry equipment process building method based on meta-scenarios provided by the present application, the following steps are included:

[0005] Step S1: Based on the motion type and processing mode of the target object in the target scenario, build multiple different meta-scenarios;

[0006] Among them, the multiple different meta-scenarios include: moving target + in-place processing meta-scenario, moving target + workbench processing meta-scenario, stationary target + in-place processing meta-scenario, and stationary target + workbench processing meta-scenario;

[0007] The moving target + in-place processing meta-scenario includes: the target object is a moving target, and the processing mode of the target object is an action processing mode in which the target object is processed in place;

[0008] The moving target + workbench processing meta-scenario includes: the target object is a moving target, and the processing mode of the target object is an action processing mode in which the target object is moved to a workbench for processing;

[0009] The stationary target + in-place processing meta-scenario includes: the target object is a stationary target, and the processing mode of the target object is an action processing mode in which the target object is processed in place;

[0010] The static target + workbench processing element scene includes: the target object is a static target, and the processing mode of the target object is a target object moving to the workbench and then being processed mode;

[0011] Step S2: respectively build a process framework and a node mode for a plurality of different element scenes, link different nodes based on the process framework, and make the device work sequentially according to the process framework and the node mode.

[0012] Preferably, the process framework and the node mode are respectively built for a plurality of different element scenes, different nodes are linked based on the process framework, and the device works sequentially according to the process framework and the node mode, including:

[0013] For the moving target + in-place processing element scene, an image acquisition and recognition node, a prediction node, and a trigger robot action node are constructed; a process framework is constructed based on the constructed image acquisition and recognition node, prediction node, and trigger robot action node;

[0014] The process framework is constructed based on the constructed image acquisition and recognition node, prediction node, and trigger robot action node, including:

[0015] First, the image acquisition and recognition node is used to acquire the moving target point, speed, and attribute information; then, the prediction node is used to predict the time for the moving target to reach a preset point based on the acquired moving target point and speed information; finally, the trigger robot action node is used to trigger the robot to follow the current pipeline speed and execute a corresponding processing action according to the predicted time for the moving target to reach the preset point.

[0016] Preferably, the process framework and the node mode are respectively built for a plurality of different element scenes, different nodes are linked based on the process framework, and the device works sequentially according to the process framework and the node mode, including:

[0017] For the moving target + workbench processing element scene, an image acquisition and recognition node, a prediction node, a moving target to workbench node, and a workbench action processing node are constructed; a process framework is constructed based on the constructed image acquisition and recognition node, prediction node, moving target to workbench node, and workbench action processing node;

[0018] The process framework is constructed based on the constructed image acquisition and recognition node, prediction node, moving target to workbench node, and workbench action processing node, including:

[0019] Firstly, the image acquisition recognition node is used to acquire the motion target point, speed and attribute information; secondly, the prediction node is used to predict the time when the motion target reaches the preset point based on the acquired motion target point and speed information; then, the mobile target to workbench node is used to trigger the equipment to move the target from the assembly line to the workbench according to the predicted time when the motion target reaches the preset point; finally, the workbench action processing node is used to execute the corresponding processing action after the target is moved to the workbench.

[0020] Preferably, the flow framework and node mode are respectively built for a plurality of different meta-scenes, different nodes are linked based on the flow framework, and the equipment sequentially works according to the flow framework and the node mode, including:

[0021] For the static target + in-place processing meta-scene, the image acquisition recognition node and the trigger robot action node are constructed; the flow framework is constructed based on the constructed image acquisition recognition node and the trigger robot action node.

[0022] The flow framework is constructed based on the constructed image acquisition recognition node and the trigger robot action node, including:

[0023] Firstly, the image acquisition recognition node is used to acquire the static target point and attribute information; then, the trigger robot action node is used to trigger the robot to move to the target position to execute the corresponding processing action according to the acquired static target point and attribute information.

[0024] Preferably, the flow framework and node mode are respectively built for a plurality of different meta-scenes, different nodes are linked based on the flow framework, and the equipment sequentially works according to the flow framework and the node mode, including:

[0025] For the static target + workbench processing meta-scene, the image acquisition recognition node, the mobile target to workbench node and the workbench action processing node are constructed; the flow framework is constructed based on the constructed image acquisition recognition node, the mobile target to workbench node and the workbench action processing node.

[0026] The flow framework is constructed based on the constructed image acquisition recognition node, the mobile target to workbench node and the workbench action processing node, including:

[0027] Firstly, the image acquisition recognition node is used to acquire the static target point and attribute information; then, the mobile target to workbench node is used to trigger the equipment to move the target from the initial position to the workbench according to the target point and attribute information; finally, the workbench action processing node is used to execute the corresponding processing action after the target is moved to the workbench.

[0028] Preferably, when the target scenario is a food production line, the visual parameters and execution parameters are adjusted accordingly for different meta-scenarios;

[0029] The adaptive adjustment of visual parameters and execution parameters includes: during the visual recognition of food materials, adaptive adjustment of visual parameters including similarity threshold and symmetry threshold;

[0030] During the hardware execution of food processing, the execution parameters, including speed, force, and motion trajectory, are adjusted accordingly.

[0031] When the meta-scene is a moving target + in-situ processing meta-scene, the visual parameter threshold is lowered to adapt to the visual recognition of food materials in motion; the execution parameters are adjusted to enable the robot to adapt to the current production line speed for follow-up work.

[0032] When the meta-scene is a moving target + workbench processing meta-scene, the visual parameter threshold is lowered to adapt to the visual recognition of food materials in motion; the execution parameters are adjusted to make the robot adapt to the current workbench processing work.

[0033] When the meta-scene is a stationary target + in-situ processing meta-scene, increase the visual parameter threshold to adapt to the visual recognition of food materials in a stationary state; adjust the execution parameters to enable the robot to adapt to the current production line speed for follow-up work.

[0034] When the meta-scene is a static target + workbench processing meta-scene, increase the visual parameter threshold to adapt to the visual recognition of food materials in a stationary state; adjust the execution parameters to make the robot adapt to the current workbench processing work.

[0035] According to the present invention, a food industry equipment process construction system based on meta-scenario includes:

[0036] Module M1: Based on the motion type and processing mode of the target object within the target scene, build various different meta-scenes;

[0037] Among them, the various different meta-scenes include: moving target + in-situ processing meta-scene, moving target + workbench processing meta-scene, stationary target + in-situ processing meta-scene, and stationary target + workbench processing meta-scene;

[0038] The moving target + in-situ processing meta-scenario includes: the target object is a moving target, and the processing mode of the target object is the target object performing action processing mode in in-situ.

[0039] The motion target + workbench processing meta-scenario includes: the target object is a motion target, and the processing mode of the target object is the action processing mode after the target object is moved to the workbench;

[0040] The static target + in-place processing element scene includes that the target object is a static target, and the processing mode of the target object is an action processing mode in which the target object performs an action in place.

[0041] The static target + workbench processing element scene includes that the target object is a static target, and the processing mode of the target object is an action processing mode in which the target object performs an action after being moved to a workbench.

[0042] The module M2 includes the following steps.

[0043] Preferably, the module M2 includes the following steps.

[0044] For the moving target + in-place processing element scene, an image acquisition and recognition node, a prediction node, and a trigger robot action node are constructed, and a flow framework is constructed based on the constructed image acquisition and recognition node, prediction node, and trigger robot action node.

[0045] The module M2 includes the following steps.

[0046] First, the image acquisition and recognition node is used to acquire the position, speed, and attribute information of the moving target. Then, the prediction node is used to predict the time at which the moving target reaches a preset position based on the acquired position and speed information of the moving target. Finally, the trigger robot action node is used to trigger the robot to follow the current flow line speed and simultaneously perform a corresponding processing action according to the predicted time at which the moving target reaches the preset position.

[0047] The module M2 includes the following steps.

[0048] For the moving target + workbench processing element scene, an image acquisition and recognition node, a prediction node, a moving target to workbench node, and a workbench action processing node are constructed, and a flow framework is constructed based on the constructed image acquisition and recognition node, prediction node, moving target to workbench node, and workbench action processing node.

[0049] The module M2 includes the following steps.

[0050] Firstly, the image acquisition recognition node is used to acquire the motion target point, speed and attribute information; secondly, the prediction node is used to predict the time when the motion target reaches the preset point based on the acquired motion target point and speed information; then, the mobile target to workbench node is used to trigger the equipment to move the target from the assembly line to the workbench according to the predicted time when the motion target reaches the preset point; finally, the workbench action processing node is used to execute the corresponding processing action after the target is moved to the workbench.

[0051] Preferably, the flow framework and node mode are respectively built for a plurality of different meta-scenes, different nodes are linked based on the flow framework, and the equipment sequentially works according to the flow framework and the node mode, including:

[0052] For the static target + in-place processing meta-scene, the image acquisition recognition node and the trigger robot action node are constructed; the flow framework is constructed based on the constructed image acquisition recognition node and the trigger robot action node.

[0053] The flow framework is constructed based on the constructed image acquisition recognition node and the trigger robot action node, including:

[0054] Firstly, the image acquisition recognition node is used to acquire the static target point and attribute information; then, the trigger robot action node is used to trigger the robot to move to the target position to execute the corresponding processing action according to the acquired static target point and attribute information.

[0055] The flow framework and node mode are respectively built for a plurality of different meta-scenes, different nodes are linked based on the flow framework, and the equipment sequentially works according to the flow framework and the node mode, including:

[0056] For the static target + workbench processing meta-scene, the image acquisition recognition node, the mobile target to workbench node and the workbench action processing node are constructed; the flow framework is constructed based on the constructed image acquisition recognition node, the mobile target to workbench node and the workbench action processing node.

[0057] The flow framework is constructed based on the constructed image acquisition recognition node, the mobile target to workbench node and the workbench action processing node, including:

[0058] Firstly, the image acquisition recognition node is used to acquire the static target point and attribute information; then, the mobile target to workbench node is used to trigger the equipment to move the target from the initial position to the workbench according to the target point and attribute information; finally, the workbench action processing node is used to execute the corresponding processing action after the target is moved to the workbench.

[0059] Preferably, when the target scene is a food pipeline, the visual parameters and the execution parameters are adjusted according to different meta-scenes.

[0060] The visual parameters and the execution parameters are adjusted according to different meta-scenes, and the adjusting comprises adjusting the visual parameters including the similarity threshold and the symmetry threshold during visual identification of the food material.

[0061] The execution parameters including the speed, the force and the motion trajectory are adjusted during hardware execution of the food material.

[0062] When the meta-scene is the moving target + in-place processing meta-scene, the visual parameter threshold is reduced to adapt to the visual identification of the moving food material, and the execution parameter is adjusted to adapt to the current pipeline speed of the robot.

[0063] When the meta-scene is the moving target + workbench processing meta-scene, the visual parameter threshold is reduced to adapt to the visual identification of the moving food material, and the execution parameter is adjusted to adapt to the current workbench processing work of the robot.

[0064] When the meta-scene is the static target + in-place processing meta-scene, the visual parameter threshold is increased to adapt to the visual identification of the static food material, and the execution parameter is adjusted to adapt to the current pipeline speed of the robot.

[0065] When the meta-scene is the static target + workbench processing meta-scene, the visual parameter threshold is increased to adapt to the visual identification of the static food material, and the execution parameter is adjusted to adapt to the current workbench processing work of the robot.

[0066] Compared with the prior art, the present application has the following beneficial effects:

[0067] 1. The present application abstracts the food industry equipment working scene into four meta-scenes, and builds a special process framework and node for each meta-scene, so that the equipment can quickly adapt to the production demand under different combination modes of moving target, static target and in-place processing, workbench processing, etc. When the production scene changes, there is no need to re-write a large amount of code, only the logic and parameters of the running part in the process need to be adjusted, which significantly improves the adaptation efficiency of the equipment to diversified production tasks, and greatly shortens the production cycle of new products or new orders.

[0068] 2. The present application builds a process framework based on a node mode, and decomposes the equipment working process into a plurality of nodes with clear functions, and the nodes are closely connected in logical order, so that the equipment can orderly and efficiently execute the work tasks. BRIEF DESCRIPTION OF DRAWINGS

[0069] Other features, objects, and advantages of the application will become more apparent from the following detailed description of non-limiting embodiments thereof, when read in conjunction with the accompanying drawings:

[0070] Figure 1 A method flow chart for building a food industry equipment process based on meta-scenarios is provided. DETAILED DESCRIPTION

[0071] The application will be described in detail below with specific embodiments. The following examples will help those skilled in the art to further understand the application, but in no way limit the application. It should be noted that for those skilled in the art, without departing from the concept of the application, a number of changes and improvements can be made. These are within the scope of the application.

[0072] Example 1

[0073] A method for building a food industry equipment process based on meta-scenarios is provided according to the application, as shown in Figure 1 , comprising:

[0074] Step S1: building a plurality of different meta-scenarios based on the motion type and processing mode of the target object in the target scene;

[0075] Among them, the plurality of different meta-scenarios include: moving target + in-place processing meta-scenario, moving target + workbench processing meta-scenario, stationary target + in-place processing meta-scenario, and stationary target + workbench processing meta-scenario;

[0076] The moving target + in-place processing meta-scenario includes: the target object is a moving target, and the processing mode of the target object is an action processing mode in which the target object is in place; for example: flow line follow-up work;

[0077] The moving target + workbench processing meta-scenario includes: the target object is a moving target, and the processing mode of the target object is an action processing mode in which the target object is moved to a workbench; for example: sorting and temporary storage, quality inspection and abnormal interception;

[0078] The stationary target + in-place processing meta-scenario includes: the target object is a stationary target, and the processing mode of the target object is an action processing mode in which the target object is in place; for example: in-situ packaging and surface treatment;

[0079] The stationary target + workbench processing meta-scenario includes: the target object is a stationary target, and the processing mode of the target object is an action processing mode in which the target object is moved to a workbench; for example: finishing processing;

[0080] Step S2: respectively build a process framework and a node mode for a plurality of different meta-scenes, link different nodes based on the process framework, so that the device works sequentially according to the process framework and the node mode.

[0081] In this embodiment, the target scene is a food processing assembly line, and the corresponding device meta-scene is selected according to the type of work to be performed by the food industry device, the customer factory requirements, and the like.

[0082] Specifically, the process framework and the node mode are respectively built for a plurality of different meta-scenes, different nodes are linked based on the process framework, so that the device works sequentially according to the process framework and the node mode, including:

[0083] For the moving target + in-place processing meta-scene, an image acquisition and recognition node, a prediction node, and a trigger robot action node are constructed; a process framework is constructed based on the constructed image acquisition and recognition node, prediction node, and trigger robot action node.

[0084] The image acquisition and recognition node: uses a camera to take pictures of a moving target to obtain image information, analyzes and processes the photographed image through AI technology, and obtains target point position, speed, and attribute information; for example, in a food assembly line, a food moving on a conveyor belt is photographed, and the position, moving speed, and type of the food are identified.

[0085] The prediction node: predicts the point position (i.e., the collision point) that the target is expected to reach in the future based on the target point position, speed, and other information obtained by the image acquisition and recognition node; for example, according to the current position and moving speed of the food, it is calculated when the food will reach a designated operation area on the conveyor belt.

[0086] The trigger robot action node: controls the robot to perform corresponding actions such as grabbing, packaging, detection, and other operations when the target reaches the predicted point position. In the moving target + in-place processing meta-scene, the robot also needs to work with the current assembly line speed to ensure the accuracy of action execution. For example, the robot moves with the assembly line and performs real-time packaging operations when the food arrives.

[0087] The process framework is constructed based on the constructed image acquisition and recognition node, prediction node, and trigger robot action node, including:

[0088] First, the image acquisition recognition node is used to acquire the position, speed and attribute information of the moving target; then, the prediction node is used to predict the time when the moving target reaches the preset position based on the acquired position and speed information of the moving target; finally, the robot action triggering node is used to trigger the robot to follow the current pipeline speed and execute corresponding processing actions according to the time when the moving target reaches the preset position predicted by the prediction node. After completing an operation, the process returns to the image acquisition recognition node to start the next round of cycle and continuously process the moving target. This enables the device to continuously detect and process food on the pipeline.

[0089] For the moving target + workbench processing element scene, an image acquisition recognition node, a prediction node, a moving target to workbench node and a workbench action processing node are constructed; a process framework is constructed based on the constructed image acquisition recognition node, prediction node, moving target to workbench node and workbench action processing node;

[0090] The image acquisition recognition node: uses a camera to take pictures of the moving target, analyzes the images by means of AI technology, and acquires the target position, speed and attribute information, which is the data source of the entire process. For example, in the food sorting scene, different types of food moving on the pipeline are photographed, and the food position, moving speed and food name, specification and other attribute information are identified.

[0091] The prediction node: calculates the target's future predicted position based on the target position, speed and other data acquired by the image acquisition recognition node. This step is to prepare for subsequent operations, such as calculating when the food on the pipeline will reach the appropriate grabbing position.

[0092] The moving target to workbench node: when the target reaches the predicted position, the device is controlled to move the target from the pipeline to the workbench, which may involve operations such as mechanical arm grabbing, conveyor transfer, etc. For example, the food on the pipeline is accurately grabbed and placed on a specific sorting workbench.

[0093] The workbench action processing node: after the target reaches the workbench, corresponding processing actions are executed, such as quality inspection, packaging, labeling, etc. For example, weight detection, defect inspection, or packaging and boxing processes are performed on the food on the workbench.

[0094] The process framework constructed based on the constructed image acquisition recognition node, prediction node, moving target to workbench node and workbench action processing node includes:

[0095] First, based on the image acquisition recognition node, the motion target point, speed and attribute information are acquired; then, based on the prediction node, the time for the motion target to reach the preset point is predicted based on the acquired motion target point and speed information; then, based on the moving target to the workbench node, the equipment is triggered to move the target from the assembly line to the workbench according to the time for the motion target to reach the preset point predicted; finally, when the target is moved to the workbench, the corresponding processing action is executed based on the workbench action processing node. After completing the processing of one target, it is judged whether to continue processing the next target. If the next target is processed, the image acquisition recognition node is returned to start the cycle, otherwise the current batch work is ended. For example, when the set number of food to be processed is completed or a stop instruction is received, the entire work flow is ended.

[0096] For the static target + in-place processing meta-scene, an image acquisition recognition node and a trigger robot action node are constructed; a flow framework is constructed based on the constructed image acquisition recognition node and trigger robot action node.

[0097] The image acquisition recognition node: through the camera, the static target is photographed, and the AI image recognition technology is used to analyze and acquire the point information (such as the specific coordinate position in the work area) and attribute information (such as the type, size and appearance state of the food) of the target. For example, when processing the static placed cakes, this node identifies the placement position of the cakes, the taste category of the cakes, the shape and size, etc., to provide basic data support for subsequent operations.

[0098] The trigger robot action node: based on the target point and attribute information acquired by the image acquisition recognition node, the robot is controlled to move to the target position and execute the corresponding operation action. For example, if the target is a cake that needs to be decorated, the robot will move to the corresponding position according to the identified cake position and execute the actions of smearing sauce, scattering decorative sugar particles, etc.; if the target is a food for quality detection, the robot will execute visual detection, weight detection, etc.

[0099] The flow framework constructed based on the constructed image acquisition recognition node and trigger robot action node includes:

[0100] First, based on the image acquisition recognition node, the static target point and attribute information are acquired; then, based on the trigger robot action node, the robot is triggered to move to the target position to perform the corresponding processing action according to the acquired static target point and attribute information. After completing the processing of one target, it is judged whether there are other targets to be processed. If there are other static targets to be processed, the process returns to the image acquisition recognition node, and the next cycle of processing the new target is started; if all targets have been processed, the entire workflow ends. For example, after completing the processing of a batch of static placed bread, it is judged whether there are other batches of bread in the warehouse to be processed, and the flow direction is determined.

[0101] For the static target + workbench processing meta-scene, an image acquisition recognition node, a target moving to workbench node, and a workbench action processing node are constructed, and a process framework is constructed based on the constructed image acquisition recognition node, target moving to workbench node, and workbench action processing node;

[0102] The image acquisition recognition node: as the starting point of the process, this node uses a camera to take pictures of the static target, and analyzes the image with the help of AI technology to obtain the point information (such as the specific coordinates in the work area) and attribute information (such as the type, size, appearance state of food, etc.) of the target. For example, when processing the static placed cans, this node can identify the position of the cans on the conveyor belt, as well as the brand, capacity, and packaging color of the cans, providing data support for subsequent operations.

[0103] The target moving to workbench node: based on the target point and attribute information obtained by the image acquisition recognition node, the mechanical arm or other handling device is controlled to move the target from the initial position to the workbench. For example, the mechanical arm accurately grasps and places the cans on the detection workbench according to the recognized position of the cans.

[0104] The workbench action processing node: after the target reaches the workbench, the corresponding processing action is performed according to the target attribute, which may be quality detection (such as checking the sealing of the cans, weight detection), packaging reinforcement, labeling, etc. For example, the sealing of the cans is detected on the workbench, or the product label is pasted.

[0105] The process framework constructed based on the constructed image acquisition recognition node, target moving to workbench node, and workbench action processing node includes:

[0106] First, based on the image acquisition recognition node to obtain the static target point and attribute information; then based on the mobile target to the workbench node according to the target point and attribute information trigger device to move the target from the initial position to the workbench; finally, after the target moves to the workbench, the workbench action processing node executes the corresponding processing action. After completing the target processing, it is judged whether there are other static targets to be processed. If there are targets to be processed, the flow image acquisition recognition node starts the next round of circulation; if all targets have been processed, the whole work flow ends. For example, after completing the processing of a batch of static cans, it is judged whether there are other batches of cans in the warehouse waiting for processing to determine the flow direction.

[0107] Specifically, the above four kinds of meta-scene working methods are completely different in the underlying logic of device operation processing. After abstracting the logic of each scene, it can basically cover most of the work in the food factory. Then, by analyzing and judging the raw materials of the work target, some basic parameters can be adjusted to adapt to different work.

[0108] More specifically, compared with other traditional industrial assembly lines, food assembly lines often have non-fixed, non-standard, and random materials. Specifically, the raw materials of traditional industries or non-food industries are often fixed in shape and position, and the processing method is relatively uniform. However, in the food industry, due to the natural differences between food materials, even the same part of meat or the same fruit on a tree, there is no complete similarity. Therefore, when performing visual identification and hardware execution, it is necessary to dynamically adjust various parameters.

[0109] Among them, the visual parameters: in the process of visual identification of food materials, the visual parameters including similarity threshold, symmetry threshold, standard part parameter value, and cutting threshold are adjusted accordingly; adjusting these values is to ensure that the device can: identify the raw materials and their positions that need to be operated, and calculate the way that will produce less waste (if any);

[0110] Execution parameter: after visual recognition, the parameter provided to the execution unit for execution, the raw materials on a food pipeline are not standard arrangement, they may appear in any position, their posture is not the same direction and the same face, so the vision sees the raw material and its state, tells the execution unit the specific position (predict the position of appearance), and the execution unit executes the corresponding action. For static / dynamic raw materials, whether it is to execute cutting, spraying, grabbing action, or in situ / workbench processing, the execution unit needs to be as fine and customized as a human hand. Therefore, adjust the speed, acceleration, force, working range, motion trajectory, angle of the execution unit to realize the standard output of non-standard raw materials in one processing. Adjust the force and depth, such as cutting, pressing; adjust the motion trajectory, angle, such as grabbing, cutting, etc.

[0111] For four different meta-scenes, construct the corresponding node and process framework, and update and save in the cloud in real time. The device obtains and downloads the corresponding process package file based on the cloud. The device executes the device action according to the downloaded process package file.

[0112] The application also provides a food industry device process building system based on a meta-scene, which can be realized by executing the process steps of the food industry device process building method based on a meta-scene, that is, those skilled in the art can understand the food industry device process building method based on a meta-scene as the preferred embodiment of the food industry device process building system based on a meta-scene.

[0113] Those skilled in the art know that in addition to implementing the system, device and each module thereof provided by the application in a pure computer readable program code manner, the same program can also be realized in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers and embedded microcontrollers by logically programming the method steps. Therefore, the system, device and each module thereof provided by the application can be considered as a hardware component, and the modules included therein for realizing various programs can also be considered as structures in the hardware component; the modules for realizing various functions can also be considered as both software programs for realizing methods and structures in the hardware component.

[0114] The specific embodiments of the application are described above. It should be understood that the application is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essential content of the application. In the case of no conflict, the embodiments of the application and the features in the embodiments can be combined with each other at will.

Claims

1. A method for constructing equipment processes in the food industry based on meta-scenarios, characterized in that, include: Step S1: Based on the motion type and processing mode of the target object within the target scene, build multiple different meta-scenes; Among them, the various different meta-scenes include: moving target + in-situ processing meta-scene, moving target + workbench processing meta-scene, stationary target + in-situ processing meta-scene, and stationary target + workbench processing meta-scene; The moving target + in-situ processing meta-scenario includes: the target object is a moving target, and the processing mode of the target object is the target object performing action processing mode in in-situ. The motion target + workbench processing meta-scenario includes: the target object is a motion target, and the processing mode of the target object is the action processing mode after the target object is moved to the workbench; The static target + in-situ processing meta-scenario includes: the target object is a static target, and the processing mode of the target object is the target object performing action processing mode in-situ. The static target + workbench processing meta-scenario includes: the target object is a static target, and the target object processing mode is the target object being moved to the workbench and then performing an action processing mode. Step S2: Build process frameworks and node methods for various different meta-scenarios, and link different nodes based on the process framework so that the device works in sequence according to the process framework and node methods.

2. The method for constructing equipment processes in the food industry based on meta-scenarios according to claim 1, characterized in that, The process framework and node methods are built for various different meta-scenarios, and different nodes are linked based on the process framework so that the device works sequentially according to the process framework and node methods, including: For the aforementioned moving target + in-situ processing scenario, an image acquisition and recognition node, a prediction node, and a robot action triggering node are constructed; based on the constructed image acquisition and recognition node, prediction node, and robot action triggering node, a process framework is built; The framework for constructing image acquisition and recognition nodes, prediction nodes, and robot action triggering nodes includes: First, the image acquisition and recognition node acquires the position, speed, and attribute information of the moving target. Then, the prediction node predicts the time it will take for the moving target to reach the preset position based on the acquired position and speed information. Finally, the robot action trigger node triggers the robot to move with the current production line speed according to the predicted time it will reach the preset position, and performs corresponding processing actions.

3. The method for constructing equipment processes in the food industry based on meta-scenarios according to claim 1, characterized in that, The process framework and node methods are built for various different meta-scenarios, and different nodes are linked based on the process framework so that the device works sequentially according to the process framework and node methods, including: For the aforementioned moving target + workbench processing scenario, an image acquisition and recognition node, a prediction node, a moving target to workbench node, and a workbench motion processing node are constructed; a process framework is built based on the constructed image acquisition and recognition node, prediction node, moving target to workbench node, and workbench motion processing node. The framework for constructing the image acquisition and recognition node, prediction node, moving target to workbench node, and workbench action processing node includes: First, the image acquisition and recognition node acquires the location, speed, and attribute information of the moving target. Then, the prediction node predicts the time it will take for the moving target to reach the preset location based on the acquired location and speed information. Next, the target-to-workbench node triggers the device to move the target from the production line to the workbench based on the predicted time it will reach the preset location. Finally, once the target has moved to the workbench, the workbench motion processing node executes the corresponding processing action.

4. The method for building equipment processes in the food industry based on meta-scenarios according to claim 1, characterized in that, The process framework and node methods are built for various different meta-scenarios, and different nodes are linked based on the process framework so that the device works sequentially according to the process framework and node methods, including: For the static target + in-situ processing meta-scenario, an image acquisition and recognition node and a robot action triggering node are constructed; a process framework is built based on the constructed image acquisition and recognition node and robot action triggering node. The framework for constructing image acquisition and recognition nodes and triggering robot action nodes includes: First, the image acquisition and recognition node acquires the location and attribute information of the stationary target; then, the robot action trigger node triggers the robot to move to the target position and perform the corresponding processing action based on the acquired location and attribute information of the stationary target.

5. The method for constructing equipment processes in the food industry based on meta-scenarios according to claim 1, characterized in that, The process framework and node methods are built for various different meta-scenarios, and different nodes are linked based on the process framework so that the device works sequentially according to the process framework and node methods, including: For the aforementioned static target + workbench processing scenario, an image acquisition and recognition node, a moving target to workbench node, and a workbench action processing node are constructed. Based on the constructed image acquisition and recognition node, moving target to workbench node, and workbench action processing node, a process framework is built. The framework for constructing the image acquisition and recognition node, the moving target to the workbench node, and the workbench action processing node includes: First, the image acquisition and recognition node acquires the location and attribute information of the stationary target; then, the moving target to workbench node triggers the device to move the target from its initial position to the workbench based on the target location and attribute information; finally, after the target moves to the workbench, the workbench action processing node executes the corresponding processing action.

6. The method for building equipment processes in the food industry based on meta-scenarios according to claim 1, characterized in that, When the target scenario is a food production line, the visual parameters and execution parameters are adjusted accordingly for different meta-scenarios. The adaptive adjustment of visual parameters and execution parameters includes: during the visual recognition of food materials, adaptive adjustment of visual parameters including similarity threshold and symmetry threshold; During the hardware execution of food processing, the execution parameters, including speed, force, and motion trajectory, are adjusted accordingly. When the meta-scene is a moving target + in-situ processing meta-scene, the visual parameter threshold is lowered to adapt to the visual recognition of food materials in motion; the execution parameters are adjusted to enable the robot to adapt to the current production line speed for follow-up work. When the meta-scene is a moving target + workbench processing meta-scene, the visual parameter threshold is lowered to adapt to the visual recognition of food materials in motion; the execution parameters are adjusted to make the robot adapt to the current workbench processing work. When the meta-scene is a stationary target + in-situ processing meta-scene, increase the visual parameter threshold to adapt to the visual recognition of food materials in a stationary state; adjust the execution parameters to enable the robot to adapt to the current production line speed for follow-up work. When the meta-scene is a static target + workbench processing meta-scene, increase the visual parameter threshold to adapt to the visual recognition of food materials in a stationary state; adjust the execution parameters to make the robot adapt to the current workbench processing work.

7. A food industry equipment process construction system based on meta-scenario, characterized in that, include: Module M1: Based on the motion type and processing mode of the target object within the target scene, build various different meta-scenes; Among them, the various different meta-scenes include: moving target + in-situ processing meta-scene, moving target + workbench processing meta-scene, stationary target + in-situ processing meta-scene, and stationary target + workbench processing meta-scene; The moving target + in-situ processing meta-scenario includes: the target object is a moving target, and the processing mode of the target object is the target object performing action processing mode in in-situ. The motion target + workbench processing meta-scenario includes: the target object is a motion target, and the processing mode of the target object is the action processing mode after the target object is moved to the workbench; The static target + in-situ processing meta-scenario includes: the target object is a static target, and the processing mode of the target object is the target object performing action processing mode in-situ. The static target + workbench processing meta-scenario includes: the target object is a static target, and the target object processing mode is the target object being moved to the workbench and then performing an action processing mode. Module M2: It builds process frameworks and node methods for various different meta-scenarios, and links different nodes based on the process framework so that the device can work in sequence according to the process framework and node method.

8. The food industry equipment process construction system based on meta-scenario as described in claim 7, characterized in that, The process framework and node methods are built for various different meta-scenarios, and different nodes are linked based on the process framework so that the device works sequentially according to the process framework and node methods, including: For the aforementioned moving target + in-situ processing scenario, an image acquisition and recognition node, a prediction node, and a robot action triggering node are constructed; based on the constructed image acquisition and recognition node, prediction node, and robot action triggering node, a process framework is built; The framework for constructing image acquisition and recognition nodes, prediction nodes, and robot action triggering nodes includes: First, the image acquisition and recognition node acquires the position, speed, and attribute information of the moving target. Then, the prediction node predicts the time it will take for the moving target to reach the preset position based on the acquired position and speed information. Finally, the robot action trigger node triggers the robot to move with the current production line speed and execute corresponding processing actions based on the predicted time it will take for the moving target to reach the preset position. The process framework and node methods are built for various different meta-scenarios, and different nodes are linked based on the process framework so that the device works sequentially according to the process framework and node methods, including: For the aforementioned moving target + workbench processing scenario, an image acquisition and recognition node, a prediction node, a moving target to workbench node, and a workbench motion processing node are constructed; a process framework is built based on the constructed image acquisition and recognition node, prediction node, moving target to workbench node, and workbench motion processing node. The framework for constructing the image acquisition and recognition node, prediction node, moving target to workbench node, and workbench action processing node includes: First, the image acquisition and recognition node acquires the location, speed, and attribute information of the moving target. Then, the prediction node predicts the time it will take for the moving target to reach the preset location based on the acquired location and speed information. Next, the target-to-workbench node triggers the device to move the target from the production line to the workbench based on the predicted time it will reach the preset location. Finally, once the target has moved to the workbench, the workbench motion processing node executes the corresponding processing action.

9. The food industry equipment process construction system based on meta-scenario as described in claim 7, characterized in that, The process framework and node methods are built for various different meta-scenarios, and different nodes are linked based on the process framework so that the device works sequentially according to the process framework and node methods, including: For the static target + in-situ processing meta-scenario, an image acquisition and recognition node and a robot action triggering node are constructed; a process framework is built based on the constructed image acquisition and recognition node and robot action triggering node. The framework for constructing image acquisition and recognition nodes and triggering robot action nodes includes: First, the image acquisition and recognition node acquires the location and attribute information of the stationary target; then, the robot action trigger node triggers the robot to move to the target position and perform the corresponding processing action based on the acquired location and attribute information of the stationary target. The process framework and node methods are built for various different meta-scenarios, and different nodes are linked based on the process framework so that the device works sequentially according to the process framework and node methods, including: For the aforementioned static target + workbench processing scenario, an image acquisition and recognition node, a moving target to workbench node, and a workbench action processing node are constructed. Based on the constructed image acquisition and recognition node, moving target to workbench node, and workbench action processing node, a process framework is built. The framework for constructing the image acquisition and recognition node, the moving target to the workbench node, and the workbench action processing node includes: First, the image acquisition and recognition node acquires the location and attribute information of the stationary target; then, the moving target to workbench node triggers the device to move the target from its initial position to the workbench based on the target location and attribute information; finally, after the target moves to the workbench, the workbench action processing node executes the corresponding processing action.

10. The food industry equipment process construction system based on meta-scenario as described in claim 7, characterized in that, When the target scenario is a food production line, the visual parameters and execution parameters are adjusted accordingly for different meta-scenarios. The adaptive adjustment of visual parameters and execution parameters includes: during the visual recognition of food materials, adaptive adjustment of visual parameters including similarity threshold and symmetry threshold; During the hardware execution of food processing, the execution parameters, including speed, force, and motion trajectory, are adjusted accordingly. When the meta-scene is a moving target + in-situ processing meta-scene, the visual parameter threshold is lowered to adapt to the visual recognition of food materials in motion; the execution parameters are adjusted to enable the robot to adapt to the current production line speed for follow-up work. When the meta-scene is a moving target + workbench processing meta-scene, the visual parameter threshold is lowered to adapt to the visual recognition of food materials in motion; the execution parameters are adjusted to make the robot adapt to the current workbench processing work. When the meta-scene is a stationary target + in-situ processing meta-scene, increase the visual parameter threshold to adapt to the visual recognition of food materials in a stationary state; adjust the execution parameters to enable the robot to adapt to the current production line speed for follow-up work. When the meta-scene is a static target + workbench processing meta-scene, increase the visual parameter threshold to adapt to the visual recognition of food materials in a stationary state; adjust the execution parameters to make the robot adapt to the current workbench processing work.

Images