Toy packaging workshop collaborative manufacturing method based on industrial internet

Abstract

The invention discloses a toy packaging workshop collaborative manufacturing method based on the industrial internet. The method comprises the steps that a cloud intelligent platform obtains production real-time data of equipment in a production workshop; the cloud intelligent platform establishes a mathematical model of dynamic collaborative manufacturing according to the production real-time data; the cloud intelligent platform sends a task to idle intelligent equipment of the equipment layer through the task layer, or sends the task to the decision layer and stores the task in a task pool; querying information of the device pool through a decision-making layer, and sending a task to the intelligent device corresponding to the idle device in the device pool; receiving the task sent by the task layer or querying the task in the task pool through the equipment layer; and the intelligent equipment receives the tasks directly sent by the task layer or queries the tasks in the task pool, locks the tasks and adds the tasks into a task queue of the intelligent equipment. According to the invention, the equipment and the cloud intelligent platform are combined to form a system, the equipment and data are combined together, and the network is cooperated for updating, so that a major breakthrough is realized in the packaging industry.

Description

Technical Field

[0001] The present invention relates to the field of industrial Internet intelligent manufacturing technology, and in particular to a toy packaging workshop collaborative manufacturing method based on industrial Internet. Background Art

[0002] At present, toy packaging is mainly complex discrete manufacturing. Complex discrete manufacturing workshops mainly adopt a top-down, centralized and planned workshop operation mode, that is, the Manufacturing Execution System (MES) is used to transmit and interact information between the upper-level Enterprise Resource Planning (ERP) system and the underlying industrial control system (or production site), and the MES is used to centrally and uniformly control the time schedule, product quality, cost, manufacturing resources, energy consumption and other factors in the manufacturing process.

[0003] However, there are still many shortcomings in complex discrete manufacturing workshops: (a) The equipment of each unit of the production and packaging production line only communicates with the master control system. There is a lack of data sharing mechanism between units, and there are "information islands". Production control relies solely on the master control system, and the unit equipment lacks autonomous control capabilities. When the master control is offline, the production line cannot continue processing; (b) The product structure is generally complex, with many parts and components under the product, and a large number of processes for products and parts. For some products, there is a situation of "development and production at the same time". At the same time, the concurrent execution of multiple tasks leads to a more prominent phenomenon of competition for key manufacturing resources; (c) There are many uncertain factors and emergencies in the production process (such as emergency orders, material shortages, equipment downtime, personnel leaving the job, tooling damage, quality fluctuations, etc.), which have a great impact on the normal execution of production tasks. Summary of the invention

[0004] The technical problem to be solved by the embodiments of the present invention is to provide a collaborative manufacturing method for a toy packaging workshop based on the industrial Internet, which can solve the defects that the traditional toy packaging production line adopts a fixed process production mode according to the material category or packaging requirements, the production line has low versatility, and the data sharing level between the equipment in each module of the production line is low, there are "information islands", and the production line equipment lacks the ability of autonomous control and collaborative manufacturing.

[0005] In order to solve the above technical problems, an embodiment of the present invention provides a toy packaging workshop collaborative manufacturing method based on the industrial Internet, comprising the following steps: S1: The cloud intelligence platform obtains real-time production data of equipment in the production workshop through the edge server; S2: The cloud intelligent platform establishes a mathematical model of dynamic collaborative manufacturing according to the real-time production data; S3: The cloud intelligent platform sends tasks to idle intelligent devices in the device layer through the task layer, or sends tasks to the decision layer and stores them in the task pool; queries the information of the device pool through the decision layer, and sends tasks to the intelligent devices corresponding to the idle devices in the device pool; receives tasks sent by the task layer through the device layer or queries tasks in the task pool; S4: The smart device receives the task sent directly by the task layer or queries the task in the task pool, locks the task and adds it to its own task queue.

[0006] Among them, S3 also includes the step of the decision layer querying the information in the task pool and the equipment pool, and arbitrating according to the optimization strategy. The decision layer determines whether the optimization strategy is global optimization or local optimization based on the real-time information of the task layer and the equipment layer. After the decision layer completes the arbitration, it generates a queue diagram to guide workshop production.

[0007] Among them, in S3, the task layer calculates the priority of the task in advance. When the priority of the task is higher than the predetermined value, the task layer sends the task to the smart device; when the priority of the task is lower than the predetermined value, the task layer stores the task in the task pool.

[0008] Among them, the local optimization includes: the decision-making layer selects the interaction strategy and arbitrates the combination of tasks and equipment based on the predetermined target benefit function, arbitrates according to the order grabbing situation of smart devices and manufacturing tasks, calculates the target benefit function value of each candidate task-equipment based on real-time data, assigns tasks to the optimal smart device to form the final task-equipment combination, and generates a real-time scheduling diagram to guide the production process of the physical workshop.

[0009] Among them, the global optimization includes: based on the real-time data of the current task progress and equipment status, and based on the existing pre-scheduling results, within a predetermined time range, all pending manufacturing tasks and smart devices are globally rescheduled in the cloud, and the optimal task-equipment combination is selected from the candidate pool based on the global scheduling results; at the same time, at the beginning of each scheduling cycle, the manufacturing tasks and smart devices are periodically globally rescheduled based on the prediction results and real-time data in the cloud, and the scheduling results are used as a reference for the arbitration team to select the optimal task-equipment combination.

[0010] The mathematical model is

[0011]

[0012]

[0013]

[0014] Where, T: total duration of production task; J: equipment station set in workshop; P: production task set; A: production capacity of each equipment; ST: production plan start time; ET: production plan completion time; D: other abnormal situation parameters that occur in the actual production process.

[0015] Implementing the embodiments of the present invention has the following beneficial effects: ① The present invention collaborates with the device and the intelligent cloud platform to generate an intelligent data system and bring into play its application value.

[0016] ② The use of intelligent data can optimize the system operation function, reduce the operation cost, and enable the maintenance and management of the entire equipment system. For example, all equipment and components are connected to the Internet, the working status of the equipment can be monitored at any time, and the parts can be delivered to the specified location as quickly as possible, reducing inventory, reducing maintenance costs, and promoting the improvement of equipment operation stability.

[0017] ③The data generated by the operation of each device or machine can form a digital system, allowing machines and equipment to understand their operation through self-learning, which cannot be done on a single device. For example, by collecting a large amount of data information such as the location and status of parts, the performance parameter information of the equipment under various requirements can be accurately grasped. Combining many machines into a system to form a large amount of intelligent data information can achieve the expansion and learning of the network system and make the system more intelligent. As long as the intelligent device obtains all the information, the intelligent system can be used to develop comprehensive business value and make accurate decisions. The combination of equipment and data, and the collaborative update of the network have enabled the packaging industry to achieve major breakthroughs. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a schematic diagram of the collaborative manufacturing application framework of the present invention; Figure 2 It is a schematic diagram of the collaborative manufacturing method of the present invention. DETAILED DESCRIPTION

[0019] In order to make the objectives, technical solutions and advantages of the present invention more clear, the present invention will be further described in detail below with reference to the accompanying drawings.

[0020] Combination Figure 1 , Figure 2 As shown, a toy packaging workshop collaborative manufacturing method based on industrial Internet according to an embodiment of the present invention is implemented through the following steps.

[0021] 1. The cloud server collects real-time production data, which is transmitted to the cloud server through the edge server. The cloud server conducts collaborative analysis and sends control instructions to the on-site workshop equipment.

[0022] In the above steps, the edge server is an intelligent system deployed on-site, which supports wired and wireless transmission. It can replace dozens of industrial protocols such as Modbus TCP with MQTT to realize real-time data collection of on-site equipment and integrate it into the cloud intelligent platform. It also supports the downlink of cloud control instructions and calculations that require real-time processing.

[0023] The cloud intelligence platform includes private servers, industrial Internet networks, and business development application layers.

[0024] 2. The cloud intelligent platform analyzes and issues control instructions to realize collaborative manufacturing in the workshop. The specific collaborative methods are as follows: Figure 2 shown.

[0025] The cloud intelligence platform includes task layer, decision layer and equipment layer.

[0026] The task layer is for service demanders; the decision layer is for combined arbitration; and the equipment layer is for service providers.

[0027] According to the main influencing factors in the workshop manufacturing process, a mathematical model of dynamic collaborative manufacturing is established

[0028]

[0029]

[0030]

[0031] In the formula, T: total duration of production task; J: workshop equipment station set; P: production task set; A: production capacity of each equipment; ST: production plan start time; ET: production plan completion time; D: other abnormal situation parameters in the actual production process. For example, changes in various parameters such as production process / procedure, number of operators, business level, operation time, equipment status, etc. can affect the results. i represents the number of the production task or a specific task, and j usually represents the number of the equipment or station.

[0032] 3. The task layer sends tasks to the idle intelligent devices in the device layer; or, the task layer sends tasks to the decision layer and stores them in the task pool.

[0033] The total production task is ,in It is a positive integer representing the number of production tasks.

[0034] Represents the production tasks of each intelligent device corresponding to the total production task, where: j ∈ J Represents total production task The number of intelligent equipment stations required in the process. Therefore, the station-level production task M is expressed as:

[0035]

[0036] Assume that the processing amount of each production task equipment is Q . Represents the total production task The corresponding The production volume of a smart device. Therefore, the station-level production task M is expressed as:

[0037]

[0038] According to the equipment production tasks obtained J And the production volume of each task corresponding to the intelligent equipment station Q , further combine the production tasks of each equipment and the corresponding production volume, and finally get the total production task P The disassembly results are expressed by the Hadamard product:

[0039] The task layer calculates the priority of the task in advance. When the priority of the task is higher than a predetermined value, the task layer sends the task to the smart device; when the priority of the task is lower than the predetermined value, the task layer stores the task in the task pool.

[0040] In the above steps, the decision layer queries the information of the equipment pool and sends tasks to the smart devices corresponding to the idle devices in the equipment pool. The task information stored in the task pool includes: one or more of quality, quantity, priority, process route constraints, time constraints or target constraints.

[0041] Specifically, the task layer acts as a service demander. When production anomalies such as new order insertion, delivery date change, personnel change, untimely material delivery, and equipment failure occur, on the one hand, it can actively make service requests to the intelligent device layer, including quality, quantity, priority (such as urgent, important, non-urgent, etc.), process route constraints (i.e. process sequence constraints), time constraints (such as delivery date), target constraints (such as delivery date priority, processing time priority) and other requirements and constraints; on the other hand, when the task priority is high, such as tasks with priority above urgent, according to the real-time status of the intelligent device pool, it actively finds the currently idle intelligent devices that meet the service requirements, and directly sends service requests to the designated intelligent devices. The intelligent devices will give feedback to the task whether they agree to accept the order based on their own situation.

[0042] Before sending tasks, the task layer classifies them according to the urgency of the task. For example, urgent or higher-level tasks are classified into one category, and less-urgent tasks are classified into another category. When sending tasks less than urgent, they are sent to the task pool for storage. The tasks in the task pool are provided for the smart devices at the device layer to grab orders. When sending urgent or higher-level tasks, the task layer queries the information of each smart device in the device pool, filters out the smart devices that meet the service requirements and are currently idle, and sends the tasks directly to the idle smart devices; the idle devices determine whether to accept the order based on the current status. For example, when the current device is in an overhaul or maintenance state, it will not receive the task. When the current device is in a state where it can process the task, it will receive the task. When the device does not accept the order, the device returns the information that it will not process the task, and the task layer reselects an idle device to send the task.

[0043] 4. The device layer receives tasks sent by the task layer or queries tasks in the task pool. The smart device screens the received tasks or the queried tasks. The device layer adds the screened tasks to the pending queue. The smart device sends the pending queue to the decision layer and stores it in the device pool. The smart device screens the tasks in the task pool based on its own attributes, capabilities, status, load, target, and the operation of other smart devices. The smart device also predicts the downtime before it stops and releases the corresponding tasks during the downtime.

[0044] The process of predicting downtime at the equipment level is implemented in the following ways to ensure the continuity of task scheduling and the timeliness of task redistribution: Historical data analysis: Smart devices use their historical operating data, including previous downtime records, failure rates, maintenance cycles, and other information to predict the next possible downtime. By analyzing historical downtime data, the average operating time or failure interval of the equipment can be obtained to estimate future downtime periods.

[0045] Sensor monitoring and real-time data analysis: The sensors equipped on the equipment can monitor key performance parameters in real time, such as temperature, vibration, energy consumption, etc. If certain parameters of the equipment are abnormal or exceed the preset threshold (such as high temperature or severe vibration), the system will predict possible shutdown and issue a shutdown warning in advance.

[0046] Machine learning prediction model: Use machine learning algorithms (such as regression analysis, time series analysis, etc.) to analyze the historical fault data and real-time monitoring data of the equipment and establish a prediction model. The model improves the prediction accuracy of downtime through continuous learning and updating.

[0047] Maintenance plan and maintenance cycle: The equipment layer usually has regular maintenance and maintenance plans. According to the established maintenance cycle, the system can estimate the planned downtime of the equipment, thereby releasing tasks in advance. For unforeseen failures, the above sensor data and historical records can be combined to further improve the prediction accuracy.

[0048] As a service provider, after the service demand is released by the manufacturing task layer, the device layer will actively discover, automatically match and actively grab orders for service demands based on its own attributes, capabilities, status, load and goals (such as minimum energy consumption, minimum idle time, etc.), and combined with the actual operation of other smart devices; on the other hand, when the equipment fails or needs to be shut down for maintenance, the downtime is predicted, and the queue of tasks to be completed that have been allocated during the downtime is released, so that the manufacturing task layer can publish updated service demands to other available smart devices, or other smart devices can actively "grab" the released task queue after obtaining relevant information. For service requests directly issued from the task layer, the smart device determines whether to agree to accept the order based on its own situation, and feeds the result back to the task in real time.

[0049] The smallest equipment working unit in the manufacturing process is the workstation. The basic unit for multi-station collaboration is also set as the workstation. In the actual manufacturing process, there is one workstation for one process, multiple workstations for one process, and multiple processes and multiple workstations. The workstation set is a set of workstations divided according to the actual production process, production process, etc., which can be expressed as:

[0050] Among them, j represents the number of equipment or workstation, and J: workshop equipment and workstation set.

[0051] Combined with the analysis of uncertain situations, if there is a need to add or cancel workstations, the workstation set can be modified and maintained according to the actual situation to improve the flexible management of the workstation set.

[0052] There are two sources of tasks for smart devices. One is to passively receive tasks sent directly by the task layer. The specific processing process of this task is described in detail in the previous step. The other is that smart devices query the tasks in the task pool and grab orders based on their own attributes, capabilities, status, load and goals (such as minimum energy consumption, minimum idle time, etc.) and the actual operation of other smart devices. Grabbing an order means locking the task and adding it to its own task queue.

[0053] Assume that the production capacity of each device is , then it is expressed as ,

[0054] In the formula represents the equipment production capacity, then the production capacity of each equipment is expressed as

[0055] In the formula Represents the equipment production capacity coefficient. The equipment production capacity is an uncertain parameter and is related to the operation time t, equipment status s, etc. When one of the variables changes, the corresponding operation capacity coefficient changes accordingly, which is expressed as .

[0056] In the above steps, the decision layer queries the information in the task pool and the equipment pool, and arbitrates according to the optimization strategy. The decision layer determines whether the optimization strategy is global optimization or local optimization based on the real-time information of the task layer and the equipment layer. After the decision layer completes the arbitration, it generates a queue diagram to guide workshop production.

[0057] The local optimization includes: the decision layer selects an interaction strategy and arbitrates the combination of tasks and equipment according to a predetermined target benefit function.

[0058] Among them, global optimization includes: based on the real-time information of the task pool and equipment pool, global scheduling of all tasks and equipment to be completed is carried out through cloud computing.

[0059] As the arbitrator of task and equipment combinations, the decision-making layer first determines whether the arbitration optimization strategy is global optimization or local optimization based on pre-set judgment conditions and real-time on-site data.

[0060] The decision conditions are used to decide whether to choose a local optimization or global optimization strategy when allocating tasks. These conditions are based on real-time data in the production process, the urgency of the task, the current status of the equipment, the priority of the task, and the overall resource utilization.

[0061] Specifically, if it is local optimization, an interactive strategy (cooperative game or non-cooperative game) is selected under the premise of a given target benefit function, and arbitration is performed based on the order grabbing situation of smart devices and manufacturing tasks (after the smart device completes the order grabbing, it enters the candidate task-equipment pool). That is, the target benefit function value of each candidate task-equipment is calculated based on real-time data, and the task is assigned to the optimal smart device to form the final task-equipment combination, and a real-time scheduling diagram is generated to guide the production process of the physical workshop.

[0062] The objective benefit function is:

[0063] Among them, T: total duration of production tasks; J: workshop equipment station set; P: production task set; A: production capacity of each equipment; ST: production plan start time; ET: production plan completion time; D: other abnormal situation parameters that occur during the actual production process; T represents the total duration of production tasks; C represents the production cost (including equipment operation and maintenance costs); E represents the energy consumption of equipment during the production process; U represents equipment utilization; α, β, γ, δ are weight coefficients.

[0064] If it is global optimization, all pending manufacturing tasks and smart devices will be globally rescheduled in the cloud within a reasonable time frame based on real-time data such as current task progress and equipment status, based on existing pre-scheduling results, and the optimal task-equipment combination will be selected from the candidate pool based on the global scheduling results. At the same time, in order to improve the operating performance of the manufacturing system, at the beginning of each scheduling cycle (pre-set according to the actual situation of the manufacturing system), manufacturing tasks and smart devices will be periodically globally rescheduled based on the prediction results and real-time data in the cloud, and the scheduling results will be used as a reference for the arbitration team to select the optimal task-equipment combination.

[0065] The toy packaging collaborative manufacturing method of the present invention can not only perform local optimization and real-time control of the edge layer through equipment-task, equipment-equipment communication, interaction and autonomous collaborative decision-making, but also perform global optimization and accurate prediction through the high-performance computing power and decision-making ability of the cloud (the generated plan can be used to assist the edge layer in real-time decision-making), thereby maximizing the workshop production efficiency and manufacturing resource utilization, reducing production costs and energy consumption, and ultimately achieving the goal of optimizing workshop operation. In addition, by giving certain intelligence to equipment and tasks, when an abnormality occurs during the operation of the workshop, two-way real-time communication and interaction can be achieved between tasks and equipment, and between equipment and equipment, and autonomous collaborative decision-making can be made based on real-time information, so as to actively and in real time solve the workshop abnormality, and truly realize real-time control and active control for smart workshops. In addition, the real-time demand of manufacturing tasks, the real-time status and capabilities of intelligent devices are virtualized and service-oriented, and the autonomous collaborative decision-making of equipment-equipment, the autonomous matching and combination optimization of task-equipment are realized, and the cloud access and service-oriented manufacturing tasks and intelligent devices are realized at the same time, so as to seamlessly connect with the cloud manufacturing mode, greatly improving the system's outward scalability and collaborative ability.

[0066] The embodiments of the present invention have the following advantages: ① Aiming at the problems existing in the current operation mode of traditional toy packaging workshops with MES as the core, such as slow abnormal response, poor real-time decision-making, low system stability, etc., this invention proposes a new operation mode of toy packaging workshops—a collaborative manufacturing mode based on the industrial Internet, and proposes solutions to the concepts, system architecture, operation mechanism, key technologies and characteristics of autonomous collaborative manufacturing in workshops.

[0067] ② Based on the interconnection of things, with intelligence-intelligence collaboration as the core, and on the basis of giving full play to the real-time optimization of the edge layer and the global optimization of the cloud, through two-way real-time autonomous communication, interaction and decision-making between manufacturing tasks and intelligent devices, as well as autonomous collaborative decision-making between intelligent devices, event-driven autonomous matching and local combination optimization of edge layer manufacturing tasks and intelligent devices, as well as cycle-driven global optimization and accurate prediction in the cloud are realized, thereby improving the response speed of exceptions, reducing the frequency of exceptions and the negative impact of production uncertainty.

[0068] ③ The concept of collaborative manufacturing system for toy packaging workshops is proposed. Based on the analysis of workshop production mode and production plan related factors, the collaborative dynamic adjustment mechanism is studied, and a mathematical model of collaborative manufacturing system is established. Dynamic collaborative adjustment can be performed when the workshop is facing abnormal situations during operation.

[0069] The above disclosure is only a preferred embodiment of the present invention, which certainly cannot be used to limit the scope of the present invention. Therefore, equivalent changes made according to the claims of the present invention are still within the scope of the present invention.

Claims

1. A toy packaging workshop collaborative manufacturing method based on industrial Internet, characterized in that: The following steps are involved: S1: The cloud intelligence platform obtains real-time production data of equipment in the production workshop through the edge server; S2: The cloud intelligent platform establishes a mathematical model of dynamic collaborative manufacturing according to the real-time production data; S3: The cloud intelligent platform sends tasks to idle intelligent devices in the device layer through the task layer, or sends tasks to the decision layer and stores them in the task pool; queries the information of the device pool through the decision layer, and sends tasks to the intelligent devices corresponding to the idle devices in the device pool; receives tasks sent by the task layer through the device layer or queries tasks in the task pool; S4: The smart device receives the task sent directly by the task layer or queries the task in the task pool, locks the task and adds it to its own task queue.

2. The toy packaging workshop collaborative manufacturing method based on the industrial Internet according to claim 1 is characterized in that: S3 also includes the step of the decision layer querying the information in the task pool and the equipment pool, and arbitrating according to the optimization strategy. The decision layer determines whether the optimization strategy is global optimization or local optimization based on the real-time information of the task layer and the equipment layer. After the decision layer completes the arbitration, it generates a queue diagram to guide workshop production.

3. The toy packaging workshop collaborative manufacturing method based on industrial Internet according to claim 1 is characterized in that: In S3, the task layer calculates the priority of the task in advance. When the priority of the task is higher than a predetermined value, the task layer sends the task to the smart device; when the priority of the task is lower than the predetermined value, the task layer stores the task in the task pool.

4. The toy packaging workshop collaborative manufacturing method based on industrial Internet according to claim 2 is characterized in that: The local optimization includes: the decision layer selects an interaction strategy and arbitrates the combination of tasks and equipment based on a predetermined target benefit function, arbitrates based on the order grabbing situation of smart devices and manufacturing tasks, calculates the target benefit function value of each candidate task-equipment based on real-time data, assigns tasks to the optimal smart device to form the final task-equipment combination, and generates a real-time scheduling diagram to guide the production process of the physical workshop.

5. The toy packaging workshop collaborative manufacturing method based on the industrial Internet according to claim 4 is characterized in that: The global optimization includes: based on the real-time data of the current task progress and equipment status, and based on the existing pre-scheduling results, within a predetermined time range, all pending manufacturing tasks and smart devices are globally rescheduled in the cloud, and the optimal task-equipment combination is selected from the candidate pool based on the global scheduling results; at the same time, at the beginning of each scheduling cycle, the manufacturing tasks and smart devices are periodically globally rescheduled based on the prediction results and real-time data in the cloud, and the scheduling results are used as a reference for the arbitration team to select the optimal task-equipment combination.

6. The toy packaging workshop collaborative manufacturing method based on the industrial Internet according to any one of claims 1 to 5, characterized in that: The mathematical model is Where, T: total duration of production task; J: Workshop equipment station set; P: Production task set; A: Production capacity of each equipment; ST: Production plan start time; ET: Production plan completion time; D: Other abnormal parameters that occur during the actual production process.

Images