A collaborative method for industrial IoT production and industrial chains based on multi-agents

Abstract

This invention provides a multi-agent-based industrial Internet of Things (IIoT) production chain and industrial chain collaboration method, relating to the field of industrial Internet technology. Under constraints such as elastic supply and dynamic demand in the external market, dynamic production status within enterprises (equipment capacity, process parameters), external resource consumption, and carbon emissions, the method aims to improve the comprehensive evaluation indicators of the entire production process, including product output, quality, energy consumption, emissions, and cost. It employs a virtual manufacturing process to achieve feedforward decision correction based on virtual simulation, and uses large-scale sensing data from industrial production for working condition identification and feedback to achieve self-optimizing decision-making and dynamic optimization decision-making through human-machine interaction, thereby establishing a multi-agent-based industrial chain and production chain collaboration mechanism.

Description

Technical Field

[0001] This invention belongs to the field of industrial internet technology, specifically relating to a collaborative method for industrial IoT production chains and industrial chains based on multi-agent systems. Background Technology

[0002] The steel industry's production and supply chains generate massive amounts of multi-source, heterogeneous data (both existing and future), requiring extensive optimization and control decisions. Therefore, it is necessary to build an intelligent system for steel enterprises, with a smart factory brain comprised of a data center, business center, and algorithm center at its core. This system should improve intelligent management across the entire process and supply chain, from suppliers and production bases to manufacturing lines and customers, through supply chain collaboration, value chain extension, and ecological integration. This will enable steel enterprises to achieve intelligent production, networked collaboration, personalized customization, and service extension, thereby optimizing operational systems, leveraging resource advantages, innovating production models, and enhancing the overall competitiveness and advantages of steel enterprises.

[0003] Chinese patent "CN202310285877.9 A Multi-Agent Collaboration Method Based on Hierarchical Communication Mechanism" provides a multi-agent collaboration method based on a hierarchical communication mechanism. First, an intra-group communication topology is established, then an inter-group communication topology is established, followed by intra-group information aggregation and inter-group information exchange, and finally, action decision-making is performed. By establishing a reasonable and efficient hierarchical communication structure among agents, the centralized data acquisition, aggregation, and utilization are broken, alleviating the curse of dimensionality problem of centralized training.

[0004] Chinese patent "CN202310895639.X Multi-agent Cooperative Computing Resource Scheduling Method, Device and System" discloses a multi-agent cooperative computing resource scheduling method. First, a task request queue is configured, and the historical running state sequence of the edge cluster is obtained and input into a multi-layered LSTM neural network. The output is a system state matrix with temporal characteristics. Then, the state matrix is ​​input into an Actor-Critic network configured in the edge cluster for multi-agent reinforcement learning to complete resource scheduling. Finally, the Actor-Critic network parameters are updated based on the task rewards, calculating the loss function and gradient. This method successfully learns the periodic states between requests when facing large-scale service requests, improving system throughput. The combination of temporal networks and policies for each task scheduling results in faster convergence and requires less training data.

[0005] The above describes the traditional multi-agent collaboration mechanism. Although it can achieve collaboration among agents at the same level, it cannot meet the needs of multi-level collaboration in the steel industry's production and industrial chains, as well as the steel industry's requirements for high-end, intelligent, and green development. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention designs an industrial IoT production chain and supply chain collaboration method based on multi-agent intelligence. Based on three layers of intelligence, the production equipment intelligence agent adaptively makes optimization decisions, optimizes resource allocation, and rationally allocates and recycles energy, forming the bottom-level intelligence agent. The production chain-level intelligence agent collaboratively controls the production equipment intelligence agent, achieving optimized allocation of process resources and worry-free production in the workshop. The supply chain-level intelligence agent collaborates with the production chain-level intelligence agent to achieve global optimization of the manufacturing process.

[0007] A multi-agent industrial IoT production chain and industrial chain collaboration method consists of three parts: production equipment intelligent agent, production chain-level collaborative intelligent agent, and industrial chain-level resource allocation intelligent agent.

[0008] The production equipment intelligent agents correspond to production equipment and include blast furnace agents, converter agents, and rolling mill agents, each completing its respective production target. The production chain-level intelligent agents include workshop production planning agents and workshop operation planning agents. On the one hand, they generate operation plans for each process of production equipment and control the production equipment intelligent agents to realize the production of the underlying production equipment. On the other hand, they obtain the company's production plan from the supply chain-level intelligent agents and are controlled by the supply chain-level intelligent agents. The supply chain-level intelligent agents include material procurement planning agents, product marketing planning agents, and manufacturing resource allocation planning agents, which optimize the overall production process of the enterprise, determine the production plan for each process, and guide each process to complete production operations.

[0009] The production equipment intelligent agent, the production chain-level collaborative intelligent agent, and the industry chain-level resource allocation intelligent agent communicate and make decisions using the Contract Net Protocol (CNP) mechanism. The Contract Net consists of n nodes, divided into two types: managers and agents. In the Contract Net, managers and agents exchange dynamic messages through negotiation. The decision-making process of the Contract Net includes the following four basic steps:

[0010] Step 1: The manager posts the task for bidding;

[0011] Step 2: The agent submits its bidding information to the manager, including production capacity and production costs;

[0012] Step 3: Managers select an agent to perform the task based on the principle of maximizing the company's profits;

[0013] Step 4: Finally, the manager confirms the contract, thus completing the bidding and tendering decision-making process;

[0014] The functions of the intelligent production equipment agent are as follows: acquiring production data information of each piece of equipment through sensors, including: production efficiency, production benefits, product quality, production equipment operating status, environmental changes, material types and quality, energy resources, human resources, and equipment resources; intelligently sensing production operating condition information, including: material input, energy consumption, and carbon emissions; and adaptively making optimization decisions, optimizing resource allocation, and rationally allocating and recycling energy using neural network models; the intelligent production equipment agents adopt a communication and information sharing strategy, and effective communication and information sharing are required between the agents to understand each other's status, intentions, and action plans. Communication is achieved through direct message transmission, sharing of knowledge bases, or sensor data.

[0015] The functions of the blast furnace agent, converter agent, and rolling mill agent are as follows:

[0016] (1) Blast Furnace Agent: Each blast furnace agent bids for its own production tasks from the workshop operation plan agent, while simultaneously sensing production data and production conditions;

[0017] (2) Converter Agent: Each converter agent bids for its own production tasks from the workshop operation plan agent, and at the same time senses its own production data and production conditions;

[0018] (3) Rolling mill agent: Each rolling mill agent bids for its own production tasks from the workshop operation plan agent, and at the same time senses its own production data and production conditions;

[0019] The production chain-level intelligent agent receives order production tasks and makes autonomous decisions using the contract network, formulating cooperation strategies. These decisions and strategies involve task allocation and require comprehensive consideration of overall goals and individual interests. Production at each process stage is coordinated and controlled by workshop production planning agents and workshop operation planning agents for ironmaking, steelmaking, and rolling processes, ensuring smooth workshop production. The workshop production planning agent formulates workshop production plans based on the production capacity of each process transmitted through the JADE platform and the enterprise's production model. The workshop operation planning agent fully utilizes the production capacity of each process, rationally organizes production, ensures the schedulability of resources for each process, and rationally arranges operations for each process, achieving coordinated matching of processes and guaranteeing the smooth operation of the entire workshop.

[0020] The workshop production planning agent and workshop operation planning agent cooperate to achieve the global goal, and their functions are as follows:

[0021] (1) Workshop Production Planning Agent: Obtains the workshop production plan by bidding to the Manufacturing Resource Allocation Planning Agent using the Contract Network Protocol;

[0022] (2) Workshop Operation Planning Agent: Bid for each production equipment agent and select a production equipment agent to complete the workshop operation plan;

[0023] The industry chain-level intelligent agent achieves the overall goal by coordinating the production of each process. It coordinates upstream and downstream enterprises to complete material procurement and product marketing plans through material procurement planning agents, product marketing planning agents, and manufacturing resource allocation planning agents. At the same time, it prioritizes the orders received by enterprises to improve the efficiency of the entire enterprise.

[0024] The Material Procurement Planning Agent, Product Marketing Planning Agent, and Manufacturing Resource Allocation Planning Agent have the following functions:

[0025] (1) Material Procurement Planning Agent: The company's production depends on the materials of the upstream companies; therefore, the company bids for the upstream companies according to the material procurement planning agent to complete the procurement of the required materials.

[0026] (2) Product Marketing Plan Agent: The production of downstream enterprises depends on the products of enterprises. Therefore, enterprises tender to downstream enterprises according to the product marketing plan agent to complete the enterprise's product marketing plan;

[0027] (3) Manufacturing Resource Allocation Planning Agent: Responsible for generating production plans based on the production orders received by the enterprise, and issuing the orders generated based on the production plans to the production planning agents at the production chain level workshop, and completing the bidding and tendering with the production chain level intelligent agents;

[0028] The production chain and industry chain collaboration method is implemented using the Java JADE framework. The basic functions of the JADE framework include an Agent Management System (AMS) responsible for monitoring and managing platform access and usage, a Directory Service Agent (DF) providing default yellow pages, and a message transmission system. The AMS monitors and controls the Agent platform; each agent must register with the AMS to obtain a valid AID. The DF provides default yellow pages services to the agents. The message transmission system is a software component within the platform that controls all message exchanges. Agents interact and communicate through these functions, while a MySQL database is used for information storage and retrieval. Production equipment agents, production chain-level agents, and industry chain-level agents are created using the basic functions of the JADE platform. Information exchange between different agents is completed through the MySQL database and the message transmission system in the JADE framework. Agent identification is achieved through the AID in the JADE framework.

[0029] A multi-agent-based method for collaborative production and supply chains in the Industrial Internet of Things (IIoT) also includes the following:

[0030] Establishing a production model for the enterprise provides a basis for the decision-making behavior of each agent. Specifically, the enterprise's planned production time is set as T, the product types include hot-rolled plate, hot-rolled flat coil plate, M specifications of wire rod, and N specifications of bar rod, and there are two ironmaking equipment: a large blast furnace and a small blast furnace. The iron output of the large blast furnace on day t is F. Iron _ b (t), the unit production cost of the large blast furnace is UC. Iron _ b The hot metal output of the small blast furnace on day t is F. Iron_s (t), the unit production cost of a small blast furnace is UC. Iron_s There are two steelmaking units, No. 1 converter and No. 2 converter. The steel output of No. 1 converter on day t is F. Steel _ o (t), the unit production cost of converter No. 1 is UC Steel _ o The steel output of converter No. 2 on day t is F. Steel_t (t), the unit production cost of converter No. 2 is UC Steel_t There is one hot rolling mill production line. The output of the hot rolling mill on day t is... The unit production cost of hot-rolled steel is UC HR There is one leveling and slitting machine production line. The output of the leveling and slitting machine on day t is... The unit production cost of hot-rolled, leveled, slitting sheet is UC. Smooth There are three bar production lines. The output of bar of specification n on day t of the j-th bar production line is... The unit production cost of the nth specification bar is There are two wire production lines. The output of wire of specification m on day t of the j-th wire production line is... The unit production cost of the m-th specification bar is The demand for hot-rolled steel plates is D. HR The price of hot-rolled steel is Up. HR The demand for hot-rolled flattened slitting sheets is D. Smooth The price of hot-rolled flat slitting sheet is Up. Smooth The demand for the nth specification bar is The price of the nth specification of wire is The demand for the m-th specification of wire is The price of the m-th specification wire is The objective function of the firm's production model is to maximize production profit, that is:

[0031]

[0032] During the production process, the material balance of ironmaking and steelmaking, the material balance of hot rolling and leveling coiling, and the material balance of wire rod and bar rolling are used as the limiting conditions of the enterprise's production model.

[0033] The material balance for ironmaking and steelmaking satisfies the following constraints: (1) The total daily output of molten iron from the large and small blast furnaces is equal to the steel output from converter No. 1 and converter No. 2, i.e., F Iron _ b (t)-F Steel _ o (t)=F Steel _ t (t)-F Iron _ s (t)t∈T;(2)The daily output of billets and slabs is equal to the sum of the steel output of converter No. 1 and converter No. 2, that is in This represents the output of slabs on day t. The slabs are used to produce hot-rolled plates and hot-rolled leveled coils. The billet output on day t is used to produce steel and bar stock.

[0034] The hot-rolled and leveled coil material balance meets the following constraints: (1) The daily hot-rolled plate output and slab inventory are equal to the previous day's inventory plus the current day's slab output, i.e. in (1) The inventory of the slab warehouse on day t; (2) The daily output of hot-rolled plates is less than or equal to its maximum output, i.e. in (3) The daily output of hot-rolled and leveled slitting plates is less than or equal to its maximum output, i.e. in (3) The maximum daily output of hot-rolled and leveled spooled plates; (4) The demand for hot-rolled plates is the sum of the difference between the daily hot-rolled output and the daily hot-rolled and leveled spooled plates, i.e. (5) The demand for hot-rolled and leveled slitting plates is the sum of the daily production output of hot-rolled and leveled slitting plates, i.e.

[0035] The material balance for wire and bar rolling should meet the following constraints: (1) On day t, bar production line j produces only one specification of bar, i.e. t∈T, where (1) Whether the j-th bar production line produces the n-th rolled bar on day t; (2) The j-th wire production line produces only one type of wire on day t, i.e. t∈T, where (2) Whether the j-th wire rod production line produces the m-th rolled wire rod on day t; (3) The bar production output of the j-th bar production line on day t is less than or equal to its maximum daily output. t∈T, n∈N, where (3) The maximum daily output of the j-th production line for producing the n-th specification of bar stock; (4) On day t, the output of the j-th wire production line is less than or equal to its maximum daily output. t∈T, m∈M where (5) The total daily output of the j-th production line for the m-th specification of wire rod; (6) The total output of the j-th production line for the t-th specification of bar rod plus the daily inventory equals the previous day's bar rod inventory plus the billet supply to the bar rod, i.e. t∈T, where The quantity of billets supplied to the j-th bar production line on day t. (6) The total output of m types of wire rod produced by the j-th production line on day t, plus the inventory of that day, equals the inventory of wire rod on the previous day plus the supply of billets to wire rod, i.e. t∈T, where The quantity of billets supplied to the j-th wire rod production line on day t. Let be the inventory of the j-th billet-wire stock on day t; (7) the slab output on day t is the sum of the slab quantities supplied to the bar production line and the wire production line, i.e. in The quantity of billets supplied to the j-th bar production line on day t. The quantity of billets supplied to the j-th wire rod production line on day t. The demand for the nth specification of bar is the sum of the daily outputs of the two production lines for that specification of bar, which is the billet output on day t. (9) The demand for wire of the m-th specification is the sum of the daily output of the two production lines for that specification, i.e.

[0036] For supply chain-level intelligent agents, the Manufacturing Resource Allocation Planning Agent uses the JADE platform and MySQL database to interact with other enterprises to exchange order information and complete the bidding process. It then obtains the enterprise's product production orders and places them into scheduling tasks according to priority. The priority calculation formula is as follows: Where p represents task priority, processtime represents processing time, endtime represents the latest start time, and nowtime represents the current time. Priorities are written into the task queue in descending order. When a new task arrives or a task fails and returns, it is inserted as an urgent order, requiring an update to the priority table. At this time, urgent orders have the highest priority. When an order is canceled, all tasks for that order are deleted, the priority table is updated, and high-priority orders are tendered to the production chain-level agents through the JADE platform. Order information includes the task quantity for each product specification, order delivery time, and price of each product specification in the order. The material procurement planning agent's decision is determined by the enterprise production model and obtained through communication with the manufacturing resource allocation planning agent via the JADE platform's message transmission system. Order information is used, and the Gurobi solver is used to solve the enterprise production model to determine the required molten iron output to complete the order. A material-molten iron output neural network model is trained using historical data on enterprise material inputs and molten iron output from a MySQL database through machine learning to determine the material input ratio per unit of molten iron output. Then, the required molten iron output for the order is used to derive the input quantities of each material, generating an enterprise material procurement plan. The material procurement plan is used to tender to upstream enterprises, selecting the lowest-priced materials with the goal of maximizing the enterprise's profit. The product marketing plan agent communicates with the production chain-level agent through the JADE platform to obtain product inventory information for each process, and tenders to downstream enterprises with the goal of maximizing the enterprise's profit, selecting the downstream enterprise with the highest bid to complete the contract.

[0037] For production chain-level agents, the JADE platform and MySQL database are used to interact with production equipment agents and supply chain-level agents and complete bidding processes. The workshop production planning agent uploads production line information, daily maximum production capacity information, production cost information, and product inventory information from each production chain-level agent to the MySQL database and bids for the manufacturing resource allocation planning agent through the JADE platform. The enterprise production model is used to determine the workload of each process using the Gurobi solver with the enterprise profitability value as the evaluation standard for the bid. The workshop operation planning agent bids for the production equipment agents through the JADE platform. Similarly, the enterprise production model is used to determine the daily output of different specifications of products of each production equipment using the enterprise profitability value as the evaluation standard for the bid. At the same time, the production data information and production status of each process production equipment-level agent are obtained through the MySQL database. If abnormal data occurs in a certain production equipment, the workshop operation planning agent reports to the manufacturing resource allocation planning agent, and the task of that equipment is put back into the task pool as a failed order.

[0038] For production equipment agents, the JADE platform and MySQL database are used to solicit bids from production chain-level agents. Each production equipment agent uploads production data and operating condition information to the MySQL database through sensor data. The JADE platform is used to conduct bidding based on the production data and operating condition information. The workshop operation planning agent uses the enterprise profitability as the evaluation standard for the bids and the Gurobi solver to determine the daily output of different specifications of products for each production equipment. During the production process, the equipment agents use the historical material input information, carbon emission information, and product output information of each production equipment in the MySQL database to train a neural network model with material input as input and carbon emission and product output as output. This model determines the relationship between the material input ratio, carbon emission factor, and unit product output of each production equipment. Each production equipment then adaptively makes optimization decisions on material input and resource allocation for each process.

[0039] Beneficial technical effects of the present invention:

[0040] In response to the need for the steel industry's production and industrial chains to meet the requirements of lean and green transformation and upgrading, as well as the unforeseen circumstances that may arise, the intelligentization of steel enterprises requires a change in the existing collaborative mechanisms of the production and industrial chains. This can be achieved by using the Contract Network protocol to construct an industrial internet-based collaborative method for the production and industrial chains, which consists of three levels of intelligent agents and is based on a multi-agent structure.

[0041] The multi-agent industrial IoT production chain and industrial chain collaboration method described in this invention is suitable for the collaborative optimization of the steel industry's industrial chain and production chain. Through collaboration between agents at the same level and between agents at multiple levels, a systematic multi-agent collaboration method is formed around production and operation, covering three levels of intelligent systems: production equipment agents, production chain-level agents, and industrial chain-level agents. Utilizing big data technology based on the Industrial Internet and artificial intelligence technology oriented towards deep perception, it aims to achieve stable product quality, timely order delivery, and maximized enterprise benefits for steel companies. It not only enables real-time production planning decisions during production but also effectively addresses unexpected situations such as equipment damage and urgent orders. Attached Figure Description

[0042] Figure 1 A schematic diagram of a collaborative method for production and industrial chains based on multi-agent industrial IoT according to an embodiment of the present invention;

[0043] Figure 2 Flowchart of the contract network protocol mechanism in an embodiment of the present invention. Detailed Implementation

[0044] The present invention will be further described below with reference to the accompanying drawings and embodiments;

[0045] Under the constraints of flexible supply and dynamic demand in the external market, dynamic production status within the enterprise (equipment capacity, process parameters), external resource consumption, and carbon emissions, this invention aims to improve the comprehensive evaluation indicators of the entire production process, including product output, quality, energy consumption, emissions, and costs. It adopts a virtual manufacturing process to achieve feedforward decision correction based on virtual simulation, and uses large-scale sensing data in industrial production for working condition identification and feedback to achieve self-optimization decision-making and dynamic optimization decision-making through human-machine interaction, thereby establishing a collaborative method for the industrial chain and production chain based on multiple intelligent agents.

[0046] A method for collaborative production and industrial chain development based on multi-agent industrial IoT, as shown in the appendix. Figure 1 As shown, it consists of three parts: production equipment intelligent agent, production chain-level collaborative intelligent agent, and industrial chain-level resource allocation intelligent agent;

[0047] Production equipment agents correspond to production equipment and include blast furnace agents, converter agents, and rolling mill agents, each fulfilling its own production goals. Production chain-level agents include workshop production planning agents and workshop operation planning agents, responsible for achieving the production goals of each process in the production chain. On one hand, they generate operation plans for each process's production equipment and control the production equipment agents to realize the production of the underlying production equipment. On the other hand, they obtain the company's production plan from the supply chain-level agents and are controlled by the supply chain-level agents. Supply chain-level agents include material procurement planning agents, product marketing planning agents, and manufacturing resource allocation planning agents, optimizing the company's overall production process, determining the production plans for each process, and guiding each process to complete its production operations.

[0048] The production equipment intelligent agent, the production chain-level collaborative intelligent agent, and the industry chain-level resource allocation intelligent agent use the Contract Net Protocol (CNP) mechanism for communication and decision-making. The Contract Net Protocol process is as follows: Figure 2 As shown; the contract network consists of several nodes, divided into two types: managers and agents. In the contract network, managers and agents exchange messages dynamically through negotiation, rather than pre-defined ones. The decision-making process of the contract network includes the following four basic steps:

[0049] Step 1: The manager posts the task for bidding;

[0050] Step 2: The agent submits its bidding information to the manager, including production capacity and production costs;

[0051] Step 3: Managers select an agent to perform the task based on the principle of maximizing the company's profits;

[0052] Step 4: Finally, the manager confirms the contract, thus completing the bidding and tendering decision-making process;

[0053] The functions of the intelligent production equipment agent are as follows: acquiring production data information of each piece of equipment through sensors, including: production efficiency, production benefits, product quality, production equipment operating status, environmental changes, material types and quality, energy resources, human resources, and equipment resources; intelligently sensing production operating condition information, including: material input, energy consumption, and carbon emissions; and adaptively making optimization decisions, optimizing resource allocation, and rationally allocating and recycling energy using neural network models; the intelligent production equipment agents adopt a communication and information sharing strategy, and effective communication and information sharing are required between the agents to understand each other's status, intentions, and action plans. Communication is achieved through direct message transmission, sharing of knowledge bases, or sensor data.

[0054] The functions of the blast furnace agent, converter agent, and rolling mill agent are as follows:

[0055] (1) Blast Furnace Agent: Each blast furnace agent bids for its own production tasks from the workshop operation plan agent, while simultaneously sensing production data and production conditions;

[0056] (2) Converter Agent: Each converter agent bids for its own production tasks from the workshop operation planning agent, while simultaneously sensing its own production data and production status.

[0057] (3) Rolling mill agent: Each rolling mill agent bids for its own production tasks from the workshop operation plan agent, and at the same time senses its own production data and production conditions;

[0058] The production chain-level intelligent agent receives order production tasks and makes autonomous decisions using a contract network, formulating cooperation strategies. These decisions and strategies involve task allocation and require comprehensive consideration of overall goals and individual interests. Production at each process stage is coordinated and controlled by workshop production planning agents and workshop operation planning agents for ironmaking, steelmaking, and rolling processes, enabling seamless workshop production. The workshop production planning agent formulates workshop production plans based on the production capacity of each process transmitted through the JADE platform and the enterprise's production model. The workshop operation planning agent fully utilizes the production capacity of each process, rationally organizes production, ensures the schedulability of resources for each process, and rationally arranges operations for each process, achieving coordinated matching of processes and ensuring smooth operation of the entire workshop.

[0059] The workshop production planning agent and workshop operation planning agent cooperate to achieve the global goal, and their functions are as follows:

[0060] (1) Workshop Production Planning Agent: Obtains the workshop production plan by bidding to the Manufacturing Resource Allocation Planning Agent using the Contract Network Protocol;

[0061] (2) Workshop Operation Planning Agent: Bid for each production equipment agent and select a production equipment agent to complete the workshop operation plan;

[0062] The supply chain-level intelligent agent achieves overall goals by coordinating the production of each process. Through material procurement planning agents, product marketing planning agents, and manufacturing resource allocation planning agents, it coordinates upstream and downstream enterprises to complete material procurement and product marketing plans. At the same time, it prioritizes the orders received by enterprises to improve the overall efficiency of the enterprise.

[0063] The Material Procurement Planning Agent, Product Marketing Planning Agent, and Manufacturing Resource Allocation Planning Agent have the following functions:

[0064] (1) Material Procurement Planning Agent: The company's production depends on the materials of the upstream companies; therefore, the company bids for the upstream companies according to the material procurement planning agent to complete the procurement of the required materials.

[0065] (2) Product Marketing Plan Agent: The production of downstream enterprises depends on the products of enterprises. Therefore, enterprises tender to downstream enterprises according to the product marketing plan agent to complete the enterprise's product marketing plan;

[0066] (3) Manufacturing Resource Allocation Planning Agent: Responsible for generating production plans based on the production orders received by the enterprise, and issuing the orders generated based on the production plans to the production planning agents at the production chain level workshop, and completing the bidding and tendering with the production chain level intelligent agents;

[0067] The production chain and industry chain collaboration method is implemented using the Java JADE framework. The basic functions of the JADE framework include an Agent Management System (AMS) responsible for monitoring and managing platform access and usage, a Directory Service Agent (DF) providing default yellow pages, and a message transmission system. The AMS monitors and controls the Agent platform; each agent must register with the AMS to obtain a valid AID. The DF provides default yellow pages services to the agents. The message transmission system is a software component within the platform that controls all message exchanges. Agents interact and communicate through these functions, while a MySQL database is used for information storage and retrieval. Production equipment agents, production chain-level agents, and industry chain-level agents are created using the basic functions of the JADE platform. Information exchange between different agents is completed through the MySQL database and the message transmission system in the JADE framework. Agent identification is achieved through the AID in the JADE framework.

[0068] A multi-agent-based method for collaborative production and supply chains in the Industrial Internet of Things (IIoT) includes the following:

[0069] A production model is established to provide a basis for the decision-making behavior of each agent. The planned production time is set as T. Product types include hot-rolled plate, hot-rolled flattened coil plate, M specifications of wire rod, and N specifications of bar rod. There are two ironmaking units: a large blast furnace and a small blast furnace. The iron output of the large blast furnace on day t is F. Iron _ b (t), the unit production cost of the large blast furnace is UC. Iron _ b The hot metal output of the small blast furnace on day t is F. Iron_s (t), the unit production cost of a small blast furnace is UC. Iron_s There are two steelmaking units, No. 1 converter and No. 2 converter. The steel output of No. 1 converter on day t is F. Steel _ o (t), the unit production cost of converter No. 1 is UC Steel _ o The steel output of converter No. 2 on day t is F. Steel _ t (t), the unit production cost of converter No. 2 is UC Steel _ t There is one hot rolling mill production line. The output of the hot rolling mill on day t is... The unit production cost of hot-rolled steel is UC HR There is one leveling and slitting machine production line. The output of the leveling and slitting machine on day t is... The unit production cost of hot-rolled, leveled, slitting sheet is UC. Smooth There are three bar production lines. The output of bar of specification n on day t of the j-th bar production line is... The unit production cost of the nth specification bar is There are two wire production lines. The output of wire of specification m on day t of the j-th wire production line is... The unit production cost of the m-th specification bar is The demand for hot-rolled steel plates is D. HR The price of hot-rolled steel is Up. HR The demand for hot-rolled flattened slitting sheets is D. Smooth The price of hot-rolled flat slitting sheet is Up. Smooth The demand for the nth specification bar is The price of the nth specification of wire is The demand for the m-th specification of wire is The price of the m-th specification wire is The objective function of the firm's production model is to maximize production profit, that is:

[0070]

[0071] During the production process, the material balance of ironmaking and steelmaking, the material balance of hot rolling and leveling coiling, and the material balance of wire rod and bar rolling are used as the limiting conditions of the enterprise's production model.

[0072] The material balance for ironmaking and steelmaking should meet the following constraints: (1) The total daily output of molten iron from the large and small blast furnaces is equal to the steel output of converter No. 1 and converter No. 2, i.e., F Iron _ b (t)-F Steel _ o (t)=F Steel _ t (t)-F Iron _ s (t)t∈T;(2)The daily output of billets and slabs is equal to the sum of the steel output of converter No. 1 and converter No. 2, that is in This represents the output of slabs on day t. The slabs are used to produce hot-rolled plates and hot-rolled leveled coils. The billet output on day t is used to produce steel and bar stock.

[0073] The material balance of hot-rolled and leveled coils should meet the following constraints: (1) The daily output of hot-rolled plates and the inventory of slabs are equal to the inventory of the previous day plus the output of slabs on the current day, i.e. in (1) The inventory of the slab warehouse on day t; (2) The daily output of hot-rolled plates is less than or equal to its maximum output, i.e. in (3) The daily output of hot-rolled and leveled slitting plates is less than or equal to its maximum output, i.e. in (3) The maximum daily output of hot-rolled and leveled spooled plates; (4) The demand for hot-rolled plates is the sum of the difference between the daily hot-rolled output and the daily hot-rolled and leveled spooled plates, i.e. T; (5) The demand for hot-rolled and leveled slitting plates is the sum of the daily production output of hot-rolled and leveled slitting plates, i.e.

[0074] The material balance for wire and bar rolling should meet the following constraints: (1) On day t, bar production line j produces only one specification of bar, i.e. t∈T, where (1) Whether the j-th bar production line produces the n-th rolled bar on day t; (2) The j-th wire production line produces only one type of wire on day t, i.e. t∈T, where (2) Whether the j-th wire rod production line produces the m-th rolled wire rod on day t; (3) The bar production output of the j-th bar production line on day t is less than or equal to its maximum daily output. t∈T, n∈N, where (3) The maximum daily output of the j-th production line for producing the n-th specification of bar stock; (4) On day t, the output of the j-th wire production line is less than or equal to its maximum daily output. t∈T, m∈M where (5) The total daily output of the j-th production line for the m-th specification of wire rod; (6) The total output of the j-th production line for the t-th specification of bar rod plus the daily inventory equals the previous day's bar rod inventory plus the billet supply to the bar rod, i.e. t∈T, where The quantity of billets supplied to the j-th bar production line on day t. (6) The total output of m types of wire rod produced by the j-th production line on day t, plus the inventory of that day, equals the inventory of wire rod on the previous day plus the supply of billets to wire rod, i.e. t∈T, where The quantity of billets supplied to the j-th wire rod production line on day t. Let be the inventory of the j-th billet-wire stock on day t; (7) the slab output on day t is the sum of the slab quantities supplied to the bar production line and the wire production line, i.e. in The quantity of billets supplied to the j-th bar production line on day t. The quantity of billets supplied to the j-th wire rod production line on day t. The demand for the nth specification of bar is the sum of the daily outputs of the two production lines for that specification of bar, which is the billet output on day t. (9) The demand for wire of the m-th specification is the sum of the daily output of the two production lines for that specification, i.e.

[0075] For supply chain-level intelligent agents, the Manufacturing Resource Allocation Planning Agent uses the JADE platform and MySQL database to interact with other enterprises to exchange order information and complete the bidding process. It then obtains the enterprise's product production orders and places them into scheduling tasks according to priority. The priority calculation formula is as follows: Where p represents task priority, processtime represents processing time, endtime represents the latest start time, and nowtime represents the current time. Priorities are written into the task queue in descending order. When a new task arrives or a task fails and returns, it is inserted as an urgent order, requiring an update to the priority table. At this time, urgent orders have the highest priority. When an order is canceled, all tasks for that order are deleted, the priority table is updated, and high-priority orders are tendered to the production chain-level agents through the JADE platform. Order information includes the task quantity for each product specification, order delivery time, and price of each product specification in the order. The material procurement planning agent's decision is determined by the enterprise production model and obtained through communication with the manufacturing resource allocation planning agent via the JADE platform's message transmission system. Order information is used, and the Gurobi solver is used to solve the enterprise production model to determine the required molten iron output to complete the order. A material-molten iron output neural network model is trained using historical data on enterprise material inputs and molten iron output from a MySQL database through machine learning to determine the material input ratio per unit of molten iron output. Then, the required molten iron output for the order is used to derive the input quantities of each material, generating an enterprise material procurement plan. The material procurement plan is used to tender to upstream enterprises, selecting the lowest-priced materials with the goal of maximizing the enterprise's profit. The product marketing plan agent communicates with the production chain-level agent through the JADE platform to obtain product inventory information for each process, and tenders to downstream enterprises with the goal of maximizing the enterprise's profit, selecting the downstream enterprise with the highest bid to complete the contract.

[0076] For production chain-level agents, the JADE platform and MySQL database are used to interact with production equipment agents and supply chain-level agents and complete bidding processes. The workshop production planning agent uploads production line information, daily maximum production capacity information, production cost information, and product inventory information from each production chain-level agent to the MySQL database and bids for the manufacturing resource allocation planning agent through the JADE platform. The enterprise production model is used to determine the workload of each process using the Gurobi solver with the enterprise profitability value as the evaluation standard for the bid. The workshop operation planning agent bids for the production equipment agents through the JADE platform. Similarly, the enterprise production model is used to determine the daily output of different specifications of products of each production equipment using the enterprise profitability value as the evaluation standard for the bid. At the same time, the production data information and production status of each process production equipment-level agent are obtained through the MySQL database. If abnormal data occurs in a certain production equipment, the workshop operation planning agent reports to the manufacturing resource allocation planning agent, and the task of that equipment is put back into the task pool as a failed order.

[0077] For production equipment agents, the JADE platform and MySQL database are used to solicit bids from production chain-level agents. Each production equipment agent uploads production data and operating condition information to the MySQL database through sensor data. The JADE platform is used to conduct bidding based on the production data and operating condition information. The workshop operation planning agent uses the enterprise profitability as the evaluation standard for the bids and the Gurobi solver to determine the daily output of different specifications of products for each production equipment. During the production process, the equipment agents use the historical material input information, carbon emission information, and product output information of each production equipment in the MySQL database to train a neural network model with material input as input and carbon emission and product output as output. This model determines the relationship between the material input ratio, carbon emission factor, and unit product output of each production equipment. Each production equipment then adaptively makes optimization decisions on material input and resource allocation for each process.

Claims

1. A method for collaborative production and industrial chain development in the Industrial Internet of Things (IIoT) based on multi-agent systems, characterized in that, It consists of three parts: intelligent production equipment agent, intelligent production chain-level collaborative agent, and intelligent industrial chain-level resource allocation agent. The production equipment intelligent agents correspond to production equipment and include blast furnace agents, converter agents, and rolling mill agents, each completing its respective production target. The production chain-level collaborative intelligent agents include workshop production planning agents and workshop operation planning agents. On the one hand, they generate operation plans for each process's production equipment and control the production equipment intelligent agents to realize the production of the underlying production equipment. On the other hand, they obtain the company's production plan from the supply chain-level resource allocation intelligent agents and are controlled by the supply chain-level resource allocation intelligent agents. The supply chain-level resource allocation intelligent agents include material procurement planning agents, product marketing planning agents, and manufacturing resource allocation planning agents. They optimize the company's overall production process, determine the production plans for each process, and guide each process to complete its production operations. For supply chain-level resource allocation agents, the manufacturing resource allocation planning agent uses the JADE platform and MySQL database to interact with other enterprises to exchange order information and complete the bidding process. It then obtains the enterprise's product production orders and places them into scheduling tasks according to priority. The priority calculation formula is as follows: Where p represents task priority, processtime represents processing time, endtime represents the latest start time, and nowtime represents the current time. Priorities are written into the task queue in descending order. When a new task arrives or a task fails and returns, it will be inserted as an emergency order, requiring an update to the priority table. At this time, emergency orders have the highest priority. When an order is canceled, all tasks for that order are deleted and the priority table is updated. At the same time, high-priority orders are tendered to the JADE platform for production chain-level agents. Order information includes the task quantity for each product specification, order delivery time, and the price of each product specification in the order. The material procurement planning agent's decisions are determined by the enterprise's production model. It communicates with the manufacturing resource allocation planning agent through the JADE platform's message transmission system to obtain order information. Under the condition of determining the order information, the Gurobi solver is used to solve the enterprise's production model to determine the amount of molten iron required to complete the order. The material-molten iron production neural network model is trained by machine learning on the enterprise's material input and molten iron production historical data in the MySQL database to determine the material input ratio per unit of molten iron production. Then, the amount of each material input is derived from the amount of molten iron required for the order, generating the enterprise's material procurement plan. The material procurement quantity in the material procurement plan is used to tender to upstream enterprises, selecting the lower-priced materials with the goal of maximizing the enterprise's interests. The product marketing plan agent communicates with the production chain level agent through the JADE platform to obtain product inventory information for each process. With the goal of maximizing the company's interests, it bids for downstream companies and selects the downstream company with the highest bid to complete the contract. For production chain-level collaborative agents, the JADE platform and MySQL database are used to interact with production equipment agents and supply chain-level agents to complete bidding and tendering processes. The workshop production planning agent uploads production line information, daily maximum production capacity information, production cost information, and product inventory information from each production chain-level agent to the MySQL database and bids for the manufacturing resource allocation planning agent through the JADE platform. The enterprise production model is used with enterprise profitability as the evaluation standard for the bid, and the Gurobi solver is used to determine the workload of each process. The workshop operation planning agent bids for the production equipment agents through the JADE platform. Similarly, the enterprise production model is used with enterprise profitability as the evaluation standard for the bid, and the Gurobi solver is used to determine the daily output of different specifications of products of each production equipment. At the same time, the production data information and production status of each process's production equipment-level agent are obtained through the MySQL database. For production equipment agents, the JADE platform and MySQL database are used to solicit bids from production chain-level agents. Each production equipment agent uploads production data and operating condition information to the MySQL database through sensor data. The JADE platform is used to conduct bidding based on the production data and operating condition information. The workshop operation planning agent uses the enterprise profitability as the evaluation standard for the bids and the Gurobi solver to determine the daily output of different specifications of products for each production equipment. During the production process, the equipment agents use the historical material input information, carbon emission information, and product output information of each production equipment in the MySQL database to train a neural network model with material input as input and carbon emission and product output as output. This model determines the relationship between the material input ratio, carbon emission factor, and unit product output of each production equipment. Each production equipment then adaptively makes optimization decisions on material input and resource allocation for each process.

2. The method for collaborative production and industrial chain development based on multi-agent industrial IoT according to claim 1, characterized in that, The production equipment intelligent agent, the production chain-level collaborative intelligent agent, and the industry chain-level resource allocation intelligent agent communicate and make decisions using the Contract Net Protocol (CNP) mechanism. The Contract Net consists of n nodes, divided into two types: managers and agents. In the Contract Net, managers and agents exchange dynamic messages through negotiation. The decision-making process of the Contract Net includes the following four basic steps: Step 1: The manager posts the task for bidding; Step 2: The agent submits its bidding information to the manager, including production capacity and production costs; Step 3: Managers select an agent to perform the task based on the principle of maximizing the company's profits; Step 4: Finally, the manager confirms the contract, thus completing the bidding and tendering decision-making process.

3. The method for collaborative production and industrial chain development based on multi-agent industrial IoT according to claim 1, characterized in that, The functions of the intelligent production equipment agent are as follows: acquiring production data information of each piece of equipment through sensors, including: production efficiency, production benefits, product quality, production equipment operating status, environmental changes, material types and quality, energy resources, human resources, and equipment resources; intelligently sensing production operating condition information, including: material input, energy consumption, and carbon emissions; and adaptively making optimization decisions, optimizing resource allocation, and rationally allocating and recycling energy using neural network models; the intelligent production equipment agents adopt a communication and information sharing strategy, and effective communication and information sharing are required between the agents to understand each other's status, intentions, and action plans. Communication is achieved through direct message transmission, sharing of knowledge bases, or sensor data. The functions of the blast furnace agent, converter agent, and rolling mill agent are as follows: (1) Blast Furnace Agent: Each blast furnace agent bids for its own production tasks from the workshop operation plan agent, while simultaneously sensing production data and production conditions; (2) Converter Agent: Each converter agent bids for its own production tasks from the workshop operation plan agent, and at the same time senses its own production data and production conditions; (3) Rolling mill agent: Each rolling mill agent bids for its own production tasks from the workshop operation plan agent, and at the same time senses its own production data and production conditions; The production chain-level collaborative intelligent agent receives order production tasks and makes autonomous decisions using a contract network, formulating cooperation strategies. These decisions and strategies involve task allocation and require comprehensive consideration of overall goals and individual interests. Production at each process stage is coordinated and controlled by workshop production planning agents and workshop operation planning agents for ironmaking, steelmaking, and rolling processes, ensuring smooth workshop production. The workshop production planning agent formulates workshop production plans based on the production capacity of each process transmitted through the JADE platform and the enterprise's production model. The workshop operation planning agent fully utilizes the production capacity of each process, rationally organizes production, ensures the schedulability of resources for each process, and rationally arranges operations for each process, achieving coordinated matching of processes and guaranteeing the smooth operation of the entire workshop. The workshop production planning agent and workshop operation planning agent cooperate to achieve the global goal, and their functions are as follows: (1) Workshop Production Planning Agent: Obtains the workshop production plan by bidding to the Manufacturing Resource Allocation Planning Agent using the Contract Network Protocol; (2) Workshop Operation Planning Agent: Bid for each production equipment agent and select a production equipment agent to complete the workshop operation plan; The industry chain-level resource allocation intelligence agent achieves global goals by coordinating the production of each process. It coordinates upstream and downstream enterprises to complete material procurement and product marketing plans through material procurement planning agents, product marketing planning agents, and manufacturing resource allocation planning agents. At the same time, it prioritizes the orders received by enterprises to improve the overall efficiency of the enterprise. The Material Procurement Planning Agent, Product Marketing Planning Agent, and Manufacturing Resource Allocation Planning Agent have the following functions: (1) Material Procurement Planning Agent: The company's production depends on the materials of the upstream companies; therefore, the company bids for the upstream companies according to the material procurement planning agent to complete the procurement of the required materials. (2) Product Marketing Plan Agent: The production of downstream enterprises depends on the products of enterprises. Therefore, enterprises tender to downstream enterprises according to the product marketing plan agent to complete the enterprise's product marketing plan; (3) Manufacturing Resource Allocation Plan Agent: Responsible for generating production plans based on the production orders received by the enterprise, and issuing the orders generated based on the production plans to the production plan agents at the production chain level workshops, and completing the bidding and tendering process with the production chain level collaborative intelligent agents.

4. The method for collaborative production and industrial chain development based on multi-agent industrial IoT according to claim 1, characterized in that, The production chain and industry chain collaboration method is implemented using the Java JADE framework. The basic functions of the JADE framework include an Agent Management System (AMS) responsible for monitoring and managing platform access and usage, a Directory Service Agent (DF) providing default yellow pages, and a message transmission system. The AMS monitors and controls the Agent platform; each agent must register with the AMS to obtain a valid AID. The DF provides default yellow pages services to the agents. The message transmission system is a software component within the platform that controls all message exchanges. Agents interact and communicate through these functions, while a MySQL database is used for information storage and retrieval. Production equipment agents, production chain-level collaborative agents, and industry chain-level resource allocation agents are created using the basic functions of the JADE platform. Information exchange between different agents is completed through the MySQL database and the message transmission system in the JADE framework. Agent identification is achieved through the AID in the JADE framework.

5. The method for collaborative production and industrial chain development based on multi-agent industrial IoT according to claim 1, characterized in that, Also includes the following: A production model is established to provide a basis for the decision-making behavior of each agent. The planned production time is set as T. Product types include hot-rolled plate, hot-rolled flattened coil plate, M specifications of wire rod, and N specifications of bar rod. There are two ironmaking units: a large blast furnace and a small blast furnace. The iron output of the large blast furnace on day t is F. Iron_b (t), the unit production cost of the large blast furnace is UC. Iron_b The hot metal output of the small blast furnace on day t is F. Iron_s (t), the unit production cost of a small blast furnace is UC. Iron_s There are two steelmaking units, No. 1 converter and No. 2 converter. The steel output of No. 1 converter on day t is F. Steel_o (t), the unit production cost of converter No. 1 is UC Steel_o The steel output of converter No. 2 on day t is F. Steel_t (t), the unit production cost of converter No. 2 is UC Steel_t There is one hot rolling mill production line. The output of the hot rolling mill on day t is... The unit production cost of hot-rolled steel is UC HR There is one leveling and slitting machine production line. The output of the leveling and slitting machine on day t is... The unit production cost of hot-rolled, leveled, slitting sheet is UC. Smooth There are three bar production lines. The output of bar of specification n on day t of the j-th bar production line is... The unit production cost of the nth specification bar is There are two wire production lines. The output of wire of specification m on day t of the j-th wire production line is... The unit production cost of the m-th specification bar is The demand for hot-rolled steel plates is D. HR The price of hot-rolled steel is Up. HR The demand for hot-rolled flattened slitting sheets is D. Smooth The price of hot-rolled flat slitting sheet is Up. Smooth The demand for the nth specification bar is The price of the nth specification of wire is The demand for the m-th specification of wire is The price of the m-th specification wire is The objective function of the firm's production model is to maximize production profit, that is: In the production process, the material balance of ironmaking and steelmaking, hot rolling and leveling coiling, and wire rod and bar rolling are used as the limiting conditions of the enterprise's production model.

6. The method for collaborative production and industrial chain development based on multi-agent industrial IoT according to claim 5, characterized in that, The material balance for ironmaking and steelmaking satisfies the following constraints: (1) The total daily output of molten iron from the large and small blast furnaces is equal to the steel output from converter No. 1 and converter No. 2, i.e., F Iron_b (t)-F Steel_o (t)=F Steel_t (t)-F Iron_s (t)t∈T;(2)The daily output of billets and slabs is equal to the sum of the steel output of converter No. 1 and converter No. 2, that is in This represents the output of slabs on day t. The slabs are used to produce hot-rolled plates and hot-rolled leveled coils. The billet output on day t is used to produce steel and bar stock. The hot-rolled and leveled coil material balance meets the following constraints: (1) The daily hot-rolled plate output and slab inventory are equal to the previous day's inventory plus the current day's slab output, i.e. in (1) The inventory of the slab warehouse on day t; (2) The daily output of hot-rolled plates is less than or equal to its maximum output, i.e. in (3) The daily output of hot-rolled and leveled slitting plates is less than or equal to its maximum output, i.e. in (3) The maximum daily output of hot-rolled and leveled spooled plates; (4) The demand for hot-rolled plates is the sum of the difference between the daily hot-rolled output and the daily hot-rolled and leveled spooled plates, i.e. (5) The demand for hot-rolled and leveled slitting plates is the sum of the daily production output of hot-rolled and leveled slitting plates, i.e. The material balance for wire and bar rolling should meet the following constraints: (1) On day t, bar production line j produces only one specification of bar, i.e. in (1) Whether the j-th bar production line produces the n-th rolled bar on day t; (2) The j-th wire production line produces only one type of wire on day t, i.e. in (2) Whether the j-th wire rod production line produces the m-th rolled wire rod on day t; (3) The bar production output of the j-th bar production line on day t is less than or equal to its maximum daily output. in (3) The maximum daily output of the j-th production line for producing the n-th specification of bar stock; (4) On day t, the output of the j-th wire production line is less than or equal to its maximum daily output. in (5) The total daily output of the j-th production line for the m-th specification of wire rod; (6) The total output of the j-th production line for the t-th specification of bar rod plus the daily inventory equals the previous day's bar rod inventory plus the billet supply to the bar rod, i.e. in The quantity of billets supplied to the j-th bar production line on day t. (6) The total output of m types of wire rod produced by the j-th production line on day t, plus the inventory of that day, equals the inventory of wire rod on the previous day plus the supply of billets to wire rod, i.e. in The quantity of billets supplied to the j-th wire rod production line on day t. Let be the inventory of the j-th billet-wire stock on day t; (7) the slab output on day t is the sum of the slab quantities supplied to the bar production line and the wire production line, i.e. in The quantity of billets supplied to the j-th bar production line on day t. The quantity of billets supplied to the j-th wire rod production line on day t. The demand for the nth specification of bar is the sum of the daily outputs of the two production lines for that specification of bar, which is the billet output on day t. (9) The demand for wire of the m-th specification is the sum of the daily output of the two production lines for that specification, i.e.

Images