Intelligent workshop production scheduling system and method

Abstract

The application relates to an intelligent workshop scheduling system and method, which comprises a workpiece intelligent agent group, a device intelligent agent group, a scheduling intelligent agent and a scheduling decision support library. The workpiece intelligent agent group is a mapping of workpiece-related resources in an intelligent workshop, the device intelligent agent group is a mapping of device-related resources in the intelligent workshop, the scheduling intelligent agent reads a current workshop production state from the workpiece intelligent agent group and the device intelligent agent group, and provides decision support for a scheduling process, the scheduling decision support library stores and provides a scheduling decision model and a prediction model required by the scheduling intelligent agent, and the workpiece intelligent agent group and the device intelligent agent group determine a final production plan through negotiation interaction. Compared with the prior art, the workpiece intelligent agent and the device intelligent agent realize bidirectional selection through an interaction process, can complete a process from order reading, device configuration to production plan implementation, and greatly improve the autonomy of the whole scheduling process.

Description

Technical Field

[0001] This invention belongs to the field of intelligent manufacturing technology and relates to an intelligent workshop production scheduling system and method. Background Technology

[0002] Shop floor scheduling has always been a crucial aspect of manufacturing, and efficient scheduling can significantly contribute to improving production efficiency and protecting corporate interests. Traditional production scheduling research has largely focused on production scheduling and optimization in deterministic environments. However, with the widespread penetration of new-generation information technologies such as intelligent equipment, the increasing variety of products, and the growing complexity of manufacturing systems, intelligent workshops have become one of the future trends in the manufacturing field. Compared to traditional workshops, the complexity of production scheduling problems in intelligent workshops has increased dramatically, rendering traditional production scheduling methods inadequate. As a research hotspot in the field of artificial intelligence, multi-agent systems (MAS) are unified systems composed of multiple interacting agents. The effective collaboration among these agents to complete tasks provides feasible technical support for solving complex and ever-changing problems and realizing intelligent manufacturing systems. MAS systems can accomplish complex production tasks through the interactive collaboration of various agents, which is beneficial for improving the flexibility and reconfigurability of manufacturing systems.

[0003] A search of Chinese Patent Publication No. CN106527373A reveals a workshop autonomous scheduling system based on multiple agents. This system includes a production information database, an equipment information database, a workpiece agent group, an equipment agent group, and a logistics agent. The scheduling process of this system involves the workpiece and equipment agents simply reading existing scheduling strategies, determining their feasibility, and executing them. It lacks a negotiation process between the equipment and workpiece agents. Regarding scheduling timing, the scheduling time in this patent is before the overall processing, meaning the scheduling strategy is fixed. Regarding scheduling content, the content is the processing task, i.e., the time and equipment where the workpiece should be during processing. Regarding the negotiation process, this patent does not involve the selection of both equipment and workpiece. In summary, this prior art suffers from low scheduling strategy flexibility and low autonomy in the scheduling process. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent workshop production scheduling system and method to overcome the problems of low scheduling strategy flexibility and low autonomy in the scheduling process.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] A smart workshop production scheduling system includes a workpiece intelligent agent group, an equipment intelligent agent group, and a scheduling intelligent agent.

[0007] The workpiece intelligent agent group includes a workpiece management intelligent agent and multiple workpiece intelligent agents. The workpiece management intelligent agent reads production information from the intelligent workshop and is used to create, manage, and destroy workpiece intelligent agents. The workpiece intelligent agents are used to send status information and priority sorting requests, receive workpiece priority information, send workpiece status, workpiece information, and processing requests, and enter a two-way selection phase.

[0008] The equipment intelligence group includes an equipment management intelligence and multiple equipment intelligences. The equipment management intelligence reads equipment information from the smart workshop and is used to create, manage, and destroy equipment intelligences, as well as receive processing information and optimal equipment combination information. The equipment intelligences receive information about workpieces to be processed, place them in a waiting queue, and send equipment performance-related information and equipment combination change requests. Based on the received workpiece status, workpiece information, and processing requests, the equipment intelligences enter the two-way selection phase to determine the feasibility of processing.

[0009] The scheduling agent receives status information and priority sorting requests, determines whether the workpiece corresponding to the workpiece agent is a workpiece to be processed, and if so, sends processing information; otherwise, it calls the scheduling decision model to obtain and send workpiece priority information, and calls the prediction model to obtain and send optimal equipment combination information.

[0010] Furthermore, during the bidirectional selection phase, if the workpiece intelligent agent receives a processing signal, it updates its own status to "processing"; if it receives a waiting signal, it updates its own status to "awaiting processing"; and if it receives a reselection signal, it selects the next activated currently processable device.

[0011] During the bidirectional selection phase, the device intelligence agent performs a processing feasibility assessment. If it is in an idle state, it selects the highest priority workpiece intelligence agent that made the request for processing, sends a processing signal, and places the remaining workpiece intelligence agents into a waiting queue. If the device intelligence agent is in a processing state but the waiting queue is not full, it selects a higher priority workpiece intelligence agent to join the waiting queue according to the capacity of the waiting queue, sends a waiting signal, and sends a reselection signal to other workpiece intelligence agents. If the device intelligence agent is in a processing state and the waiting queue is full, it sends a reselection signal.

[0012] Furthermore, the intelligent workshop production scheduling system also includes a scheduling decision support library, which stores the scheduling decision models and prediction models that the scheduling agent needs to call when making decisions, and receives data generated by the workpiece agent group and the equipment agent group, and the scheduling decision models and prediction models are continuously updated based on the data.

[0013] Furthermore, the production information includes production content information, production constraint information, and real-time product information.

[0014] The production information includes order information, product types, processing quantities, and work-in-process quantities.

[0015] The production constraint information includes material feeding time and submission time.

[0016] The real-time product information includes the current process and the remaining processing time for the current process.

[0017] Furthermore, the equipment information includes equipment configuration information, equipment production information, and equipment real-time status information.

[0018] The equipment configuration information includes the equipment name, equipment type, and the number of workpieces that can be processed simultaneously.

[0019] The equipment production information includes equipment processing time and process name.

[0020] The real-time status information of the device includes activation status, fault status, and remaining repair time.

[0021] Furthermore, the multiple device intelligent agents are controlled by the device management intelligent agent, forming a centralized organizational structure.

[0022] Furthermore, the multiple workpiece intelligent agents are controlled by the workpiece management intelligent agent, forming a centralized organizational structure.

[0023] Furthermore, the device intelligence agent and the workpiece intelligence agent are organized in a distributed structure.

[0024] A method for a smart workshop production scheduling system includes the following steps:

[0025] Step S1: The workpiece intelligent agent obtains production information and generates a workpiece intelligent agent corresponding to the workpiece being processed based on the production information. The equipment management intelligent agent obtains equipment information and activates the equipment intelligent agent corresponding to the physical equipment based on the equipment information.

[0026] Step S2: The scheduling agent receives the status information and priority sorting request sent by the workpiece agent. Based on the status information, it determines whether the workpiece corresponding to the workpiece agent is a workpiece to be processed. If so, it sends processing information; otherwise, it requests the scheduling decision model from the decision support library to obtain and send the workpiece priority information.

[0027] Step S3: The intelligent device places the workpiece to be processed into a waiting queue according to the received processing information, and sends equipment performance information and equipment combination change requests in real time.

[0028] Step S4: The scheduling agent requests the invocation of the prediction model from the decision support library, and obtains and sends the optimal equipment combination information based on the equipment performance information and the equipment combination change request.

[0029] Step S5: The device management agent sends an activation or deactivation signal based on the optimal device combination information, and the device agent completes the activation or deactivation based on the activation or deactivation signal.

[0030] Step S6: The workpiece intelligent agent interacts with the activated device intelligent agent based on the workpiece priority information to determine and execute the plan.

[0031] Furthermore, the interaction process includes the following steps:

[0032] Step S61: The workpiece agent selects an activated device agent and sends the workpiece status, workpiece information, and processing request to it.

[0033] Step S62: Based on the workpiece status, workpiece information and processing request, the device agent performs a processing feasibility judgment. If it is in an idle state, proceed to step S63; if it is in a processing state but the waiting queue is not full, proceed to step S64; if it is in a processing state and the waiting queue is full, proceed to step S65.

[0034] Step S63: The device agent selects the highest priority workpiece agent that issued the request for processing, sends a processing signal, and puts the remaining workpiece agents into the waiting queue. The workpiece agents update their status to processing.

[0035] Step S64: The device agent selects the workpiece agent with higher priority to join the waiting queue according to the waiting queue capacity, sends a waiting signal, and sends a reselection signal to the remaining workpiece agents. The workpiece agent updates its own status to be ready for processing.

[0036] Step S65: The device agent sends a reselection signal to the workpiece agent, then returns to step S61.

[0037] When the device's intelligent agent malfunctions, the currently processed workpiece is removed from the first position of the waiting queue and processing continues after the device malfunction is repaired.

[0038] Compared with the prior art, the present invention has the following characteristics:

[0039] 1. This invention divides the on-site resources of the smart workshop into workpiece intelligent agent groups and equipment intelligent agent groups. Corresponding intelligent agents are generated for each individual workpiece and each individual device. The workpiece intelligent agents and equipment intelligent agents can achieve bidirectional selection through an interactive process, which can complete the process from order reading and equipment configuration to production plan implementation, thereby greatly improving the autonomy of the entire scheduling process.

[0040] 2. Within the intelligent agent group of the equipment and the intelligent agent group of the workpiece in this invention, the intelligent agents have a centralized organizational structure, and all intelligent agents are controlled by the management intelligent agent, thereby improving the adaptability of the production scheduling method.

[0041] 3. The intelligent agents of the equipment and the intelligent agents of the workpiece in this invention have a distributed organizational structure. The production plan is obtained through negotiation between the intelligent agents. There is no centralized control. Therefore, it has high robustness and scalability and will not cause global paralysis due to local failure.

[0042] 4. The intelligent agent of this invention covers all processed workpieces and equipment, thus it can detect the production status and equipment operation, respond promptly when the system fails, and adaptively adjust according to changes in the workshop environment, further improving the robustness of the system. Attached Figure Description

[0043] Figure 1 This is a system structure diagram of the present invention;

[0044] Figure 2 This is a flowchart of a multi-agent-based intelligent workshop scheduling method provided in an embodiment of the present invention. Detailed Implementation

[0045] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. These embodiments are based on the technical solution of the present invention and provide detailed implementation methods and specific operating procedures. However, the scope of protection of the present invention is not limited to the following embodiments.

[0046] Example:

[0047] This invention provides an intelligent workshop production scheduling method and system, which divides the resources involved in the workshop into workpiece intelligent agent groups, equipment intelligent agent groups, and scheduling intelligent agents. Through the mutual coordination and cooperation among the workpiece intelligent agent groups, equipment intelligent agent groups, and scheduling intelligent agents, the entire process from production order input to product output is completed.

[0048] like Figure 1 As shown, the intelligent workshop production scheduling system of the present invention mainly includes a workpiece intelligent agent group, an equipment intelligent agent group, a scheduling intelligent agent, and a scheduling decision support library.

[0049] The workpiece intelligent agent group is configured to retrieve production content information, production constraints, and real-time product information from the intelligent workshop via a MySQL database, and send the retrieved workpiece information to the scheduling intelligent agent for priority sorting, and further negotiate and interact with the equipment intelligent agent group. The workpiece intelligent agent group includes a workpiece management intelligent agent and a workpiece intelligent agent. The production content information includes order information, product type, processing quantity, and work-in-process quantity. The production constraints include material feeding time and submission time. The real-time product information includes the current process and the remaining processing time for the current process.

[0050] The equipment intelligence group is configured to retrieve equipment configuration information, equipment production information, and real-time equipment status information sent by the smart workshop through a MySQL database. It then sends the equipment status information and calculated equipment utilization and other performance information to the scheduling intelligence agent, which calculates the optimal equipment combination and further negotiates and interacts with the workpiece intelligence agent. The equipment intelligence group includes an equipment management intelligence agent and equipment intelligence agents. The equipment configuration information includes equipment name, equipment type, and the number of workpieces that can be processed simultaneously. The equipment production information includes equipment processing time and process name. The real-time equipment status information includes activation status, fault status, and remaining repair time.

[0051] The scheduling agent is configured to receive scheduling requests from the workpiece agent group and the equipment agent group, and, based on the received workpiece and equipment information, retrieve appropriate scheduling knowledge from the scheduling decision support library to return the corresponding workpiece priority and optimal equipment combination to the workpiece agent group and the equipment agent group. Furthermore, after processing is completed, the scheduling agent can collect processing data accumulated in the smart workshop (such as the negotiation results between the workpiece agent group and the scheduling agent group) and production performance information (such as equipment utilization rate, average waiting time in the processing area, average production cycle, etc.), and update the scheduling decision model and prediction model in the scheduling decision support library accordingly.

[0052] The scheduling decision support library is configured to store decision support information, including metaheuristic rules, scheduling decision models, and prediction models. Metaheuristic rules include, but are not limited to, rules such as First-In-First-Out (FIFO) and shortest processing cycle. Scheduling decision models include, but are not limited to, pre-trained machine learning models that can output weights of combined scheduling rules based on production status and optimization objectives. Predictive models include, but are not limited to, pre-trained machine learning models that can predict the workshop production performance and the feasibility of the production plan when activating different equipment, based on production status and equipment status. Furthermore, the models within the scheduling decision support library can be continuously updated based on the processing data and production performance collected by the scheduling agent.

[0053] In this embodiment, the workpiece management intelligent agent and the equipment management intelligent agent are located in the workshop control center. Both agents have data storage, wired and wireless communication permissions. The workpiece intelligent agent is fixed to the workpiece by adsorption or clamping, and possesses computing power, data storage functions, and the authority to monitor the workpiece status. The equipment intelligent agent directly reads data from within the intelligent device and executes the scheduling commands of the equipment intelligent agent through the control module of the intelligent device. The intelligent agents communicate directly with each other via the TCP / IP protocol.

[0054] Specifically, each workpiece intelligent agent corresponds to a workpiece being processed. The workpiece intelligent agent monitors the current step, priority, processing status, location, pending processing status, and completion status of the workpiece, and sends this information to the scheduling intelligent agent, the workpiece management intelligent agent, and the device intelligent agent. Each device intelligent agent corresponds to a device. The device intelligent agent monitors the current fault status, processing status, processed workpieces, and the number of workpieces in the buffer zone of the device, and sends the status data of the processing device to the scheduling intelligent agent, the device management intelligent agent, and the workpiece intelligent agent.

[0055] Specifically, the workpiece management intelligent agent is responsible for managing all workpiece intelligent agents, directly receiving order information and matching production information from the workshop control center, and generating workpiece intelligent agents based on the order information and work-in-process information. Simultaneously, the workpiece management intelligent agent can monitor the real-time status of workpieces in the workshop by receiving data from the workpiece intelligent agents. While monitoring the status of workpieces in the workshop, the workpiece management intelligent agent can query the current production progress, workpiece production cycle, etc., in the workshop.

[0056] Specifically, the equipment management intelligent agent is responsible for managing all equipment intelligent agents. It reads equipment configuration information (such as equipment name, equipment type, fault information, etc.) from the workshop control center, generates an equipment status information table, and activates the equipment intelligent agents corresponding one-to-one with the configured equipment. The equipment management intelligent agent can receive data uploaded by the equipment intelligent agents, thereby monitoring the workshop status in real time and promptly reporting equipment status information to workshop management personnel. The equipment management intelligent agent receives the optimal equipment combination sent by the scheduling intelligent agent, and sends activation or deactivation signals to the corresponding equipment intelligent agents based on the combination. The equipment intelligent agents receive the signals and perform power-on or power-off operations on the corresponding equipment.

[0057] On the other hand, the present invention provides an intelligent workshop scheduling method based on the system. For example... Figure 2 As shown, the method includes the following steps:

[0058] Step S1: The workpiece management agent reads order information, product information, and process information from the smart workshop and generates workpiece agents corresponding one-to-one with the workpieces being processed. In this step, the workpiece management agent reads the order information, generates a product information table required for the order based on the product types required in the order, and supplements the product information table with information such as product delivery dates. Simultaneously, the workpiece management agent obtains production constraints such as processable equipment from the product process information and supplements the product information table with information such as the total number of processes and product processing time. After the product information table is complete, the workpiece management agent generates workpiece agents corresponding one-to-one with the actual workpieces based on the product information table, and sends the information of each workpiece in the table to the corresponding workpiece agent as initialization information.

[0059] The equipment management agent reads physical equipment information from the smart workshop and generates a corresponding equipment agent for each physical device. In this step, the equipment management agent reads information about all physical equipment in the workshop and builds an equipment status information table containing all activated devices based on the equipment activation data in the smart workshop configuration information. Simultaneously, the equipment management agent supplements the equipment status information table with information such as equipment repair time using the physical equipment information. After the equipment status information table is supplemented, the equipment management agent activates the corresponding equipment agent for each physical device.

[0060] Step S2: The workpiece agent determines its own state. If the workpiece is in an unprocessed state and has not entered the equipment's waiting queue, it sends its own state information and constraint information to the scheduling agent and makes a priority sorting request.

[0061] After receiving the workpiece information and priority ranking request from the workpiece agent, the scheduling agent determines the current workpiece's pending processing status. If the current workpiece is pending processing, the scheduling agent sends information such as the product processing time from the workpiece status to the corresponding device agent; if the current workpiece is not pending processing, the scheduling agent selects a scheduling decision model from the scheduling decision support library, obtains the workpiece's priority ranking based on its delivery date and other status data, and sends it to the workpiece agent.

[0062] Step S3: After receiving the workpiece information, the intelligent agent adds the workpiece to the last position of the current device waiting queue. If the device waiting queue is full, it is expanded.

[0063] Once activated, the device agent will send its own performance data, such as device utilization and queue length, to the scheduling agent in real time, and will also send a device combination change request to the scheduling agent.

[0064] Step S4: After receiving the performance data and equipment combination change request sent by the equipment agent, the scheduling agent selects a feasible equipment combination based on the product process information, selects a prediction model from the scheduling decision support library, predicts the feasibility of the production plan and the production performance results under different equipment combinations, and selects the optimal equipment combination to send to the equipment management agent.

[0065] Step S5: The device management agent receives the optimal device combination from the scheduling agent and compares the combination with the active devices in the current device status information table. If there are inactive devices, the device management agent activates the device agent and updates the device status information table; if there are devices that need to be deregistered, the device management agent queries the device status. When the device is idle, the device management agent stores the device's performance data and historical negotiation information, sends a deregistration command, and updates the device status information table. The device agent receives signals and performs power-on or power-off operations on the corresponding devices.

[0066] Step S6: The workpiece intelligent agent negotiates with the activated equipment intelligent agent based on the received priority data to determine and execute the production plan.

[0067] Specifically, step S6 further includes the following steps:

[0068] Step S61: The workpiece agent selects an activated device agent and sends the workpiece status, workpiece information, and processing request to it.

[0069] Step S62: Based on the workpiece status, workpiece information, and processing request, if the device agent is in an idle state, proceed to step S63; if it is in a processing state but the waiting queue is not full, proceed to step S64; if it is in a processing state and the waiting queue is full, proceed to step S65.

[0070] Step S63: The device agent selects the highest priority workpiece agent that issued the request for processing, sends a processing signal, and puts the remaining workpiece agents into the waiting queue. The workpiece agents update their status to processing.

[0071] Step S64: The device agent selects the workpiece agent with higher priority to join the waiting queue according to the waiting queue capacity, sends a waiting signal, and sends a reselection signal to the remaining workpiece agents. The workpiece agent updates its own status to be ready for processing.

[0072] Step S65: The device agent sends a reselection signal to the workpiece agent, and returns to step S61.

[0073] Specifically, when the device intelligence agent malfunctions, the currently processed workpiece is removed from the first position of the waiting queue and processing continues after the device malfunction is repaired.

[0074] In this embodiment, after the workpiece is processed, the workpiece management agent reads the performance data stored in the workpiece agent, such as the completion time, and then cancels the workpiece agent.

[0075] In this embodiment, after production is completed, the activated equipment agent sends equipment performance data, such as equipment utilization rate and average waiting queue length, along with negotiation information to the scheduling agent. The workpiece management agent sends stored performance data, such as processing cycle, on-time delivery rate, and number of movement steps, along with negotiation information to the scheduling agent. The scheduling agent integrates the information and sends it to the scheduling decision support library for model updates. Based on the performance data, the scheduling decision support library adjusts the evaluation score of the negotiation process between the workpiece agent and the equipment agent, thereby updating the scheduling decision model and prediction model selected by the scheduling agent in this instance.

[0076] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0077] The above description of the embodiments is provided to enable those skilled in the art to understand and use the invention. It will be apparent to those skilled in the art that various modifications can be made to these embodiments, and the general principles described herein can be applied to other embodiments without inventive effort. Therefore, the present invention is not limited to the above embodiments, and any improvements and modifications made by those skilled in the art based on the disclosure of the present invention without departing from the scope of the invention should be within the protection scope of the present invention.

Claims

1. An intelligent shop floor production scheduling system, characterized by, The workpiece agent group, the device agent group and the scheduling agent, The workpiece agent group, the device agent group and the scheduling agent, The device agent group, including a device management agent and a plurality of device agents, the device management agent reads device information from the intelligent workshop, and is used for creating, managing and destroying the device agents, receiving processing information and optimal device combination information; the device agent is used for receiving information of workpieces to be processed, putting the workpieces to be processed into a waiting queue, and sending device performance related information and device combination change request, and the device agent enters the bidirectional selection stage to judge processing feasibility according to the received workpiece state, workpiece information and processing request, The scheduling agent receives state information and priority sorting request, judges whether the workpiece corresponding to the workpiece agent is a workpiece to be processed, if yes, sends processing information, if not, calls a scheduling decision model to obtain and send workpiece priority information, calls a prediction model to obtain and send optimal device combination information; The device agent and the workpiece agent are in a distributed organization structure. The intelligent workshop production scheduling system further includes a scheduling decision support library, which is used for storing a scheduling decision model and a prediction model that need to be called by the scheduling agent when making decisions, and receiving data generated by the workpiece agent group and the device agent group, and the scheduling decision model and the prediction model are continuously updated according to the data.

2. The intelligent shop floor production scheduling system of claim 1, wherein, In the bidirectional selection stage, if the workpiece agent receives a processing signal, the workpiece agent updates its own state to processing, if the workpiece agent receives a waiting signal, the workpiece agent updates its own state to be processed, and if the workpiece agent receives a reselection signal, the workpiece agent selects the next activated current processable device, In the bidirectional selection stage, the device agent judges processing feasibility, if the device agent is in an idle state, the device agent selects the highest priority workpiece agent to process, sends a processing signal, and puts the remaining workpiece agents into a waiting queue, if the device agent is in a processing state but the waiting queue is not full, the device agent selects a workpiece agent with a higher priority to join the waiting queue according to the capacity of the waiting queue, sends a waiting signal, and sends a reselection signal to other workpiece agents, and if the device agent is in a processing state and the waiting queue is full, the device agent sends a reselection signal.

3. The intelligent shop floor production scheduling system of claim 1, wherein, The production information includes production content information, production constraint information and product real-time information, The production content information includes order information, production product types, processing quantity and work-in-process quantity, The production constraint information includes feeding time and submission time, The product real-time information includes a current process and a remaining processing time of the current process.

4. The intelligent shop floor production scheduling system of claim 1, wherein, The device information includes device configuration information, device production information, and device real-time state information, The device configuration information includes a device name, a device type, and a number of workpieces that can be processed simultaneously, The device production information includes a device processing time and a process name, The device real-time state information includes an active state, a fault state, and a remaining repair time.

5. A scheduling method based on the intelligent workshop production scheduling system according to claim 1, characterized in that, The method comprises the following steps: In step S1, a workpiece management intelligent agent acquires production information, generates a workpiece intelligent agent corresponding to a workpiece to be processed according to the production information, and a device management intelligent agent acquires device information and activates a device intelligent agent corresponding to a physical device according to the device information; In step S2, a scheduling intelligent agent receives state information and a priority sorting request sent by the workpiece intelligent agent, determines whether the workpiece corresponding to the workpiece intelligent agent is a workpiece to be processed according to the state information, sends processing information if the workpiece is a workpiece to be processed, requests a decision support library to call a scheduling decision model if the workpiece is not a workpiece to be processed, obtains workpiece priority information, and sends the workpiece priority information; In step S3, a device intelligent agent puts a workpiece to be processed into a waiting queue according to the received processing information, and sends device performance information and a device combination change request in real time; In step S4, the scheduling intelligent agent requests the decision support library to call a prediction model, obtains optimal device combination information according to the device performance information and the device combination change request, and sends the optimal device combination information; In step S5, a device management intelligent agent sends an activation or deregistration signal according to the optimal device combination information, and a device intelligent agent completes activation or deregistration according to the activation or deregistration signal; In step S6, a workpiece intelligent agent enters an interaction process with an activated device intelligent agent according to the workpiece priority information, determines a plan, and executes the plan.

6. The method of claim 5, wherein, The interaction process comprises the following steps: In step S61, a workpiece intelligent agent selects an activated device intelligent agent and sends workpiece state, workpiece information, and a processing request to the device intelligent agent; In step S62, the device intelligent agent performs a processing feasibility judgment according to the workpiece state, the workpiece information, and the processing request, enters step S63 if the device intelligent agent is in an idle state, enters step S64 if the device intelligent agent is in a processing state but a waiting queue is not full, or enters step S65 if the device intelligent agent is in a processing state and the waiting queue is full; In step S63, the device intelligent agent selects a workpiece intelligent agent with the highest priority to process, sends a processing signal, and puts remaining workpiece intelligent agents into the waiting queue, and a workpiece intelligent agent updates its own state to processing; In step S64, the device intelligent agent selects workpiece intelligent agents with higher priorities to join the waiting queue according to the capacity of the waiting queue, sends a waiting signal, and sends a reselection signal to the remaining workpiece intelligent agents, and a workpiece intelligent agent updates its own state to being processed; In step S65, the device intelligent agent sends a reselection signal to a workpiece intelligent agent, and returns to step S61. When the device intelligent agent fails, a workpiece currently being processed exits the first position of the waiting queue, and continues to be processed after the device failure is repaired.

Images