Factory equipment cooperative control method based on swarm intelligence

Abstract

The invention relates to the technical field of industrial automation and intelligent manufacturing, in particular to a factory equipment cooperative control method based on swarm intelligence, which comprises the following steps: step 1, agent abstraction and task release: abstracting each independently controllable equipment in a factory into an agent with an autonomous decision-making capability; the upper-layer production management system decomposes a total production task into a plurality of ordered sub-tasks, each sub-task comprises required equipment types, process parameter requirements, predicted time consumption and priority information, and the sub-tasks are published to a global task pool; according to the method, a consensus decision-making mechanism of swarm intelligence is creatively applied to the cooperative control field of the roller kiln assembly line, the swarm intelligence algorithm is mostly applied to the fields of robot cluster control, traffic flow optimization and the like in the past, the algorithm is introduced into roller kiln assembly line control, the application scene of the algorithm is expanded, and the application prospect is wide. And a new thought and method are provided for solving the problem of complex cooperative control of the roller kiln assembly line.

Description

Technical Field

[0001] This invention relates to the fields of industrial automation and intelligent manufacturing technology, specifically to a collaborative control method for factory equipment based on swarm intelligence. Background Technology

[0002] In industries such as ceramics, lithium battery electrodes, and powder metallurgy, roller kiln production lines are core production equipment. They typically need to work in conjunction with various other equipment, such as loading and unloading robotic arms, testing instruments, and packaging machines, to complete the entire production process.

[0003] Traditional control methods employ a centralized approach, where a central controller (such as a Manufacturing Execution System (MES) or a Programmable Logic Controller (PLC) centrally plans the production tasks and schedules of all equipment. However, this centralized control scheme has significant drawbacks, as follows:

[0004] Bottlenecks and Single Point of Failure Risks: The central controller, being the core of the entire system, becomes a bottleneck for system performance. Its computing power directly limits the system's scale. As the number of production devices increases and production tasks become more complex, the central controller struggles to efficiently process large amounts of data and task instructions. More seriously, if the central controller fails, the entire production line may be paralyzed. For roller kiln production lines, a production line shutdown can lead to the scrapping of products being processed in the kiln due to the inability to be processed, resulting in significant economic losses for the company.

[0005] Rigidity and poor adaptability: Production plans under centralized control are pre-defined and static, unable to flexibly respond to dynamic changes in the production process. When dynamic disturbances occur, such as the insertion of urgent orders, sudden equipment failures (e.g., a roller jamming in a roller kiln), or the need for dynamic adjustment of process parameters (e.g., requiring accelerated heating in a certain temperature zone of the roller kiln to meet special production needs), the entire system needs to be shut down, and the central controller must re-plan the overall system. This process is not only slow in response but also severely impacts production efficiency. Furthermore, frequent start-ups and shutdowns and fluctuations in process parameters can adversely affect product quality.

[0006] High communication load: In a centralized control architecture, the status information of all devices needs to be reported to the central node in real time, and all control commands issued by the central node also need to be transmitted to each device. As the scale of factory production expands and the number of devices increases, this bidirectional large-scale data transmission can lead to excessive network communication load, easily causing data transmission delays and losses, which in turn affects the stability and reliability of the entire production system.

[0007] Existing technologies include some distributed control schemes that attempt to address some of the problems of centralized control. However, the collaborative strategies of these distributed control schemes are often too simplistic and cannot handle the complex multi-objective optimization problems in roller kiln production processes. For example, how to simultaneously achieve multiple objectives such as reducing energy consumption, optimizing firing curves, and balancing equipment load while ensuring product quality. Therefore, these schemes struggle to cope with complex and ever-changing factory production environments and cannot meet the high requirements for production continuity and stability of roller kiln production lines. To address this, a collaborative control method for factory equipment based on swarm intelligence is proposed. Summary of the Invention

[0008] In view of this, the present invention provides a collaborative control method for factory equipment based on swarm intelligence to solve or alleviate the technical problems existing in the prior art, and at least provides a beneficial option.

[0009] The technical solution of this invention is implemented as follows: a collaborative control method for factory equipment based on swarm intelligence, comprising the following steps: Step 1, Agent Abstraction and Task Deployment: Each independently controllable piece of equipment in the factory is abstracted into an agent with autonomous decision-making capabilities; the upper-level production management system decomposes the total production task into multiple ordered sub-tasks, each sub-task containing the required equipment type, process parameter requirements, estimated time consumption, priority information, and publishes it to the global task pool. Step 2, Task Bidding and Decision-Making: Each agent perceives the sub-tasks available for execution in the global task pool through local communication. For the perceived sub-task j, it calculates its own bidding value Bid_i(j) using the formula Bid_i(j) = α*(T_ij^setup + T_j^process) + βU_i + γP_j + δ*E_ij, where T_ij^setup is the preparation time for agent i to execute task j, T_j^process is the estimated execution time of task j, U_i is the current utilization rate of agent i, P_j is the priority coefficient of task j, E_ij is the estimated energy cost for agent i to execute task j, and α, β, γ, and δ are adjustable weight coefficients. Each agent broadcasts its bidding value for all tasks of interest to surrounding agents or lightweight coordination nodes. Step 3, Conflict Resolution and Task Allocation: Based on the bidding information received from other agents, each agent uses a consensus-based group decision-making algorithm to determine the agent with the lowest bid value as the executor for the same task j. This agent removes task j from the task pool and adds it to its own task queue. Step 4, Collaborative Execution and Dynamic Adjustment: The agent that wins the bid executes the task and periodically broadcasts its own state; when a dynamic event occurs, the affected agent triggers a new round of Step 2 and Step 3 to rebid and reassign unassigned or reassigned tasks.

[0010] More preferably, in step one, the independently controllable equipment includes a robotic arm, an independent temperature zone module of the roller kiln, a transmission group of the roller kiln, a detector, and a packaging machine.

[0011] More preferably, in step two, for the roller kiln temperature zone intelligent agent, T_ij^setup is the heating or cooling time required for the intelligent agent to reach the target process temperature; for the roller kiln drive group intelligent agent, T_ij^setup is the time required for the intelligent agent to adjust to the target speed.

[0012] More preferably, in step two, the consensus-based group decision-making algorithm is a bee honey-collecting model.

[0013] More preferably, in step four, the agent periodically broadcasts its own state, including current temperature, speed, energy consumption, working status (busy / idle), and remaining task queue.

[0014] More preferably, in step four, dynamic events include process formula changes, temperature zone heating element failures, transmission speed synchronization adjustments, emergency order insertions, and equipment maintenance requests.

[0015] In a further preferred embodiment, in step four, when a certain temperature zone agent of the roller kiln malfunctions, the agent in that temperature zone releases its sintering task back to the task pool. Adjacent temperature zone agents participate in the redistribution of the sintering task through re-bidding and complete the sintering task by adjusting their own process parameters in a collaborative manner.

[0016] More preferably, in step two, the lightweight coordination node is only used for information aggregation and forwarding, and does not participate in the specific decision-making process.

[0017] In a further preferred embodiment, in step one, the upper-level production management system includes an Enterprise Resource Planning (ERP) system and a Manufacturing Execution System (MES).

[0018] More preferably, this method is applicable to production systems in the ceramics, lithium battery electrode, and powder metallurgy industries that include roller kiln production lines.

[0019] The embodiments of the present invention have the following advantages due to the adoption of the above technical solutions: I. This invention creatively applies the consensus decision-making mechanism of swarm intelligence to the field of collaborative control of roller kiln production lines. Swarm intelligence algorithms have been used in robot swarm control, traffic flow optimization and other fields in the past. Introducing them into the control of roller kiln production lines is an expansion of the application scenarios of this algorithm and provides new ideas and methods for solving the complex collaborative control problems of roller kiln production lines.

[0020] Second, this invention designs a multi-objective bidding function that integrates process preparation time (T_ij^setup) and energy consumption cost (E_ij). In the production process of roller kiln, process preparation time directly affects production efficiency, while energy consumption cost is related to the enterprise's production cost and environmental protection requirements. Previous control methods often find it difficult to simultaneously take into account these two important factors. The bidding function of this invention comprehensively considers multiple objectives such as time, load, priority, and energy consumption, and can effectively solve the multi-objective optimization problem unique to the field of roller kiln production line control.

[0021] Third, this invention enables a near-globally optimal production scheduling scheme to be autonomously decided by a group of equipment. Through local interaction and group consensus among various intelligent agents, the optimized scheduling of the entire production system can be achieved without the need for a central controller to perform global planning, which is non-obvious in the field of roller kiln production line control.

[0022] The above overview is for illustrative purposes only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features of the invention will become readily apparent from the accompanying drawings and the following detailed description. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a system architecture diagram of the present invention; Figure 2 This is a structural diagram of the intelligent agent execution layer and the physical device layer of the present invention; Figure 3 This is a flowchart illustrating the overall process of the method of the present invention. Figure 4 This is a sequence of the task bidding and allocation process of the present invention. Figure 1 ; Figure 5 This is a sequence of the task bidding and allocation process of the present invention. Figure 2 ; Figure 6 This is a schematic diagram of the dynamic adjustment process of the present invention. Detailed Implementation

[0025] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0026] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0027] like Figure 1-6 As shown, this embodiment of the invention provides a collaborative control method for factory equipment based on swarm intelligence, including the following steps: Step 1, Agent Abstraction and Task Deployment: Each independently controllable piece of equipment in the factory is abstracted into an agent with autonomous decision-making capabilities; the upper-level production management system decomposes the total production task into multiple ordered sub-tasks, each sub-task containing the required equipment type, process parameter requirements, estimated time consumption, priority information, and publishes it to the global task pool. Step 2, Task Bidding and Decision-Making: Each agent perceives the sub-tasks available for execution in the global task pool through local communication. For the perceived sub-task j, it calculates its own bidding value Bid_i(j) using the formula Bid_i(j) = α*(T_ij^setup + T_j^process) + βU_i + γP_j + δ*E_ij, where T_ij^setup is the preparation time for agent i to execute task j, T_j^process is the estimated execution time of task j, U_i is the current utilization rate of agent i, P_j is the priority coefficient of task j, E_ij is the estimated energy cost for agent i to execute task j, and α, β, γ, and δ are adjustable weight coefficients. Each agent broadcasts its bidding value for all tasks of interest to surrounding agents or lightweight coordination nodes. Step 3, Conflict Resolution and Task Allocation: Based on the bidding information received from other agents, each agent uses a consensus-based group decision-making algorithm to determine the agent with the lowest bid value as the executor for the same task j. This agent removes task j from the task pool and adds it to its own task queue. Step 4, Collaborative Execution and Dynamic Adjustment: The agent that wins the bid executes the task and periodically broadcasts its own state; when a dynamic event occurs, the affected agent triggers a new round of Step 2 and Step 3 to rebid and reassign unassigned or reassigned tasks.

[0028] In one embodiment, in step one, the independently controllable equipment includes a robotic arm, an independent temperature zone module of the roller kiln, a transmission group of the roller kiln, a detector, and a packaging machine.

[0029] In one embodiment, in step two, for the roller kiln temperature zone intelligent agent, T_ij^setup is the heating or cooling time required for the intelligent agent to reach the target process temperature; for the roller kiln drive group intelligent agent, T_ij^setup is the time required for the intelligent agent to adjust to the target speed.

[0030] In one embodiment, in step two, the consensus-based group decision-making algorithm is a bee honey-collecting model.

[0031] In one embodiment, in step four, the agent periodically broadcasts its own state, including current temperature, speed, energy consumption, working status (busy / idle), and remaining task queue.

[0032] In one embodiment, in step four, dynamic events include process recipe changes, temperature zone heating element failures, transmission speed synchronization adjustments, emergency order insertions, and equipment maintenance requests.

[0033] In one embodiment, in step four, when a certain temperature zone agent of the roller kiln malfunctions, the agent releases its sintering task back to the task pool. Adjacent temperature zone agents participate in the redistribution of the sintering task through re-bidding and complete the sintering task by adjusting their own process parameters in a collaborative manner.

[0034] In one embodiment, in step two, the lightweight coordination node is only used for information aggregation and forwarding and does not participate in the specific decision-making process.

[0035] In one embodiment, in step one, the upper-level production management system includes an Enterprise Resource Planning (ERP) system and a Manufacturing Execution System (MES).

[0036] In one embodiment, the method is applicable to production systems in the ceramics, lithium battery electrode, and powder metallurgy industries that include roller kiln production lines.

[0037] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope disclosed in the present invention, and these should all be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A collaborative control method for factory equipment based on swarm intelligence, characterized in that: Includes the following steps: Step 1, Agent Abstraction and Task Issuance: Abstract each independently controllable piece of equipment in the factory into an agent with autonomous decision-making capabilities; The upper-level production management system decomposes the total production task into multiple ordered sub-tasks. Each sub-task includes the required equipment type, process parameter requirements, estimated time, and priority information, and publishes them to the global task pool. Step 2, Task Bidding and Decision-Making: Each agent perceives the sub-tasks available for execution in the global task pool through local communication. For the perceived sub-task j, it calculates its own bidding value Bid_i(j) using the formula Bid_i(j) = α*(T_ij^setup + T_j^process) + βU_i + γP_j + δ*E_ij, where T_ij^setup is the preparation time for agent i to execute task j, T_j^process is the estimated execution time of task j, U_i is the current utilization rate of agent i, P_j is the priority coefficient of task j, E_ij is the estimated energy cost for agent i to execute task j, and α, β, γ, and δ are adjustable weight coefficients. Each agent broadcasts its bidding value for all tasks of interest to surrounding agents or lightweight coordination nodes. Step 3, Conflict Resolution and Task Allocation: Based on the bidding information received from other agents, each agent uses a consensus-based group decision-making algorithm to determine the agent with the lowest bid value as the executor for the same task j. This agent removes task j from the task pool and adds it to its own task queue. Step 4, Collaborative Execution and Dynamic Adjustment: The agent that wins the bid executes the task and periodically broadcasts its own state; when a dynamic event occurs, the affected agent triggers a new round of Step 2 and Step 3 to rebid and reassign unassigned or reassigned tasks.

2. The collaborative control method for factory equipment based on swarm intelligence according to claim 1, characterized in that: In step one, the independently controllable equipment includes a robotic arm, an independent temperature zone module of the roller kiln, a transmission group of the roller kiln, a detector, and a packaging machine.

3. The collaborative control method for factory equipment based on swarm intelligence according to claim 1, characterized in that: In step two, for the roller kiln temperature zone smart body, T_ij^setup is the heating or cooling time required for the smart body to reach the target process temperature; for the roller kiln drive group smart body, T_ij^setup is the time required for the smart body to adjust to the target speed.

4. The collaborative control method for factory equipment based on swarm intelligence according to claim 1, characterized in that: In step two, the consensus-based group decision-making algorithm is a bee honey-collecting model.

5. The collaborative control method for factory equipment based on swarm intelligence according to claim 1, characterized in that: In step four, the agent periodically broadcasts its own state, including current temperature, speed, energy consumption, working status (busy / idle), and remaining task queue.

6. The collaborative control method for factory equipment based on swarm intelligence according to claim 1, characterized in that: In step four, dynamic events include process formula changes, temperature zone heating element failures, transmission speed synchronization adjustments, emergency order insertions, and equipment maintenance requests.

7. The collaborative control method for factory equipment based on swarm intelligence according to claim 1, characterized in that: In step four, when a certain temperature zone agent of the roller kiln malfunctions, the agent releases its sintering task back to the task pool. Adjacent temperature zone agents participate in the redistribution of the sintering task through re-bidding and complete the sintering task by adjusting their own process parameters in a collaborative manner.

8. The collaborative control method for factory equipment based on swarm intelligence according to claim 1, characterized in that: In step two, the lightweight coordination node is only used for information aggregation and forwarding and does not participate in the specific decision-making process.

9. The collaborative control method for factory equipment based on swarm intelligence according to claim 1, characterized in that: In step one, the upper-level production management system includes the Enterprise Resource Planning (ERP) system and the Manufacturing Execution System (MES).

10. The collaborative control method for factory equipment based on swarm intelligence according to claim 1, characterized in that: This method is applicable to production systems in the ceramics, lithium battery electrode, and powder metallurgy industries that include roller kiln production lines.

Images