WO2023207060A1

WO2023207060A1 - Distributed production line scheduling method and system based on dispatching rules

Info

Publication number: WO2023207060A1
Application number: PCT/CN2022/134150
Authority: WO
Inventors: 孙铮; 张健琛; 徐俊; 孙孝飞; 梅雪松
Original assignee: 西安交通大学
Priority date: 2022-04-24
Filing date: 2022-11-24
Publication date: 2023-11-02
Also published as: CN114995301A

Abstract

Disclosed are a distributed production line scheduling method and system based on dispatching rules. A local area network for workshop device communication is constructed according to the workshop device composition, the workshop layout and an operation logic and information collection and transmission architecture, and is connected to the cloud; the local area network issues information through the cloud, the information comprising tasks, task priorities and a scheduling optimization target; an agent uses available devices to establish a temporary device network according to the issued information, the processing information of the tasks being only sent and received in the currently established temporary device network; realizing local scheduling of distributed devices according to the processing information sent and received in the temporary device network; overlapping the multiple temporary device networks for local scheduling, and arranging the material processing of multiple tasks according to the rules and priorities so as to achieve distributed scheduling of the production line. Global scheduling is eventually realized by achieving local scheduling and combining all local scheduling activities. Therefore, the problems of unknown global scheduling information, workshop disturbance and poor real-time response are effectively solved.

Description

A production line distributed scheduling method and system based on dispatch rules

Technical field

The invention belongs to the field of workshop scheduling technology, and specifically relates to a production line distributed scheduling method and system based on dispatch rules.

Background technique

Currently, the main scheduling methods used in production lines include global static scheduling and a combination of global static and dynamic scheduling. Global static scheduling must obtain all information required for calculation before production, and cannot respond and adjust to disturbances generated in production. The method of combining global static and dynamic scheduling can adjust the production plan according to disturbances, but there are problems such as slow response time and unreasonable adjusted production plan.

The main scheduling methods used in existing workshop production are global static scheduling before production and dynamic scheduling during the production process. Global static requires obtaining materials, equipment and other information in advance, and the calculation is complex and time-consuming. After the production plan is broken by disturbance, the initial scheduling is difficult to apply. Dynamic scheduling is based on the original static scheduling plan and adjusts the scheduling plan according to different disturbances to make the scheduling plan closer to actual production. However, dynamic scheduling is prone to problems such as excessive modification of the production plan, frequent rescheduling, and unreasonable rescheduling.

technical problem

The technical problem to be solved by the present invention is to provide a production line distributed scheduling method and system based on dispatch rules to achieve local scheduling and combine all local scheduling to achieve the final global scheduling in view of the above-mentioned deficiencies in the prior art. The invention can effectively overcome the problems of unknowable global scheduling information, workshop disturbance, poor real-time response and other problems.

Technical solutions

A production line distributed scheduling method based on dispatch rules, including the following steps:

S1. Construct a local area network for workshop equipment communication based on the workshop equipment composition, workshop layout and operation logic, and information collection and transmission architecture, and connect to the cloud;

S2. The local area network constructed in step S1 delivers information through the cloud. The information includes tasks, task priorities, and scheduling optimization goals;

S3. The agent uses the available equipment to establish a temporary equipment network based on the information sent in step S2. The processing information of the task is only sent and received within the currently established temporary equipment network;

S4. Implement local scheduling of distributed equipment based on the processing information sent and received in the temporary equipment network in step S3;

S5. Multiple temporary equipment networks that are partially scheduled through step S4 are overlapped. The underlying workshop operating equipment arranges the material processing of multiple tasks according to the assignment rules and priorities to achieve distributed scheduling of the production line.

Specifically, step S1 is as follows:

All equipment in the workshop is connected to form a local area network, and the local area network is connected to the cloud through switches and routers to realize information transmission.

Specifically, in step S2, task priorities are divided into urgent tasks, more urgent tasks and general tasks; the optimization goals include the shortest processing time, minimized cost, minimum energy consumption and minimum delay.

Specifically, in step S3, the temporary equipment network is composed of all the equipment used for a certain task within the local area network and is used for the transmission of scheduling information related to the current task, and the temporary equipment network is only used for the current task, and each temporary equipment network receives The optimization goals and corresponding scheduling rules are delivered to the cloud, and the optimization goals and corresponding scheduling rules only apply to the current temporary device network.

Specifically, in step S4, the local scheduling of distributed devices is as follows:

In a temporary device network, after the current device is completed, it sends the processed information through the LAN. All downstream devices that meet the requirements receive the processed information and send their own status information to the current device. The current device relies on its own edge devices. The corresponding rules stored in it rely on edge computing to determine the best output path for the workpiece. The status information includes its own processing parameters, processing status, time required for transfer and the end time of the current task.

Furthermore, the optimal output path is specifically:

The current device obtains the status information of the next-level device and considers whether any device in the next-level device is idle. If there are multiple devices that are idle, the nearest device is selected. If there is no idle device, the device with the smallest congestion waiting time is selected. ;

The current equipment obtains the status information of the next-level equipment, including material transfer time, blocking waiting time and processing time, calculates the total delay time, and calculates the best workpiece output equipment;

The current equipment obtains all equipment status information from the next level to the next N level, including transit time, congestion waiting time, and equipment processing time. N can be changed by the cloud; the next level to the next N level is obtained through evolutionary algorithms or reinforcement learning methods. The best material output path between the paths is selected, and the first-level equipment in the path is selected for material transportation. If there are multiple best paths, one of them is randomly selected.

Specifically, in step S4, when an emergency check is required, the cloud sends the plug-in information to the designated agent, and the agent forms a new temporary device network. When a device is occupied by multiple device networks, the device passes Determine the task priority and give priority to processing resources to high-level tasks; when equipment fails, the faulty equipment does not provide feedback to the upstream; when material congestion occurs, in the parallel production line, the decision-making equipment considers the obstruction of the next-level equipment According to the situation, the workpiece is output according to the scheduling rules; in a non-parallel production line, the decision-making equipment evaluates the blocking situation of an entire branch line; when a task's raw materials are delayed, causing the equipment in the corresponding equipment network to be idle, Process workpieces from other tasks.

Specifically, step S5 is as follows:

S501. After constructing a temporary equipment network, for equipment that exists in multiple equipment networks at the same time, if it is not the first level of any task, the leveling principle will be implemented for the input task workpieces of any priority, and the arrival of materials according to different tasks will be implemented. Processing in order;

S502. When device A is the intermediate device of task α and the first-level device in task β; device A sends a network coincidence signal in the temporary network of task α to the first-level device in the network; task α is temporary The decision-making of the first-level equipment in the network is executed based on the network coincidence signal. The network coincidence signal is the priority difference P=P _β -P _α between task α and task β. If the first level of task α and task β overlap, the device Send yourself a network coincidence signal.

Furthermore, when the priority difference between the two tasks is P=0, equipment A adopts a leveling strategy to receive materials; when the priority difference between the two tasks is P=1, equipment A limits the feeding frequency of task α; when the priority difference between the two tasks is P=2, equipment A stops the reception of task α material; when the priority difference between the two tasks is P=-1, equipment A limits the feeding frequency of task β; when the priority difference between the two tasks is P=-2, equipment A stops the task Receipt of beta materials.

In the second aspect, embodiments of the present invention provide a production line distributed scheduling system based on dispatch rules, including:

The network module builds a local area network for workshop equipment communication based on the workshop equipment composition, workshop layout and operation logic, and information collection and transmission architecture, and connects to the cloud;

Information module, the local area network built by the network module delivers information through the cloud. The information includes tasks, task priorities and scheduling optimization goals;

In the sending module, the agent uses the available equipment to build a temporary equipment network based on the information sent by the information module. The processing information of the task is only sent and received within the currently established temporary equipment network;

The receiving module implements local scheduling of distributed equipment based on the processing information sent and received by the sending module within the temporary equipment network;

The scheduling module overlaps multiple temporary equipment networks that are partially scheduled through the receiving module. The underlying workshop operating equipment arranges the material processing of multiple tasks based on dispatch rules and priorities to achieve distributed scheduling of the production line.

beneficial effects

The present invention is a production line distributed scheduling method based on dispatch rules, which connects the bottom equipment of the workshop to each other to form a local area network, and then connects the equipment local area network to the local area network where the cloud is located. This division method makes it easier for the cloud to deliver task-related information to the workshop, and allows scheduling decisions to be completed within the local area network without relying on the cloud. In complex workshops, there are many devices, and it is difficult for the cloud to collect the operating status of each device in real time. Moreover, the messy data will burden the cloud, and the resulting global scheduling results may also deviate from the actual device operation. The distributed workshop architecture constructed by the present invention can solve this problem well, and utilizes the underlying equipment to make self-decisions to improve the real-time performance of dispatching and the ability to respond to disturbances.

Further, in step S1, building a temporary device network can allow complex local scheduling information to be circulated and processed within the local area network, avoiding uploading to the cloud and increasing the computing burden on the cloud processor, and reducing the difficulty of data processing. The agent integrates the internal scheduling information of an entire temporary device network, and then quickly uploads the overall task scheduling status through switches and routers.

Furthermore, setting different task priorities and optimization goals can further improve the flexibility of workshop production and help the scheduling system better achieve different scheduling optimizations according to different needs of the workshop.

Furthermore, for different tasks, they have different optimization goals and task urgency. Establishing a temporary network of multiple devices can ensure that the devices use the correct optimization goals and parameters when performing scheduling calculations. At the same time, for different tasks, the best candidate processing equipment is different, so when a new task is issued, the agent needs to re-plan the temporary equipment network.

Furthermore, scheduling calculation requires status information of workshop equipment. The traditional collection method is to obtain global workshop status data through upper-layer equipment and then perform global scheduling, which results in a heavy burden on upper-layer equipment and a large delay in scheduling. The present invention uses Information exchange completes device status information acquisition and local scheduling, which can reduce information delay and use edge devices to quickly implement local scheduling. At the same time, it can avoid the cloud collecting scattered workshop status data, which will cause the cloud to be overburdened.

Furthermore, setting different scheduling and assignment rules can achieve scheduling of different complexity and different optimization goals for different tasks, making workshop scheduling more flexible. Rule 1 has the lowest computational complexity, but takes into account the current state of the next-level equipment and has fewer factors to consider. It is suitable for production workshops with simple structures and low requirements. Rules 2 and 3 are more complex, take into account more time costs, and are suitable for production workshops with complex structures. There are many designs of dispatch rules. The present invention only lists three typical categories. In an actual workshop, more complex dispatch rules can be designed to meet different optimization goals and workshop production modes.

Furthermore, due to the structure of the temporary equipment network, information transmission between upstream and downstream is real-time, which helps equipment making scheduling decisions to obtain real-time status data of other equipment, so that upstream equipment can quickly learn about disturbances in downstream equipment. and take disturbances into account in scheduling decisions. This enables equipment in distributed workshops to have high real-time performance and anti-disturbance capabilities.

Furthermore, by setting rules, we can avoid possible recurrences when multiple temporary device networks overlap. By judging the task priorities of two overlapping device networks, if they are of the same priority, a leveling strategy will be adopted to allow the two overlapping device networks to The materials of the task pass through the overlapping equipment in parallel. If the task priorities are different, the first level of the overlapping equipment will limit or stop accepting materials for low-level tasks to achieve the purpose of completing high-level tasks first. This can avoid the phenomenon that when the network overlaps, the overlapped equipment only processes one task, causing the processing of another task to be stagnant, and improves the comprehensive utilization of all equipment.

Further, by dividing task priorities in detail, the priority difference can better determine which strategy to use to avoid conflicts caused by temporary device network overlap.

To sum up, the present invention uses a distributed workshop architecture to enable the underlying workshop equipment to make self-decisions on scheduling, improve the calculation speed of scheduling and resistance to disturbances. By setting up a temporary equipment network, it accepts task information from the cloud and utilizes different Optimize goals and scheduling rules to make workshop scheduling more flexible.

Description of drawings

Figure 1 is a distributed scheduling communication network architecture diagram of the present invention;

Figure 2 is a flow chart of cloud delivery tasks according to the present invention;

Figure 3 is a diagram of the temporary equipment network communication mode of the present invention;

Figure 4 is a flow chart of materials output by the upstream equipment of the present invention;

Figure 5 is a flow chart for receiving materials by downstream equipment of the present invention;

Figure 6 is a flow chart for modifying equipment status information according to the present invention;

Figure 7 is a flow chart of the scheduling strategy ① of the present invention;

Figure 8 is a flow chart of the scheduling strategy ② of the present invention;

Figure 9 is a flow chart of the scheduling strategy ③ of the present invention;

Figure 10 is a schematic diagram of a temporary equipment network coincidence processing method according to the present invention;

Figure 11 is a schematic diagram of the second temporary equipment network coincidence processing method of the present invention;

Figure 12 is a flow chart of temporary equipment network coincidence processing according to the present invention;

Figure 13 is a simulation diagram implemented using a production line distributed scheduling system based on dispatch rules;

Figure 14 is a simulation diagram of the workshop structure;

Figure 15 is a Gantt chart of the final processing situation obtained from normal production in the workshop;

Figure 16 shows the affected workpieces and the final scheduling results.

Embodiments of the invention

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

In the description of the present invention, it is to be understood that the terms "comprising" and "including" indicate the presence of described features, integers, steps, operations, elements and/or components, but do not exclude one or more other features, The existence or addition of an integer, a step, an operation, an element, a component, and/or a collection thereof.

It should also be understood that the terminology used in the description of the present invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this specification and the appended claims, the singular forms "a", "an" and "the" are intended to include the plural forms unless the context clearly dictates otherwise.

It will be further understood that the term "and/or" as used in the specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items. , for example, A and/or B can mean: A alone exists, A and B exist simultaneously, and B exists alone. In addition, the character "/" in this article generally indicates that the related objects are an "or" relationship.

It should be understood that although the terms first, second, third, etc. may be used to describe the preset ranges and the like in embodiments of the present invention, these preset ranges should not be limited to these terms. These terms are only used to distinguish preset ranges from each other. For example, without departing from the scope of the embodiments of the present invention, the first preset range may also be called a second preset range, and similarly, the second preset range may also be called a first preset range.

Depending on the context, the word "if" as used herein may be interpreted as "when" or "when" or "in response to determination" or "in response to detection." Similarly, depending on the context, the phrase "if determined" or "if (stated condition or event) is detected" may be interpreted as "when determined" or "in response to determining" or "when (stated condition or event) is detected )" or "in response to detecting (a stated condition or event)".

Various structural schematic diagrams according to disclosed embodiments of the present invention are shown in the accompanying drawings. The drawings are not drawn to scale, with certain details exaggerated and may have been omitted for purposes of clarity. The shapes of the various regions and layers shown in the figures and the relative sizes and positional relationships between them are only exemplary. In practice, there may be deviations due to manufacturing tolerances or technical limitations, and those skilled in the art will base their judgment on actual situations. Additional regions/layers with different shapes, sizes, and relative positions can be designed as needed.

The invention provides a production line distributed scheduling method based on dispatch rules. The production line distributed scheduling collects processing status information of surrounding equipment through the edge computing device on each processing equipment, and obtains the processing status information after the current workpiece based on rule calculations. Best output path. This kind of information collection and scheduling path calculation are performed in real time. There is no need to obtain all scheduling related information before production, and this kind of real-time scheduling can effectively avoid most disturbances.

The present invention is a production line distributed scheduling method based on dispatch rules, which includes the following steps:

S1. Build a workshop communication LAN architecture, which is used to build workshop equipment composition, workshop layout and operation logic, and information collection and transmission architecture;

Workshop equipment is divided into processing units, material transportation units, and material storage units.

Among them, the processing unit is used to process or assemble workpieces and can complete one or more processes;

Material transportation units include AGV material trucks, conveyor belts, etc., which are used to transfer materials between processing units;

The material storage unit includes a material warehouse and a temporary material accumulation point, which is used for the storage and temporary buffer stacking of materials.

Please refer to Figure 1 to construct a workshop layout and equipment operation logic. The workshop layout is used to explain the mutual positional relationship between the processing units within the workshop, and the workshop operation logic is used to explain the relationship between the processing units, processes, and material processing trajectories; the present invention Applicable workshop layout structures include flexible work workshops, parallel workshops, mixed flow workshops, etc. Device operating logic is used as the basis for edge computing devices to make scheduling decisions.

Information collection and transmission architecture is used to explain the information collection and transmission methods related to distributed scheduling, including the following steps:

S101. Construction of a local area network, which is composed of all equipment in the workshop and is used to realize the internal transmission of scheduling-related information;

Build a local area network to connect all the equipment in the workshop to form a local area network. According to the process of staged products, connect all the equipment in the process section to form a local area network. The devices in the LAN are connected to the switch through the wireless AP, and the switch is responsible for the communication of physical entities in the LAN.

S102. The temporary equipment network is composed of all the equipment within the local area network that can be used for a certain task. It is used for the transmission of scheduling information related to the current task, and the temporary equipment network is only used for the current task;

A temporary equipment network is constructed to complete specific tasks issued by the cloud, and the network is automatically released after completing the task; the equipment will be divided into upstream and downstream relationships according to the order of the process. In the flexible processing workshop, since a single equipment can realize multiple Different processes, so the division of upstream and downstream will be more flexible. For example: the device itself can be its own downstream, and the current device A's upstream device B can also be the downstream device of A; a single device can be requisitioned by multiple temporary device networks, As a candidate machine for processing tasks, a single device cannot perform two processing tasks at the same time. The upstream and downstream relationships are defined by the order of the processes. If process a requires equipment A for processing, process b requires equipment B for processing, and process a is sequenced before process b, then equipment A is the upstream of equipment B.

S103. Connection to the cloud, used for information transmission between the cloud and the workshop equipment LAN, to achieve the issuance of tasks, priorities, and optimization goals.

The connection with the cloud is based on the processing tasks, used for the issuance of processing tasks, optimization goals and priorities, and is connected to the switches and routers of the local area network, and finally connected to the cloud.

S2. Send relevant information about the current task to the agent through the cloud, including tasks, task priorities, and scheduling optimization goals, which are used as a basis for decision-making on edge device scheduling plans;

Please refer to Figure 2. The cloud formulates a processing task and sends the task information to the corresponding LAN through the TCP/IP protocol. The agent analyzes the task, builds a temporary equipment network, and establishes a superior-subordinate relationship. If the current task is completed, the agent sends a completion signal to the cloud.

Task priorities are divided into urgent tasks, more urgent tasks, and general tasks, which serve as the basis for edge computing to allocate specific processing resources to different tasks.

Optimization goals include shortest processing time, minimized cost, minimum energy consumption, minimum delay, etc. Each temporary equipment network will be issued an optimization goal and corresponding scheduling rules by the cloud for scheduling calculation, and the optimization goal and corresponding scheduling Rules only apply to the current temporary device network.

S3. Build a temporary equipment network. The agent will build a LAN based on the assigned task attributes. The processing information related to the task will only be sent and received within the current LAN;

The details of building a temporary device network are:

According to the work instructions issued from the cloud, including normal workpiece processing tasks, as well as order modifications, emergency orders and other disturbances that affect normal processing, the agent belonging to the equipment network searches for all equipment that meets the requirements according to the instructions, and then sends upstream and downstream relationships to these equipments. Data and workpiece scheduling rules; the equipment network organized by the agent is temporary, and the same equipment can be included in multiple temporary equipment networks at the same time; this feature enables distributed scheduling of mixed line production; the processes of different tasks are managed by the same The equipment is used for processing, but the processing processes belong to different temporary equipment networks.

S4. Sending and obtaining processing information within the local area network, including methods of sending and obtaining processing data, communication methods between devices, and realizing local scheduling of distributed equipment;

Please refer to Figure 3, Figure 4 and Figure 5. For the completed upstream device A, confirm all downstream devices that can output, identify them through the LAN IP, send signals to the downstream devices, and wait to receive status information fed back by the downstream devices and parse the status information. Used to calculate the best material output device, send a signal to the best material output device to confirm that it can receive materials, and output the material after feedback confirmation. If there is no feedback, recalculate. For downstream equipment B, it waits to receive the upstream signal. After receiving it, it feeds back its own status information and re-enters the waiting state. After receiving the material signal, it feeds back to the upstream equipment whether it can be received. The device stores its own status information in the database of its own edge device to be called and changed. When the device status changes, the device will modify the status information in the database and enter the waiting state again, as shown in Figure 6.

Distributed device local scheduling. In a temporary device network, the corresponding devices will cooperate to complete the process required by the workpiece; each device has edge computing capabilities and the ability to communicate with other devices in the network; when a device completes a task After a specific process, the workpiece will be output and transported to the next equipment to complete the subsequent process; after the current equipment A is completed, the processing completion information will be sent to all downstream equipment B that meets the requirements through the local area network; after the downstream equipment B receives the information , sends its own status information to device A. Device A relies on edge computing to decide the best output path of the workpiece based on the corresponding rules stored in its own edge device, which specifically includes the following steps:

S401. Acquisition and sending of status data is used to enable edge devices to obtain scheduling-related information and communicate between multiple edge devices;

Acquisition of status data; the status information fed back by the equipment includes its own processing parameters, processing status, time required for transfer, and current task end time; for automation equipment, the current task end time can be calculated based on the remaining work steps and historical data. For batch processing, the end time can be estimated based on the number of remaining unprocessed workpieces; the processing parameters exist in the edge computing database; the processing status data of the equipment exists in the edge computing database to be called. When an event is triggered, its The status data will be converted, otherwise no changes will occur.

Sending of status data; the current device A will find the next-level device or several lower-level devices that meet the conditions according to the upstream and downstream relationships pre-entered by the agent, and send completion status signals to these devices through wireless APs and switches; when the downstream After receiving the signal from Device A, the device will actively feed back the current status information to Device A; Device A will only include the devices that have fed back the information into the scope of the workpiece output device, and actively eliminate the devices without feedback.

S402, material output path decision-making, used to explain the operation mode of local scheduling of distributed equipment;

Material output path decision-making; the current machine A calculates the optimal workpiece output path based on the status information fed back by the downstream equipment and the scheduling rules issued by the agent; based on the computing power of the edge device, scheduling decisions can be divided into the following types:

S4021. Under the simplest decision rule ①, as shown in Figure 7, the current device A obtains the status information of the next-level device and considers whether any device in the next-level device is idle. If there are multiple devices idle, then Select the nearest device. If there is no idle device, select the device with the smallest congestion waiting time;

S4022. Under the more complex decision rule ②, as shown in Figure 8, the current equipment A obtains the status information of the next-level equipment, including the transfer material time, congestion waiting time and processing time, and calculates the total delay time. Best workpiece output equipment;

S4023. Under the more complex decision rule ③, as shown in Figure 9, the current equipment A obtains all equipment status information from the next level to the next N level, including transfer time, congestion waiting time, and equipment processing time. N can be obtained from the cloud Change. The optimal material output path between the next level and the next N levels is obtained through an evolutionary algorithm or reinforcement learning method, and the first-level equipment in the path is selected for material transportation. If multiple best paths appear, one of them is randomly selected.

Current equipment A not only needs to consider the feedback information of the next-level equipment, but also needs to take into account the equipment status of the following levels; in a non-parallel flow workshop, part of the processing path of the workpiece is often unique and cannot be adjusted. , so once an equipment failure or material blockage occurs in the middle link, both the upstream and downstream of the problematic equipment will be affected. For this type of production line, it is necessary to consider the current status and processing parameters of the entire branch production line as a whole; In parallel production lines and flexible production lines, the processing paths of workpieces are more diverse, and the above situation is not easy to occur. However, in actual production, there is a transfer time and installation and disassembly time due to the transfer of workpieces from one equipment to another. Etc., better production scheduling plans tend to choose to let a single piece of equipment complete multiple processes, or when transferring workpieces, give priority to downstream locations with less transfer time.

S403, disturbance countermeasures strategy enables workshop equipment to cope with sudden disturbances during production and processing, and can adjust the production plan in an orderly manner to achieve normal processing.

Distributed workshop scheduling is used to solve a variety of disturbances, including but not limited to the following:

Emergency order

When there is a need for emergency order checking, the cloud will send the plug-in information to the designated agent, and the agent will form a new temporary equipment network. When the equipment is occupied by multiple equipment networks, the equipment will provide emergency processing resources. Inserted device network;

Equipment failure

When a device fails, the faulty device does not provide feedback to the upstream, and the upstream device can avoid the faulty device when selecting subsequent processing paths;

Material blocking

When material congestion occurs, it is necessary to evaluate the congestion situation and try to select non-blocking equipment or equipment that is about to end blocking; in a parallel production line, the decision-making device only needs to consider the blocking situation of the next-level equipment and based on the scheduling rules Carry out workpiece output; in a non-parallel production line, the workpiece cannot change its processing path after entering, so the decision-making equipment needs to evaluate the blocking situation of an entire branch line;

Raw material delays

When a task's raw materials are delayed and the equipment in the corresponding equipment network is idle, since the same equipment can be requisitioned by multiple equipment networks, workpieces in other tasks can still be processed to ensure maximum equipment utilization.

S5. When multiple temporary equipment networks overlap, the equipment arranges material processing for multiple tasks according to rules and priorities to avoid conflicts and blockages.

Please refer to Figure 10, Figure 11 and Figure 12. The solution to conflicts and blocking when multiple temporary networks overlap is to avoid them by prioritizing tasks. The specific steps are as follows:

S501. After building a temporary device network, for devices that exist in multiple device networks at the same time, if they are not the first level of any task, any input priority task artifacts will not be prioritized. The principle of leveling is implemented, and materials for different tasks are processed according to the order in which they arrive;

S502. When a certain device becomes the first-level device in a temporary device network, assume that device A is an intermediate-level device in a certain task α and the first-level device in another task β; Define rules so that device A sends a network coincidence signal to the first level in the temporary network of α; the decision of the first level device B in the temporary network of α will be based on the network coincidence signal, and the network coincidence signal is based on two tasks The priority difference P=P _β -P _α ; according to the task priority score in step S3, 1 represents a general task, 2 represents a more urgent task, and 3 represents an urgent task. When the priority difference between the two tasks is P=0, Equipment A adopts a leveling strategy to receive materials; when the priority difference between the two tasks is P=1, equipment A limits the material delivery frequency of task α; when the priority difference between the two tasks is P=2, equipment A stops receiving materials for task α; when When the priority difference between the two tasks is P=-1, equipment A limits the material delivery frequency of task β; when the priority difference between the two tasks is P=-2, equipment A stops receiving materials for task β; for the restricted task output in the present invention The decision of material frequency is made through the cloud based on the overall operation of the workshop, and the specific frequency is modified. The specific operations are as follows:

S5021. When the priority difference P between the two tasks is 2, the first-level device B in the α temporary network will stop receiving the workpieces of task α, and device A treats all received workpieces according to the principle of equality; in the α temporary network, only The first-level equipment stops receiving, and the subsequent-level equipment is still processing;

S5022. When the priority difference P between the two tasks is 1, the first-level device B in the α temporary network will limit the reception of task α workpieces and reduce production capacity; device A uses the leveling principle to treat all received workpieces;

S5023. When the priority difference P between the two tasks is 0, the first-level device B in the α temporary network adopts a leveling strategy;

S5024. When the priority difference P between the two tasks is -1 or -2, the first-level device B in the α temporary network adopts the equalization strategy, while the first-level device A in the β temporary network adopts the priority difference. Reduce the receiving frequency or stop receiving raw materials;

If there is a first-level overlap between task α and task β, first-level device A sends a network coincidence signal to itself.

In yet another embodiment of the present invention, a production line distributed scheduling system based on dispatch rules is provided. The system can be used to implement the above production line distributed dispatch method based on dispatch rules. Specifically, the production line based on dispatch rules The distributed scheduling system includes a network module, an information module, a sending module, a receiving module and a scheduling module.

Among them, the network module builds a local area network for workshop equipment communication based on the workshop equipment composition, workshop layout and operation logic, and information collection and transmission architecture, and connects to the cloud;

Figure 13 is one of the simulations implemented using a production line distributed scheduling system based on dispatch rules. In this model, the workshop is a 9*5 mixed flow workshop, with a total of 9 processes, and each process has 5 parallel processing machine tools. , 25 tasks are reached at random times, and the number of workpieces within each task is also randomly distributed. Different transit times between equipment are considered in this model. The agent divides the best processing candidate equipment for each task, forming a temporary equipment network for each task. Through the preset three types of dispatching scheduling rules and the strategy to resolve network conflicts of multiple temporary devices, each device can self-determine the best output object of the workpiece. The final scheduling simulation results are shown in Figure 1. Figure 1 shows well that when the task arrival time and task processing content cannot be obtained in advance, workshop equipment can schedule the workpieces arriving in real time through self-decision within the temporary equipment network to ensure the orderly progress of production. A production conflict occurs.

Figure 14, Figure 15 and Figure 16 respectively show the second simulation of the production line distributed scheduling system based on dispatch rules. Figure 14 shows the workshop structure. This model is a hybrid flow workshop with 5 process sections. Each process section has 5, 3, 3, 3, and 5 parallel processing equipment respectively. A total of 20 random time arrival tasks, each task has one machining workpiece. This model simulates the workshop scheduling situation of the production line distributed scheduling system under two disturbances: random arrival of tasks and equipment failure. Figure 15 is a Gantt chart of the final processing situation obtained by normal production in the workshop without equipment failure. Figure 16 shows the final production and processing results obtained by the workshop in response to the equipment failure after a random piece of equipment failed at a certain moment in the production process of Figure 3. In the results of this simulation, the equipment that failed was a piece of equipment in the third process section. After the equipment failure, the processing of some tasks was not affected, and the affected workpieces and the final scheduling results are shown in Figure 16 middle. Compared with the global scheduling algorithm that requires time-consuming rescheduling after each disturbance, the production line distributed scheduling system based on dispatch rules can well cope with the occurrence of multiple disturbances, such as random arrival of tasks and equipment failure. The device responds quickly to spontaneous disturbances through self-decision and handles disturbances with extremely low latency.

In summary, the present invention is a production line distributed scheduling method and system based on dispatching rules. It obtains status information between underlying processing equipment and performs real-time scheduling self-decision-making, effectively realizing tasks of different scales under different optimization goals. mixed line production.

The above contents are only for illustrating the technical ideas of the present invention and cannot be used to limit the protection scope of the present invention. Any changes made based on the technical ideas proposed by the present invention and based on the technical solutions shall fall within the scope of the claims of the present invention. within the scope of protection.

Claims

A production line distributed scheduling method based on dispatch rules, which is characterized by including the following steps:

S1. Construct a local area network for workshop equipment communication based on the workshop equipment composition, workshop layout and operation logic, and information collection and transmission architecture, and connect to the cloud;

S2. The local area network constructed in step S1 delivers information through the cloud. The information includes tasks, task priorities, and scheduling optimization goals;

S3. The agent uses the available equipment to establish a temporary equipment network based on the information sent in step S2. The processing information of the task is only sent and received within the currently established temporary equipment network;

S4. Implement local scheduling of distributed equipment based on the processing information sent and received in the temporary equipment network in step S3;

S5. Multiple temporary equipment networks that are partially scheduled through step S4 are overlapped. The underlying workshop operating equipment arranges the material processing of multiple tasks according to the assignment rules and priorities to achieve distributed scheduling of the production line.
The production line distributed scheduling method based on dispatch rules according to claim 1, characterized in that step S1 is specifically:

All equipment in the workshop is connected to form a local area network, and the local area network is connected to the cloud through switches and routers to realize information transmission.
The production line distributed scheduling method based on dispatch rules according to claim 1, characterized in that, in step S2, task priorities are divided into urgent tasks, more urgent tasks and general tasks; the optimization goals include the shortest processing time, minimization cost, minimum energy consumption and minimum delay.
The production line distributed scheduling method based on dispatch rules according to claim 1, characterized in that, in step S3, the temporary equipment network is composed of all the equipment used for a certain task within the local area network, and is used for the current task-related scheduling information. transmission, and the temporary device network is only used for the current task. Each temporary device network receives the optimization target and corresponding scheduling rules issued by the cloud, and the optimization target and corresponding scheduling rules only apply to the current temporary device network.
The production line distributed scheduling method based on dispatch rules according to claim 1, characterized in that in step S4, the local scheduling of distributed equipment is specifically:

In a temporary device network, after the current device is completed, it sends the processed information through the LAN. All downstream devices that meet the requirements receive the processed information and send their own status information to the current device. The current device relies on its own edge devices. The corresponding rules stored in it rely on edge computing to determine the best output path for the workpiece. The status information includes its own processing parameters, processing status, time required for transfer and the end time of the current task.
The production line distributed scheduling method based on dispatch rules according to claim 5, characterized in that the optimal output path is specifically:

The current device obtains the status information of the next-level device and considers whether any device in the next-level device is idle. If there are multiple devices that are idle, the nearest device is selected. If there is no idle device, the device with the smallest congestion waiting time is selected. ;

The current equipment obtains the status information of the next-level equipment, including material transfer time, blocking waiting time and processing time, calculates the total delay time, and calculates the best workpiece output equipment;

The current equipment obtains all equipment status information from the next level to the next N level, including transit time, congestion waiting time, and equipment processing time. N can be changed by the cloud; the next level to the next N level is obtained through evolutionary algorithms or reinforcement learning methods. The best material output path between the paths is selected, and the first-level equipment in the path is selected for material transportation. If there are multiple best paths, one of them is randomly selected.
The production line distributed scheduling method based on dispatch rules according to claim 1, characterized in that, in step S4, when an emergency order check occurs, the cloud sends the plug-in information to the designated agent, and the agent is formed In the new temporary equipment network, when a device is occupied by multiple device networks, the device determines the task priority and gives priority to processing resources to high-level tasks; when a device fails, the faulty device does not provide feedback to the upstream; when a material failure occurs When blocked, in a parallel production line, the decision-making device considers the blocking situation of the next-level equipment and outputs workpieces according to scheduling rules; in a non-parallel production line, the decision-making device evaluates the blocking situation of an entire branch line; when The delay in raw materials for a task causes the equipment in the corresponding equipment network to be idle, processing workpieces in other tasks.
The production line distributed scheduling method based on dispatch rules according to claim 1, characterized in that step S5 is specifically:

S501. After constructing a temporary equipment network, for equipment that exists in multiple equipment networks at the same time, if it is not the first level of any task, the leveling principle will be implemented for the input task workpieces of any priority, and the arrival of materials according to different tasks will be implemented. Processing in order;

S502. When device A is the intermediate device of task α and the first-level device in task β; device A sends a network coincidence signal in the temporary network of task α to the first-level device in the network; task α is temporary The decision-making of the first-level equipment in the network is executed based on the network coincidence signal. The network coincidence signal is the priority difference P=P β -P α between task α and task β. If the first level of task α and task β overlap, the device Send yourself a network coincidence signal.
The production line distributed scheduling method based on dispatch rules according to claim 8, characterized in that when the priority difference between the two tasks is P=0, equipment A adopts a leveling strategy to receive materials; when the priority difference between the two tasks is P =1, equipment A limits the material delivery frequency of task α; when the priority difference between the two tasks P=2, equipment A stops receiving materials for task α; when the priority difference between the two tasks P=-1, equipment A limits the output of task β Material frequency; when the priority difference between the two tasks is P=-2, equipment A stops receiving materials for task β.
A production line distributed scheduling system based on dispatch rules, which is characterized by including:

The network module builds a local area network for workshop equipment communication based on the workshop equipment composition, workshop layout and operation logic, and information collection and transmission architecture, and connects to the cloud;

Information module, the local area network built by the network module delivers information through the cloud. The information includes tasks, task priorities and scheduling optimization goals;

In the sending module, the agent uses the available equipment to build a temporary equipment network based on the information sent by the information module. The processing information of the task is only sent and received within the currently established temporary equipment network;

The receiving module implements local scheduling of distributed equipment based on the processing information sent and received by the sending module within the temporary equipment network;

The scheduling module overlaps multiple temporary equipment networks that are partially scheduled through the receiving module. The underlying workshop operating equipment arranges the material processing of multiple tasks based on dispatch rules and priorities to achieve distributed scheduling of the production line.