A Controller and Control Method for Industrial Embodied Intelligent Manufacturing

Abstract

The present invention discloses a controller and a control method for industrial embodied intelligent manufacturing, including: a sensing module for acquiring multi-modal sensing data during the execution of an industrial production task; a computing module for inferring the current attribute data of at least one target object in the industrial production task according to the multi-modal sensing data, and generating a corresponding control decision according to the current attribute data of each target object and the preset target attribute data corresponding to each target object at present; a control module for converting the control decision into a control signal for functionally controlling intelligent machine equipment; and a driving module for driving the intelligent machine equipment to perform corresponding operations in the industrial production task according to the control signal. The technical solution in the embodiment of the present invention realizes intelligent and efficient production in industrial manufacturing.

Description

Technical Field

[0001] The present invention relates to the field of intelligent controllers, and particularly to a controller and a control method for industrial embodied intelligent manufacturing. Background Art

[0002] In the field of industrial manufacturing, traditional controllers are designed based on fixed programming logic and preset rules, and it is difficult to automatically and flexibly adjust control strategies in complex process scenarios; in multi-device collaboration and overall optimization of production lines, numerous mechanical devices, sensors, and automation devices need to cooperate closely, but traditional controllers can only control individual machines and it is difficult to perform global optimization scheduling. Traditional controllers lack the capabilities of autonomous perception and decision-making in complex industrial environments, are difficult to adapt to dynamic changes, do not have edge computing and intelligent analysis functions, cannot quickly process data locally and analyze production situations, and are difficult to respond to the demands of the production site in a timely manner, failing to meet the urgent needs of intelligent and efficient industrial manufacturing production and becoming an important bottleneck in industrial upgrading. Summary of the Invention

[0003] The present invention provides a controller and a control method for industrial embodied intelligent manufacturing to at least solve the deficiencies existing in the above prior art.

[0004] According to one aspect of the present invention, there is provided a controller for industrial embodied intelligent manufacturing, comprising:

[0005] A sensing module, configured to obtain multi-modal sensing data during the execution of an industrial production task;

[0006] A computing module, configured to infer the current attribute data of at least one target object in the industrial production task according to the multi-modal sensing data, and generate corresponding control decisions according to the current attribute data of each target object and the preset target attribute data corresponding to each target object at present;

[0007] A control module, configured to convert the control decision into a control signal for functionally controlling intelligent machine devices;

[0008] A driving module, configured to drive the intelligent machine devices to perform corresponding operations during the execution of the industrial production task according to the control signal.

[0009] Furthermore, the computing module includes: an inference unit and a decision-making unit;

[0010] The inference unit is used to preprocess the multi-modal sensing data to obtain standardized sensing data, and input the standardized sensing data into a pre-configured inference model to obtain the current attribute data of at least one corresponding target object; the inference model is a pre-trained model that can perform inference based on the input standardized sensing data and output the current attribute data of at least one target object in the industrial production task;

[0011] The decision-making unit is used to generate corresponding control decisions according to the differences between the current attribute data of each target object and the preset target attribute data corresponding to each target object at present.

[0012] Furthermore, the inference unit is also used to load a model configuration file from an edge device, and the model configuration file includes the inference model and the data format of the standardized sensing data input into the inference model;

[0013] Among them, the data format includes: high-precision FP32 format, low-precision INT8 format.

[0014] Furthermore, the decision-making unit is specifically used to generate a first control decision when the differences between the current attribute data of each target object and the preset target attribute data corresponding to each target object at present are within a preset reasonable difference range; and, when the differences between the current attribute data of each target object and the preset target attribute data corresponding to each target object at present are outside the preset reasonable difference range, generate a second control decision;

[0015] Among them, one of the first control decision and the second control decision is used to control the intelligent machine device to perform a target operation, and the other of the two is used to control the intelligent machine device not to perform a target operation.

[0016] Furthermore, the decision-making unit is specifically used to input the current attribute data of each target object and the preset target attribute data corresponding to each target object at present into a pre-configured reinforcement learning model to obtain corresponding control decisions;

[0017] The control decision can control the intelligent machine device to perform corresponding operations so that each target object generates corresponding changed attribute data, and the goal of the reinforcement learning model is to make the current attribute data of each target object plus the corresponding changed attribute data can reach the corresponding preset target attribute data as much as possible.

[0018] Furthermore, the inference model is selected from one of a mechanism model, a neural network model, and a mechanism-neural network model;

[0019] The mechanism model is used to perform data conversion on the input standardized sensing data according to the process principle and constraint conditions, generate mechanism constraint data, and use it as the current attribute data of at least one target object;

[0020] The neural network model is used to perform feature recognition on the input standardized sensing data, generate high-dimensional feature representation data, and use it as the current attribute data of at least one target object;

[0021] The mechanism-neural network model is used to first perform data screening and conversion on the input standardized sensing data to generate mechanism constraint data; then perform feature recognition on the mechanism constraint data to obtain high-dimensional feature representation data, and use it as the current attribute data of at least one target object.

[0022] Further, the control module is specifically used to convert the control decision into functional control of intelligent machine equipment, construct functional control programs for each functional control through programming languages; when the functional control programs are running, generate control signals;

[0023] The functional control includes at least one of logical control, timing control, counting control, analog quantity control, closed-loop control, position control, speed control, and sequential control;

[0024] The programming languages include at least one of instruction list, structured text, function block diagram, ladder diagram, and sequential function chart.

[0025] Further, the control signals include at least one of digital control instructions, sensor analog signals, and servo motor configuration signals;

[0026] The drive module includes a digital signal drive unit, an analog signal drive unit, and a servo control unit;

[0027] The digital signal drive unit is used to convert digital control instructions into drive signals recognizable by intelligent machine equipment;

[0028] The analog signal drive unit is used to convert sensor analog signals into control inputs recognizable by intelligent machine equipment;

[0029] The servo control unit is used to convert servo motor configuration signals into servo motor configurations recognizable by intelligent machine equipment.

[0030] Further, it further includes:

[0031] A communication module, which is used to communicate with a cloud server to receive a model configuration file generated by the cloud server and store the model configuration file at the edge device; the cloud server is used to collect the execution results of intelligent machine devices and generate a new model configuration file according to the execution results;

[0032] The communication module supports wired, wireless and Internet of Things communications and is compatible with Modbus and OPC UA protocols.

[0033] According to another aspect of the present invention, there is provided a control method for a controller for industrial embodied intelligent manufacturing, including:

[0034] S1. Obtain multi-modal sensing data during the execution of an industrial production task;

[0035] S2. Infer the current attribute data of at least one target object in the industrial production task according to the multi-modal sensing data, and generate corresponding control decisions according to the current attribute data of each target object and the preset target attribute data corresponding to each target object at present;

[0036] S3. Convert the control decision into a control signal for functionally controlling an intelligent machine device;

[0037] S4. Drive the intelligent machine device to perform corresponding operations in the industrial production task according to the control signal.

[0038] According to another aspect of the present invention, there is provided an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, the steps of any one of the above methods are implemented.

[0039] According to another aspect of the present invention, there is provided a computer-readable storage medium, and the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of any one of the above methods are implemented.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] In the technical solution of the embodiment of the present invention, through the powerful reasoning and decision-making functions of the computing module, it can quickly adapt to environmental changes, optimize the collaborative work between intelligent machine devices, and significantly improve product quality; the cooperation between the sensing module and the control module ensures the fine regulation of the production process, improves human-machine cooperation, reduces the operation difficulty, enhances flexibility, enhances system compatibility and scalability, and brings significant economic benefits and competitiveness to enterprises. It can greatly improve industrial production efficiency.

[0042] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present invention, nor is it used to limit the scope of the present invention. Other features of the present invention will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0044] Figure 1 is a schematic structural diagram of a controller for industrial embodied intelligent manufacturing according to an embodiment of the present invention;

[0045] Figure 2 is a schematic flow diagram of a cold rolling process applicable to an embodiment of the present invention;

[0046] Figure 3 is a schematic flow diagram of a control method of a controller for industrial embodied intelligent manufacturing according to an embodiment of the present invention;

[0047] Figure 4 is a schematic structural diagram of an electronic device for implementing the control method of the controller for industrial embodied intelligent manufacturing according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0048] In order to enable those skilled in the art to better understand the solution of the present invention, the following clearly and completely describes the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0049] It should be noted that it should be understood that when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0050] It should also be understood that the terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application. As used in the specification of this application and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include the plural forms. Embodiment 1

[0051] Figure 1 It is a schematic structural diagram of a controller for industrial embodied intelligent manufacturing according to an embodiment of the present invention. As Figure 1 shown, the controller includes: a sensing module 100, a computing module 200, a control module 300, a driving module 400, and a communication module 500.

[0052] The sensing module 100 is used to obtain multi-modal sensing data during the execution of industrial production tasks.

[0053] In the embodiment of the present invention, industrial production tasks include: industrial production operations such as steel smelting and cold rolling production. The sensing module 100 can perform sensor data fusion on the data collected from different sensors to obtain multi-modal sensing data. Among them, the types of sensors include: tactile sensors, temperature sensors, inclination sensors, pressure sensors, pH value sensors, vision sensors, displacement sensors, speed sensors, atmosphere sensors, tension sensors, and position sensors, etc.; the multi-modal sensing data includes: various sensing signals such as switch signals, pulse signals, analog signals, audio signals, and image signals.

[0054] Sensor fusion is a process of integrating data from multiple sensors to provide more accurate and reliable information. By carrying out all-round and multi-modal information collection work at the industrial production site, the all-round and accurate collection of information is realized. In robot navigation, sensor fusion can help the robot more accurately locate its own position and the environment, thereby improving the efficiency and safety of navigation. In an industrial automation system, sensor fusion can real-time monitor various parameters during the production process to ensure the stability and safety of the production process.

[0055] The computing module 200 is used to infer the current attribute data of at least one target object in the industrial production task according to the multi-modal sensing data, and generate corresponding control decisions according to the current attribute data of each target object and the preset target attribute data corresponding to each target object at present.

[0056] It should be noted that the target object refers to various reference objects that can be directly or indirectly observed in industrial production tasks; the attribute data refers to the reference attributes that can directly or indirectly reflect production information in industrial production tasks. For example: in steel smelting, the target object can be molten steel, etc., and the attribute data can be the temperature range, composition distribution, etc. of the molten steel; while in cold rolling production, the target object can be acid solution, strip steel, rolling rolls, etc., and the attribute data can be the consumption rate of the acid solution, pickling effect of the strip steel, deformation law of the strip steel, tissue transformation process of the strip steel, type of surface defects of the strip steel, wear condition of the rolling rolls, stability of the rolling rolls, etc.

[0057] Since the attribute data of most target objects cannot be directly obtained through sensors or mathematical calculations, for example: pickling effect of strip steel, wear condition of rolling rolls, stability of rolling rolls, etc. Therefore, the technical solution in the embodiments of the present invention cleverly designs a variety of inference models, which can, through inference, based on the multi-modal sensing data obtained during the execution of industrial production tasks, infer the current attribute data of at least one target object in the industrial production task.

[0058] In the embodiments of the present invention, the calculation module 200 can perform high-speed parallel calculations and has an inference function and a decision-making function, including: an inference unit 2001 and a decision-making unit 2002.

[0059] The inference unit 2001 is used to preprocess the multi-modal sensing data to obtain standardized sensing data, and input the standardized sensing data into a pre-configured inference model to obtain the corresponding current attribute data of at least one target object; the inference model is a pre-trained model that can infer based on the input standardized sensing data and output the current attribute data of at least one target object in the industrial production task.

[0060] In a preferred embodiment, the inference unit 2001 is further used to load a model configuration file from an edge device, and the model configuration file includes the inference model and the data format of the standardized sensing data input to the inference model. Among them, the data format includes: high-precision FP32 format and low-precision INT8 format.

[0061] In the embodiments of the present invention, the inference unit 2001 supports mixed-precision operations, covering FP32, INT8, etc. FP32 provides high-precision calculations and is suitable for critical algorithms with strict precision requirements. INT8 performs operations with lower precision, greatly reducing the amount of calculation and storage requirements, and is suitable for processing a large amount of conventional data. Through mixed-precision operations, the operation precision can be flexibly switched according to different operation scenarios, while ensuring control precision, greatly improving the overall operation performance and meeting the complex requirements of industrial embodied intelligent manufacturing.

[0062] In a preferred embodiment, the inference model is selected from a mechanism model, a neural network model, and a mechanism-neural network model. Among them, the mechanism model is used to perform data conversion on the input standardized sensing data according to the process principle and constraint conditions to generate mechanism constraint data, which is used as the current attribute data of at least one target object; the neural network model is used to perform feature recognition on the input standardized sensing data to generate high-dimensional feature representation data, which is used as the current attribute data of at least one target object; the mechanism-neural network model is used to first perform data screening and conversion on the input standardized sensing data to generate mechanism constraint data, and then perform feature recognition on the mechanism constraint data to obtain high-dimensional feature representation data, which is used as the current attribute data of at least one target object.

[0063] Among them, the inference basis of the mechanism model comes from the industrial production principle, which can accurately analyze the operation rules in the production process and provide theoretical support for the control strategy; the neural network model inference can process complex non-linear data, adaptively adjust the weights through a large amount of data learning, and realize the intelligent prediction and control of the production process; the mechanism-neural network model is suitable for scenarios that involve both industrial production principles and data prediction.

[0064] In the embodiment of the present invention, industrial data and industrial knowledge involved in the industrial production process are pre-stored in the mechanism model. Based on the industrial data and industrial knowledge, the process principle and constraint conditions involved in different production scenarios are determined. Corresponding data operations can be performed according to the process principle, and corresponding data screening can be performed according to the constraint conditions. Industrial data comes from all links of the product life cycle and provides a basis for control decisions; industrial knowledge covers professional content in multiple fields. Industrial data and industrial knowledge together lay a foundation for data insight and digital representation to improve industrial process productivity, efficiency, and innovation.

[0065] Further, when the inference model is configured as a mechanism-neural network model, the mechanism constraint data enters the neural network model for anomaly detection: if an anomaly is detected, strategy re-planning is triggered, the constraint conditions of the mechanism model are adjusted reversely, and the standardized data is re-input into the mechanism model; if the detection is normal, high-dimensional feature representation data is obtained and used as the current attribute data of at least one target object.

[0066] The decision-making unit 2002 is used to generate corresponding control decisions according to the differences between the current attribute data of each target object and the preset target attribute data corresponding to each target object at present.

[0067] In a preferred embodiment, the decision-making unit 2002 is specifically configured to generate a first control decision when the difference between the current attribute data of each target object and the preset target attribute data corresponding to each target object currently is within a preset reasonable difference range; and generate a second control decision when the difference between the current attribute data of each target object and the preset target attribute data corresponding to each target object currently is outside the preset reasonable difference range. One of the first control decision and the second control decision is used to control the intelligent machine device to perform a target operation, and the other is used to control the intelligent machine device not to perform the target operation.

[0068] The embodiment of the present invention is applicable to scenarios such as temperature measurement and sampling tasks. The current attribute data of the target object will naturally tend to the preset target attribute data corresponding to the target object currently. In the temperature measurement and sampling task scenario, the steel block is smelted in the smelting furnace, and its temperature continuously increases and will naturally reach the preset target temperature range; when the temperature is within the preset target temperature range, control the intelligent machine device to perform the target operation to perform molten steel sampling; when the temperature is not within the preset target temperature range, no molten steel sampling is performed.

[0069] In a preferred embodiment, the decision-making unit 2002 is specifically configured to input the current attribute data of each target object and the preset target attribute data corresponding to each target object currently into a pre-configured reinforcement learning model to obtain a corresponding control decision. The control decision can control the intelligent machine device to perform corresponding operations so that each target object generates corresponding changed attribute data. The goal of the reinforcement learning model is to make the current attribute data of each target object plus the corresponding changed attribute data be able to reach the corresponding preset target attribute data as much as possible.

[0070] The embodiment of the present invention is applicable to scenarios such as strip steel pickling tasks. The current attribute data of the target object will not tend to the preset target attribute data corresponding to the target object currently. In the strip steel pickling task scenario, the concentration and temperature of the acid solution will gradually deviate from the preset concentration and temperature as the strip steel is pickled; the reinforcement learning model controls the intelligent machine device to perform corresponding operations so that the concentration and temperature of the acid solution tend to the preset concentration and temperature.

[0071] In the embodiment of the present invention, the flexibility and adaptability of the decision-making unit 2002 directly affect the intelligent level of the controller. According to the changes in the environment and task requirements, the intelligent machine device is controlled in real time to perform corresponding operations. Through the reinforcement learning model, continuously learning and optimizing the decision-making strategy in the actual production process can effectively coordinate and control other modules to ensure the decision-making efficiency to adapt to different production working conditions and environmental changes.

[0072] The control module 300 is configured to convert the control decision into a control signal for functionally controlling the intelligent machine device.

[0073] Furthermore, the control module 300 converts the control decision into the function control of the intelligent machine device, and constructs the function control programs of each function control through programming languages; when the function control programs are running, control signals are generated; the function control includes: logical control, timing control, counting control, analog quantity control, closed-loop control, position control, speed control, sequence control, etc.; the programming languages include: instruction list, structured text, function block diagram, ladder diagram, sequence function diagram, etc. It has obvious advantages in development efficiency, algorithm implementation and resource management, and can cope with complex industrial scenarios.

[0074] Among them, logical, timing, and counting control can implement the basic equipment operation rules and quantity statistics; analog quantity and closed-loop control can accurately adjust industrial parameters; position and speed control ensure the precise movement of the intelligent machine device; sequence control enables the production steps to proceed in an orderly manner. Through diversified control methods, it can fully adapt to different industrial scenarios, significantly improving the production automation and intelligence level.

[0075] The driving module 400 is used to drive the intelligent machine device to perform corresponding operations in industrial production tasks according to the control signal.

[0076] It should be noted that the control signal includes at least one of digital control instructions, sensor analog signals, and servo motor configuration signals. The driving module 400 includes: a digital signal driving unit 4001, an analog signal driving unit 4002, and a servo control unit 4003. Among them, the digital signal driving unit 4001 is used to convert the digital control instruction into a driving signal recognizable by the intelligent machine device; the analog signal driving unit 4002 is used to convert the sensor analog signal into a control input recognizable by the intelligent machine device; the servo control unit 4003 is used to convert the servo motor configuration signal into a servo motor configuration recognizable by the intelligent machine device.

[0077] In the embodiment of the present invention, the driving module 400 is the execution unit of the controller, responsible for receiving the control signal and executing specific tasks. Among them, the digital signal driving unit 4001 has the digital signal driving ability, and can efficiently and accurately convert the digital control instruction issued by the controller into a driving signal recognizable by the intelligent machine device, ensuring the stable operation of the intelligent machine device according to the predetermined logic; the analog signal driving unit 4002 can adapt and amplify various analog signals from the sensor, provide a suitable input for the operation of the intelligent machine device, and ensure the precise control of the analog quantity; the servo control unit 4003 realizes the precise driving of the high-precision moving parts in the intelligent machine device by accurately controlling the rotation speed, torque and position of the servo motor, meets the high requirements for the movement of the intelligent machine device in the complex industrial production process, and thus effectively improves the production precision and efficiency.

[0078] The communication module 500 is used to communicate with the cloud server to receive the model configuration file generated by the cloud server and store the model configuration file at the edge device; the cloud server is used to collect the execution results of intelligent machine devices and generate a new model configuration file according to the execution results; the communication module supports wired, wireless and Internet of Things communications and is compatible with Modbus and OPC UA protocols.

[0079] In the embodiment of the present invention, the communication module 500 supports the edge-cloud collaboration mechanism. The cloud server collects the key data in the execution results of intelligent machine devices to update the model configuration file; the edge device deploys the updated inference model. By adopting the edge-cloud collaboration mechanism, when the environment changes, the controller can be adjusted in real time to generate a new model configuration file. After being segmented, the updated model configuration file is dynamically deployed to the controller and associated edge devices to form a "cloud training-edge inference" closed loop, continuously improving the control accuracy and scene adaptability.

[0080] Preferably, the transfer learning technology can also be used to quickly apply the experience and knowledge learned in one production scenario to similar production scenarios, reducing the time and cost of re-learning.

[0081] In the embodiment of the present invention, the communication module 500 also supports real-time data transmission and information interaction with external systems, ensuring that data can be quickly and accurately transmitted between different devices and systems to achieve collaborative control of the production process. The communication module 500 supports multiple communication protocols, such as Modbus, OPC UA, etc., and can seamlessly connect the controller with devices from different manufacturers, improving the compatibility and scalability of the controller system.

[0082] Furthermore, the communication module 500 also supports distributed interoperability. Different communication nodes can automatically coordinate their work according to preset rules. Even if some communication nodes fail, other communication nodes can still maintain the basic operation of industrial production, greatly enhancing the reliability and fault tolerance of the industrial production system.

[0083] The technical solution of the embodiment of the present invention significantly improves the product quality, greatly improves the production efficiency, improves the human-machine collaboration, reduces the operation difficulty, and enhances the flexibility through the collaborative cooperation of each module inside the controller. The powerful function of the computing module can quickly adapt to environmental changes, optimize device parameters and collaborative work; the cooperation of the sensing module, control module and driving module ensures the fine regulation of the production process; the communication module facilitates the establishment of communication between external devices or systems and the controller system, enhancing the compatibility and scalability of the controller, and strongly promoting the industrial manufacturing towards intelligence and high efficiency, bringing significant economic benefits and competitiveness to the enterprise. Embodiment 2

[0084] Embodiment 2 of the present invention is applicable to a scenario where the current attribute data of the target object will naturally tend to the preset target attribute data corresponding to the target object at present. Combining with the temperature measurement and sampling process in the steel smelting process, Embodiment 2 of the present invention will further introduce the controller in Embodiment 1 of the present invention. The industrial production task is set to measure the temperature and sample the molten steel, the intelligent machine device is selected as the temperature measurement and sampling robot, and the controller in Embodiment 1 of the present invention is used in combination with the temperature measurement and sampling robot, so that the temperature measurement and sampling robot is given powerful embodied intelligence characteristics, thus significantly optimizing the work process.

[0085] Perception module 100: It is used to obtain multi-modal sensing data during the execution of the industrial production task.

[0086] In the temperature measurement and sampling task, the multi-modal sensing data includes: multi-spectral images collected by a vision sensor, three-dimensional attitude angles of the ladle collected by an inclination sensor, floating-point temperature values collected by a temperature sensor, etc.

[0087] The various data collected by the sensors help the temperature measurement and sampling robot to accurately locate and sense the environment. Among them, the vision sensor can accurately identify the target positions such as the molten steel ladle and the casting blank in the high-temperature and dusty steel production environment to ensure that the temperature measurement and sampling probe reaches accurately; the inclination sensor obtains the three-dimensional attitude angle of the ladle in real time; the temperature sensor monitors the ambient temperature and its own equipment temperature in real time to ensure the stable operation of the temperature measurement and sampling robot.

[0088] Inference unit 2001: It is used to preprocess the multi-modal sensing data to obtain standardized sensing data, and input the standardized sensing data into a pre-configured inference model to obtain the current attribute data of at least one corresponding target object; the inference model is a pre-trained model that can perform inference based on the input standardized sensing data and output the current attribute data of at least one target object in the industrial production task.

[0089] In the embodiment of the present invention, the inference model selects a mechanism-neural network model, which is used to first perform high-precision FP32 operations on the input floating-point temperature value and the three-dimensional attitude angle of the ladle, and perform data conversion according to the thermal conductivity coefficient corresponding to the steel grade in the process principle and the target temperature range of the molten steel in the constraint conditions to obtain the spatial coordinates of the safe temperature measurement area and the temperature field prediction matrix; then, perform low-precision INT8 operations on the spatial coordinates of the safe temperature measurement area, the temperature field prediction matrix, and the multi-spectral image, and obtain the temperature of the current optimal temperature measurement point of the molten steel by identifying the dynamic characteristics of the molten pool surface.

[0090] In the embodiment of the present invention, the mechanism model part is based on the physical model of molten steel heat conduction, and combines the boundary conditions of the temperature field distribution to provide reliable constraints for the safe temperature measurement area. The floating-point temperature value of the temperature sensor and the three-dimensional ladle attitude angle of the inclination sensor are input into the mechanism model. According to the thermal conductivity corresponding to the steel grade code (Q235B) and the target temperature range of the molten steel (1550 - 1580 °C), the spatial coordinates of the safe temperature measurement area and the temperature field prediction matrix (temperature distribution map of a 128×128 grid) are output.

[0091] The neural network model part can select the ResNet-18 model. The multi-spectral image, the temperature field prediction matrix, and the spatial coordinates of the safe temperature measurement area are input into the model to identify the dynamic characteristics of the molten pool surface, such as the thickness of the slag layer, the position of the bubbles, etc., to achieve precise point selection, and output the temperature and confidence score of the current optimal temperature measurement point of the molten steel.

[0092] In addition, after the spatial coordinates of the safe temperature measurement area and the temperature field prediction matrix enter the neural network model, anomaly detection is performed: if the spatial coordinates of the safe temperature measurement area and the temperature field prediction matrix do not meet the safety range preset by the neural network model, the constraint conditions (the target temperature range of the molten steel) of the mechanism model are adjusted; if they meet, feature recognition is performed on the multi-spectral image, the spatial coordinates of the safe temperature measurement area, and the temperature field prediction matrix to obtain the temperature and coordinates of the optimal temperature measurement point. The spatial coordinates of the safe temperature measurement area and the temperature field prediction matrix output by the mechanism model limit the search space of ResNet-18 to prevent the probe from contacting the furnace wall or the slag layer.

[0093] In the embodiment of the present invention, the inference unit 2001 can quickly infer the multi-modal sensing data transmitted from the sensing module 100 based on the mechanism-neural network model and the model trained with a large amount of steel production data. For example: the quality recognition model infers information such as the temperature range and composition distribution of the molten steel based on the color, flow state, and temperature change trend of the molten steel, and anticipates possible quality problems in advance.

[0094] The decision-making unit 2002 is used to generate corresponding control decisions according to the current attribute data of each target object and the preset target attribute data corresponding to each target object at present.

[0095] The decision-making unit 2002 is specifically configured to generate a first control decision when the difference between the current attribute data of each target object and the preset target attribute data corresponding to each target object currently is within the preset reasonable difference range; and, when the difference between the current attribute data of each target object and the preset target attribute data corresponding to each target object currently is outside the preset reasonable difference range, generate a second control decision; wherein, one of the first control decision and the second control decision is used to control the intelligent machine device to perform the target operation, and the other of the two is used to control the intelligent machine device not to perform the target operation.

[0096] In the embodiment of the present invention, the decision-making unit 2002 generates a control decision according to the difference between the temperature of the current optimal temperature measurement point of the molten steel and the preset temperature measurement and sampling temperature of the molten steel, and controls the temperature measurement and sampling robot to perform or not perform the target operation. For example: according to the smelting stage and process requirements of the molten steel, the appropriate temperature measurement time point, sampling depth and position are used as the target operation.

[0097] The control module 300: is used to convert the control decision into a control signal for functionally controlling the intelligent machine device.

[0098] In the embodiment of the present invention, the control module 300 performs complex logic operations and sequential control on the actions of the temperature measurement and sampling robot by using programming languages such as ladder diagrams and statement lists to ensure precise control and stable operation. The control module 300 supports the implementation of function controls such as logic, timing, counting, analog quantity, closed-loop, position, speed, and sequence. By setting a high-precision timer, the time interval of temperature measurement and sampling is precisely controlled to ensure the continuity and accuracy of data; by accurately counting the pulse signal, specific actions of the temperature measurement and sampling robot are triggered, such as the extension and retraction of the probe, etc.; through the analog quantity control ability, signals such as temperature and pressure are collected and processed, and the decision-making instructions of the calculation module 200 are quickly converted into logic control program instructions of the temperature measurement and sampling robot to achieve precise control of the temperature measurement and sampling robot. At the same time, an efficient closed-loop control system is constructed to feedback the operation state of the equipment in real time to ensure the stable operation of the temperature measurement and sampling robot.

[0099] The driving module 400: is used to drive the intelligent machine device to perform corresponding operations in the industrial production task according to the control signal.

[0100] In the embodiment of the present invention, the driving module 400 is used as an execution unit to ensure that the robot can stably and accurately complete the temperature measurement and sampling actions in the steel production environment with high temperature and strong magnetic field, improving the accuracy and efficiency of the work.

[0101] Among them, the digital signal driving unit 4001 efficiently converts the digital control instructions issued by the controller into driving signals recognizable by the temperature measurement and sampling robot; the analog signal driving unit 4002 adapts and amplifies the analog signals to provide support for precise control; the servo control unit 4003 precisely drives the joints and other high-precision moving parts of the temperature measurement and sampling robot by precisely controlling the rotation speed, torque, and position of the servo motor.

[0102] Communication module 500: It is used to communicate with the cloud server to receive the model configuration file generated by the cloud server and store the model configuration file at the edge device; the cloud server is used to collect the execution results of intelligent machine devices and generate a new model configuration file according to the execution results.

[0103] In the embodiment of the present invention, the communication module 500 adopts industrial Ethernet or 5G technology to realize real-time data transmission and information interaction between the temperature measurement and sampling robot and other devices in the steelmaking workshop and the upper monitoring system; quickly transmit the temperature measurement and sampling data to the quality control system to timely adjust the production process; at the same time, receive the instructions and production plan change information of the upper monitoring system and flexibly adjust the work tasks.

[0104] Through the collaborative work of the above-mentioned modules in the technical solution of the embodiment of the present invention, the temperature measurement and sampling robot, empowered by the controller, has the characteristics of embodied intelligence and can autonomously, efficiently, and precisely complete the temperature measurement and sampling work in a complex steel production environment, providing a strong guarantee for the improvement of steel production quality. Embodiment Three

[0105] Embodiment Three of the present invention is applicable when the current attribute data of the target object does not tend to the preset target attribute data corresponding to the target object at present. Combining the cold rolling production process, Embodiment Three of the present invention will further introduce the controller in the embodiment of the present invention.

[0106] Figure 2 It is a schematic flow chart of a cold rolling process applicable to the embodiment of the present invention. As Figure 2 shown, Embodiment Three of the present invention divides the cold rolling production into four process stages.

[0107] Sensing module 100: It is used to obtain multi-modal sensing data during the execution of industrial production tasks.

[0108] Inference unit 2001 is used to preprocess the multi-modal sensing data to obtain standardized sensing data, and input the standardized sensing data into a pre-configured inference model to obtain the current attribute data of at least one corresponding target object; the inference model is a pre-trained model that can perform inference based on the input standardized sensing data and output the current attribute data of at least one target object in the industrial production task.

[0109] The first process stage: pickling process.

[0110] Sensing module 100: In the pickling process, different types of sensors are deployed in the pickling tank body and its surrounding areas. For example, a pH sensor is used to monitor the acid concentration in real time, and a temperature sensor is used to control the acid temperature; a high-definition vision sensor is used to check the residual scale on the strip surface, accurately perceiving the real-time state of the pickling process.

[0111] Inference unit 2001: Select a mechanism model with high-precision operation, which has a small amount of calculation and low latency. The inputs of the mechanism model include: strip running parameters, acid solution state parameters, etc. Among them, the strip running parameters include: real-time strip speed, strip thickness, material (Q235B), surface scale thickness, strip temperature distribution (obtained through an infrared sensor), etc.; the acid solution state parameters include: acid concentration (pH value, Fe²⁺ ion concentration), acid temperature, etc. The process principle and constraint conditions include: equipment and process indicators, such as: loop stock volume, speed limits of each section of the unit, etc. The model outputs include: the consumption rate of the acid solution, the pickling effect of the strip, etc., as the current attribute data of the target object in the first process stage.

[0112] The second process stage: rolling process.

[0113] Sensing module 100: In the rolling process, pressure sensors, displacement sensors, speed sensors, and high-precision flatness detectors are installed at key parts of the rolling mill to monitor the rolling force, the position and speed of the rolls, and the changes in the flatness and thickness of the strip in real time, comprehensively grasping the real-time state of the rolling process.

[0114] Inference unit 2001: Select a mechanism-neural network model, with high precision for the mechanism model part and low precision for the neural network model part.

[0115] In the mechanism model, the Stone rolling force model can be used. The inputs of the model include: real-time sensor data, such as: rolling force, roll gap, and strip inlet / outlet thickness, etc.; the settings of the process principle and constraint conditions include: strip yield strength, friction coefficient; the model outputs include: the theoretical rolling force range, roll gap adjustment range.

[0116] When the output data of the mechanism model is input into the neural network model, if the actual flatness quality deviation is greater than 5% or the rolling force exceeds the theoretical range, the friction coefficient in the mechanism model is updated in reverse, and the ResNet-50 flatness detection model simultaneously fine-tunes the weights of the network.

[0117] In the neural network model, a combination of a ResNet-50 shape detection model and a reinforcement learning model can be adopted. In the ResNet-50 shape detection model, the model inputs include: shape detector images, strip thickness distribution matrices, etc.; the model outputs include: shape defect classification codes, defect position heat maps, etc. In the reinforcement learning model, the model inputs include: current rolling parameters, shape quality scores, equipment status (roll temperature, vibration amplitude), planned production volume, etc. The model outputs include: the deformation law of the strip, the wear condition of the rolls, and the stability during the roll rolling process, as the current attribute data of the target object in the second process stage.

[0118] The third process stage: annealing process.

[0119] Perception module 100: In the annealing process, temperature sensors and atmosphere sensors are installed inside the annealing furnace, and strip tension sensors and position sensors are set outside the furnace. The temperature distribution and gas composition inside the furnace, as well as the tension and position changes of the strip during annealing, are monitored in real time to comprehensively grasp the information inside the annealing furnace.

[0120] Inference unit 2001: A neural network model is selected. The inputs of the neural network model include: current temperature curve (heating rate, holding time), strip performance indicators (yield strength, elongation), energy consumption data (gas consumption, power consumption), delivery cycle constraint (total annealing time ≤ 120 minutes), quality requirements (hardness HRB ≤ 75), etc.; the output of the model is: the heat exchange situation inside the annealing furnace, the microstructure transformation process of the strip, and possible defects, as the current attribute data of the target object in the third process stage.

[0121] The fourth process stage: finishing process.

[0122] Perception module 100: In the finishing process, surface defect detectors, dimensional gauges, and flatness gauges are installed on the finishing equipment. The surface quality, dimensional accuracy, and flatness of the strip are detected in real time to comprehensively grasp the quality status of the strip during the finishing process.

[0123] Inference unit 2001: A neural network model with low-precision operations is selected. The neural network model can be a combination of the YOLOv8 surface defect detection model and the Vision Transformer quality assessment model.

[0124] In the YOLOv8 surface defect detection model, the model inputs include: spectral surface images, strip motion parameters (speed, acceleration), etc.; the model outputs include: defect types and coordinates, etc.

[0125] In the Vision Transformer quality assessment model, the inputs of the model include: surface defect heat maps, thickness / width time-series curves, and quality standards; the outputs of the model include: the types of strip surface defects, the causes of generation, and the trend of dimensional deviation, as the current attribute data of the target objects in the fourth process stage.

[0126] The decision-making unit 2002 is used to generate corresponding control decisions according to the differences between the current attribute data of each target object and the preset target attribute data corresponding to each target object at present.

[0127] Specifically, the decision-making unit 2002 is used to input the current attribute data of each target object and the preset target attribute data corresponding to each target object at present into a pre-configured reinforcement learning model to obtain corresponding control decisions; the control decisions can control intelligent machine equipment to perform corresponding operations so that each target object generates corresponding changed attribute data, and the goal of the reinforcement learning model is to make the current attribute data of each target object plus the corresponding changed attribute data can reach the corresponding preset target attribute data as much as possible.

[0128] In the embodiment of the present invention, the decision-making unit 2002 performs global production rhythm scheduling through a reinforcement learning model, including: accurately matching the pickling feeding rate and the rolling consumption rate, adjusting the acid solution pH, rolling force, annealing temperature, and finishing shearing amount to eliminate quality deviations, etc., and constructing a process-equipment-quality association knowledge graph. In addition, the reinforcement learning model can also perform reverse parameter correction on the inference models of each process stage. For example: when the finishing quality score does not meet the standard, perform reverse detection on the finishing process → annealing process → rolling process → pickling process, and when it is detected that the yield strength of the steel block is insufficient, adjust the constraint conditions of each process stage, for example: increase the annealing holding time, reduce the rolling speed to improve the deformation uniformity, and adjust the pickling temperature.

[0129] Specifically, the decision-making unit 2002 can generate control decisions: determine whether acid solution needs to be supplemented, adjust the acid solution temperature, or change the traveling speed of the strip in the first process stage; during long-term operation, optimize the acid solution addition and temperature control to meet the pickling requirements of different batches of strips. Develop the optimal rolling force, roll speed, and roll gap adjustment plan in the second process stage; improve the rolling quality and efficiency when facing rolling tasks of different specifications and materials, continuously improve the rolling process model, and achieve precise control of the rolling process. Develop the best heating curve, holding time, and furnace atmosphere control plan in the third process stage; improve the annealing quality and performance of the strip through continuous analysis of annealing data. Develop the best finishing process plan in the fourth process stage, such as shearing length, straightening force, etc. When facing strips with different specifications and quality requirements, improve the finishing quality and production efficiency, continuously improve the finishing process model, and achieve precise control of the strip quality.

[0130] Control module 300: It is used to convert control decisions into control signals for functionally controlling intelligent machine devices.

[0131] In the embodiment of the present invention, the control module 300 precisely controls the opening and closing of the acid solution addition pump, the power adjustment of the heating device, and the rotational speed of the strip conveyor motor in the first process stage. By setting a high-precision timer, the time accuracy of each link operation is ensured. Using the analog control ability, analog signals such as acid solution concentration and temperature are collected and processed in real time to achieve closed-loop control of the pickling process. In the second process stage: Using programming languages such as ladder diagrams and statement lists to precisely control each actuator of the rolling mill. By precisely counting the pulse signals, precise adjustment of the roll position is achieved. Using the analog control ability, analog signals such as rolling force and speed are collected and processed in real time to construct an efficient closed-loop control system to ensure the stable operation of the rolling mill and the rolling quality of the strip meets the standards. In the third process stage: Precisely control the heating elements, gas flow regulating valves, strip tension regulating devices, etc. of the annealing furnace. Using complex logical operations and sequential control to ensure that each device operates according to the predetermined annealing process curve. By setting a high-precision timer, precisely control the heating and heat preservation time. Using the analog control ability, analog signals such as temperature and gas flow are collected and processed in real time to achieve closed-loop control of the annealing process. In the fourth process stage: Precisely control the actuators such as the shearing machine and straightening machine of the finishing equipment. Using programming languages such as ladder diagrams and statement lists to achieve logical control and sequential control of the equipment. By precisely counting the pulse signals, ensure the accuracy of the shearing length. Using the analog control ability, analog signals such as straightening force are collected and processed in real time to construct an efficient closed-loop control system to ensure the finishing quality of the strip.

[0132] Drive module 400: It is used to drive the intelligent machine device to perform corresponding operations in industrial production tasks according to the control signal.

[0133] In the technical solution of the embodiment of the present invention, by distributing controllers in each process section of the cold rolling production line, each process section has the ability of autonomous perception, intelligent decision-making, and precise control. And the intelligent controllers between each process section conduct real-time data interaction and collaborative operation through the communication module, combined with the unified scheduling of the upper monitoring system, making the entire cold rolling production line form an organic intelligent whole, realizing a high degree of intelligent transformation, and greatly improving production efficiency, product quality, and the competitiveness of the enterprise. Embodiment 4

[0134] Figure 3 It is a schematic flowchart of a control method for a controller used in industrial embodied intelligent manufacturing according to an embodiment of the present invention, as Figure 3As shown, the control method includes:

[0135] S1. Obtain multimodal sensing data during the execution of industrial production tasks.

[0136] S2. Infer the current attribute data of at least one target object in the industrial production task based on the multimodal sensing data, and generate corresponding control decisions according to the current attribute data of each target object and the preset target attribute data corresponding to each target object at present.

[0137] S3. Convert the control decision into a control signal for functionally controlling the intelligent machine device.

[0138] S4. Drive the intelligent machine device to perform corresponding operations in the industrial production task according to the control signal.

[0139] Preferably, it further includes:

[0140] S5. Communicate with the cloud server to receive the model configuration file generated by the cloud server, and store the model configuration file at the edge device; the cloud server is used to collect the execution results of the intelligent machine device and generate a new model configuration file according to the execution results.

[0141] The technical solution in the embodiment of the present invention proposes a control method for a controller for industrial embodied intelligent manufacturing. Based on a distributed edge intelligent architecture, it deeply integrates the closed-loop collaborative mechanism of perception, computing, control, and execution to achieve dynamic optimization and real-time response of industrial embodied intelligent manufacturing. Embodiment Five

[0142] Figure 4 The structural schematic diagram of the electronic device 10 that can be used to implement the embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0143] As Figure 4As shown, the electronic device 10 includes at least one processor 11 and a memory communicatively connected to the at least one processor 11, such as read-only memory (ROM) 12, random access memory (RAM) 13, etc. Among them, the memory stores a computer program executable by the at least one processor. The processor 11 can execute various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 into the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, ROM 12, and RAM 13 are connected to each other via a bus 14. The input / output (I / O) interface 15 is also connected to the bus 14.

[0144] Multiple components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disc, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0145] The processor 11 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the control method of a controller for industrial embodied intelligent manufacturing.

[0146] In some embodiments, the control method of a controller for industrial embodied intelligent manufacturing can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as the storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the control method of the controller for industrial embodied intelligent manufacturing described above can be executed. Alternatively, in other embodiments, the processor 11 can be configured to execute the control method of the controller for industrial embodied intelligent manufacturing by any other appropriate means (for example, by means of firmware).

[0147] The various embodiments of the systems and techniques described above in this document can be implemented in digital electronic circuitry, integrated circuit systems, field programmable gate arrays (FPGA), application specific integrated circuits (ASIC), application specific standard products (ASSP), systems on a chip (SOC), complex programmable logic devices (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special-purpose or general-purpose programmable processor that receives data and instructions from a storage system, at least one input device, and at least one output device, and transmits the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0148] The computer programs for implementing the methods of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, a special purpose computer, or other programmable data processing device, such that the computer programs, when executed by the processor, cause the functions / operations specified in the flowchart and / or block diagram to be implemented. The computer programs can be executed entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine, or entirely on the remote machine or server.

[0149] In the context of the present invention, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer-readable storage medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, the computer-readable storage medium can be a machine-readable signal medium. More specific examples of the machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0150] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the electronic device. Other kinds of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0151] The systems and techniques described herein can be implemented in a computing system including backend components (e.g., as a data server), or a computing system including middleware components (e.g., an application server), or a computing system including frontend components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system including any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected to each other by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local area network (LAN), wide area network (WAN), blockchain network, and the Internet.

[0152] The computing system can include a client and a server. The client and the server are generally far from each other and typically interact through a communication network. The client - server relationship is created by computer programs running on the respective computers and having a client - server relationship with each other. The server can be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system and solves the defects of difficult management and weak business scalability existing in traditional physical hosts and VPS services.

[0153] It should be understood that various forms of the processes shown above can be used, with steps reordered, added, or deleted. For example, the steps recited in the present invention can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved, and no limitation is imposed herein.

[0154] The above - mentioned specific implementation manners do not constitute a limitation on the protection scope of the present invention. Those skilled in the art should understand that various modifications, combinations, sub - combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A controller for industrial embodied intelligent manufacturing, characterized in that, Including: A perception module, configured to obtain multi-modal sensing data during the execution of an industrial production task. The industrial production task includes measuring the temperature and sampling molten steel. The multi-modal sensing data includes: multi-spectral images collected by a vision sensor, three-dimensional attitude angles of a ladle collected by an inclination sensor, and floating-point temperature values collected by a temperature sensor; A calculation module, configured to infer the current attribute data of at least one target object in the industrial production task based on the multi-modal sensing data, and generate corresponding control decisions according to the current attribute data of each target object and the preset target attribute data corresponding to each target object at present. The at least one target object includes: a molten steel ladle and a billet. The current attribute data includes: the temperature of the current optimal temperature measurement point of the molten steel. The preset target attribute data includes: a preset temperature measurement and sampling temperature; The calculation module includes: an inference unit. The inference unit is configured to preprocess the multi-modal sensing data to obtain standardized sensing data, and input the standardized sensing data into a pre-configured inference model to obtain the current attribute data of the corresponding at least one target object. The inference model includes: a mechanism-neural network model. The mechanism-neural network model includes: a mechanism model part and a neural network model part. The mechanism model part is configured to output the spatial coordinates of a safe temperature measurement area and a temperature field prediction matrix according to the floating-point temperature value of the input temperature sensor and the three-dimensional attitude angles of the ladle of the inclination sensor. The neural network model part selects a ResNet-18 model and is configured to output the temperature of the current optimal temperature measurement point of the molten steel according to the input multi-spectral images, temperature field prediction matrix, and spatial coordinates of the safe temperature measurement area; A control module, configured to convert the control decision into a control signal for functionally controlling an intelligent machine device. The intelligent machine device includes: a temperature measurement and sampling robot; A driving module, configured to drive the intelligent machine device to perform corresponding operations during the execution of the industrial production task according to the control signal.

2. The controller according to claim 1, characterized in that The calculation module further includes: a decision-making unit. The decision-making unit is configured to generate corresponding control decisions according to the differences between the current attribute data of each target object and the preset target attribute data corresponding to each target object at present.

3. The controller according to claim 1, wherein The inference unit is further configured to load a model configuration file from an edge device. The model configuration file includes the inference model and the data format of the standardized sensing data input to the inference model; Wherein, the data format includes: a high-precision FP32 format and a low-precision INT8 format.

4. The controller according to claim 2, wherein The decision-making unit is specifically configured to generate a first control decision when the differences between the current attribute data of each target object and the preset target attribute data corresponding to each target object at present are within a preset reasonable difference range; And generate a second control decision when the differences between the current attribute data of each target object and the preset target attribute data corresponding to each target object at present are outside the preset reasonable difference range. One of the first control decision and the second control decision is used to control the intelligent machine device to perform a target operation, and the other is used to control the intelligent machine device not to perform the target operation.

5. The controller according to claim 2, characterized in that, The decision-making unit is specifically configured to input the current attribute data of each target object and the preset target attribute data corresponding to each target object currently into a pre-configured reinforcement learning model to obtain a corresponding control decision. The control decision can control the intelligent machine device to perform corresponding operations so that each target object generates corresponding changed attribute data. The goal of the reinforcement learning model is to make the current attribute data of each target object plus the corresponding changed attribute data reach the corresponding preset target attribute data as much as possible.

6. The controller according to claim 2, wherein The inference model is selected from one of a mechanism model, a neural network model, and a mechanism-neural network model. The mechanism model is used to perform data conversion on the input standardized sensing data according to process principles and constraint conditions to generate mechanism constraint data, and use it as the current attribute data of at least one target object. The neural network model is used to perform feature recognition on the input standardized sensing data to generate high-dimensional feature representation data, and use it as the current attribute data of at least one target object. The mechanism-neural network model is used to first perform data screening and conversion on the input standardized sensing data to generate mechanism constraint data; then perform feature recognition on the mechanism constraint data to obtain high-dimensional feature representation data, and use it as the current attribute data of at least one target object.

7. The controller according to claim 1, characterized in that, The control module is specifically configured to convert the control decision into a function control of the intelligent machine device, and construct a function control program for each function control through a programming language; when the function control program runs, a control signal is generated. The function control includes at least one of logical control, timing control, counting control, analog quantity control, closed-loop control, position control, speed control, and sequence control. The programming language includes at least one of instruction list, structured text, function block diagram, ladder diagram, and sequential function chart.

8. The controller according to claim 1 or 6, wherein The control signal includes at least one of a digital control instruction, a sensor analog signal, and a servo motor configuration signal. The drive module includes a digital signal drive unit, an analog signal drive unit, and a servo control unit. The digital signal drive unit is used to convert a digital control instruction into a drive signal recognizable by the intelligent machine device. The analog signal drive unit is used to convert a sensor analog signal into a control input recognizable by the intelligent machine device. The servo control unit is used to convert a servo motor configuration signal into a servo motor configuration recognizable by the intelligent machine device.

9. The controller according to claim 1, wherein It further includes: A communication module, which is used to communicate with a cloud server to receive a model configuration file generated by the cloud server and store the model configuration file at the edge device. The cloud server is used to collect the execution results of the intelligent machine device and generate a new model configuration file according to the execution results. The communication module supports wired, wireless, and Internet of Things communications and is compatible with Modbus and OPC UA protocols.

10. A control method for a controller used in industrial embodied intelligent manufacturing, which is applied to the controller for industrial embodied intelligent manufacturing according to any one of claims 1-9, characterized in that, The method includes: S1. Obtain multi-modal sensing data during the execution of an industrial production task, where the industrial production task includes measuring the temperature and taking samples of molten steel. The multi-modal sensing data includes multi-spectral images collected by a vision sensor, three-dimensional attitude angles of a ladle collected by an inclination sensor, and floating-point temperature values collected by a temperature sensor; S2. Infer the current attribute data of at least one target object in the industrial production task based on the multi-modal sensing data, and generate corresponding control decisions according to the current attribute data of each target object and the preset target attribute data corresponding to each target object at present. The at least one target object includes a ladle and a slab. The current attribute data includes the temperature of the current optimal temperature measurement point of the molten steel, and the preset target attribute data includes the preset temperature measurement and sampling temperature; The process of inferring the current attribute data of at least one target object in step S2 specifically includes: preprocessing the multi-modal sensing data to obtain standardized sensing data, and inputting the standardized sensing data into a pre-configured inference model to obtain the corresponding current attribute data of at least one target object. The inference model includes a mechanism-neural network model, and the mechanism-neural network model includes a mechanism model part and a neural network model part. The mechanism model part is used to output the spatial coordinates of a safe temperature measurement area and a temperature field prediction matrix according to the input floating-point temperature value of the temperature sensor and the three-dimensional attitude angle of the ladle of the inclination sensor. The neural network model part selects a ResNet-18 model and is used to output the temperature of the current optimal temperature measurement point of the molten steel according to the input multi-spectral image, temperature field prediction matrix, and spatial coordinates of the safe temperature measurement area; S3. Convert the control decision into a control signal for functionally controlling an intelligent machine device, where the intelligent machine device includes a temperature measurement and sampling robot; S4. Drive the intelligent machine device to perform corresponding operations in the industrial production task according to the control signal.

Images