Distributed Scheduling System and Scheduling Method Based on Competition

Through competitive distributed scheduling system and two-door algorithm, the problems of multi-workstation information synchronization and low decision-making efficiency in intelligent factories are solved, and the system is high robustness and scalability is achieved, ensuring the stability of the production process and Optimized resource configuration.

CN112529260BActive Publication Date: 2024-11-26EFORT INTELLIGENT EQUIP CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202011354205.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2020-11-26
Publication Date
2024-11-26
Estimated Expiration
2040-11-26

AI Technical Summary

Technical Problem

In intelligent factories, as the number of concurrent workstations increases, the scheduling system needs to synchronize the amount of information from multiple workstations, resulting in information synchronization failure and other errors, resulting in system robustness and low decision-making efficiency.

Method used

A competitive distributed scheduling system and a two-door algorithm are adopted to achieve information synchronization and high responsiveness through competitive resource allocation between producers and consumers.

Benefits of technology

Improves the robustness and decision-making efficiency of the system, and can automatically troubleshoot problems when some equipment of producers and consumers occur, ensure the normal operation of the system, and re-engage in the decision-making and resource application process after troubleshooting, achieving high robustness and scalability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN112529260B_ABST
    Figure CN112529260B_ABST
Patent Text Reader

Abstract

The present invention relates to the technical field of distributed scheduling, in particular to a competitive distributed scheduling system and a scheduling method thereof, comprising: a consumer set, a producer set, and a scheduling center, which are responsible for communicating with all consumers and producers and making decisions to schedule producer resources to consumers; a resource transfer control is responsible for transferring consumer and producer resources according to instructions from the scheduling center, and the specific steps are as follows: S1, dynamic allocation of workstation tasks; S2, multiple producers compete for production; S3, multiple consumers compete for resources; S4, the current resources of the system are allocated to the optimal position. The present invention divides the equipment of the entire production line according to producers, consumers, a scheduling center and a transfer control function, utilizes a system architecture based on competition to synchronize and manage the production and consumption relationship of multiple devices, processes a dual-gate algorithm for distributed concurrent information synchronization, and is designed for virtual devices that adapt to the robustness of the system.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of distributed scheduling, and in particular to a distributed scheduling system and a scheduling method thereof based on competition. Background Art

[0002] At present, with the improvement and popularization of the Internet of Things and big data infrastructure, assembly line production factories are beginning to develop towards intelligence and unmanned operation. Intelligent factory assembly lines require smooth information exchange between equipment, smooth automated flow between objects, and digital recording and management of all production and transportation links. More importantly, the intelligent system can intelligently manage and schedule the production and transportation links of the entire factory. In the production line, one process can often be handled by multiple workers, such as sorting tasks. In an unmanned intelligent factory, one person's position is replaced by a functionally specialized workstation. In order for multiple workstations to handle the same process at the same time, the scheduling system needs to concurrently and synchronously manage, schedule, distribute, tasks and materials. At the same time, it can monitor the work progress of each device and sub-workstation at all times, so as to estimate the various states of the entire production line and make decisions for the scheduling of the next resource and task. With the increase in the number of concurrent workstations, the amount of information that the scheduling system needs to synchronize multiple workstations increases synchronously, and the complexity of summarizing and analyzing the overall data content increases geometrically. In this process, once a workstation fails to synchronize information or encounters other errors, it poses a major challenge to the analysis of the scheduling system. Therefore, a mechanism is needed to process information concurrently to improve the robustness of the system. Summary of the invention

[0003] In order to solve the above problems, the present invention proposes a distributed scheduling system and a scheduling method based on competitiveness.

[0004] A competitive distributed scheduling system includes a host computer for distributing tasks, including:

[0005] Consumer set, a collection of multiple consumers, used to apply for resources;

[0006] Producer set: a set of multiple producers, used to produce resources, apply for and wait for resources to be scheduled;

[0007] The dispatch center is responsible for communicating with all consumers and producers and making decisions to dispatch producer resources to consumers;

[0008] Resource transfer control is responsible for transferring resources between consumers and producers according to the instructions of the dispatch center.

[0009] The specific steps of the scheduling method based on the competitive distributed scheduling system are as follows:

[0010] S1. Dynamic allocation of workstation tasks:

[0011] a: First, the host computer sends tasks to the scheduling system from time to time. After the scheduling system receives the tasks, it broadcasts the tasks to all workstations. Idle workstations decide whether to apply for the tasks based on their own conditions and calculate the scores based on the degree of demand.

[0012] b: The scheduling system obtains the demand scores of all workstations that apply for the task through the double-gate algorithm, and makes decisions based on the scores, selects the best workstation, and notifies the accepted workstation to get the task, so that the tasks of the workstation can be dynamically allocated;

[0013] S2. Multiple producers compete for production:

[0014] a: The scheduling system obtains resource transfer requests from all applicant producers through a double-gate algorithm;

[0015] b: Notify the producer whose request is accepted. The scheduling system determines the best producer based on the producer's request information. The notified producer immediately prepares and waits for the resources to be transported away by the transport equipment.

[0016] S3, multiple consumers compete for resources:

[0017] a: The scheduling system broadcasts the resource information to be received by the producer currently accepted to all consumers. Consumers request the resource through the double-gate algorithm based on their own status;

[0018] b: The scheduling system decides the best consumer of the current resources based on the consumer's request information and notifies the accepted consumer that the resources are available;

[0019] c: The consumer immediately prepares and waits for the resources to be transported to its loading port by the transport equipment. Then, the scheduling system notifies the conveying equipment to transport the resources from the corresponding producer's unloading port to the corresponding loading port of the consumer.

[0020] S4. The system's current resources are allocated to the optimal location.

[0021] The dual-gate algorithm of step S1 is to realize the one-to-many information exchange, information synchronization and high information response requirements. The system synchronization information is submitted by the producer or consumer to the system's mutually exclusive memory, that is, the critical area.

[0022] The dual-gate algorithm includes:

[0023] Definition of critical section: It is an accessible shared resource that cannot be accessed by multiple threads or processes at the same time. The critical section is represented by M.

[0024] Thread: The smallest unit in the system that can independently calculate and schedule. It can independently apply for and access resources in the scheduling system. Producers and consumers are independent threads. The set of all threads is represented by P.

[0025] System intercept gate: The operating system is used to intercept, sleep and wake up thread signals. There are two groups of signals in the scheduling system, represented as D1 and D2;

[0026] Entering thread sleep: the entering thread enters the sleep state, no longer works, and waits to be awakened;

[0027] Incoming thread wake-up: The incoming thread is awakened by the operating system and continues to work;

[0028] When threads enter the resource critical section in turn, it is represented as ~M.

[0029] The specific operation steps of the dual-gate algorithm are as follows:

[0030] S1, first initialize all threads, i.e. p0, p1, p2...pn∈P, initialize the system synchronization interception gates D1, D2 and critical section M;

[0031] S2, all threads are intercepted by the first gate D1 and enter the sleep state;

[0032] S3, the system opens the first door D1, and the awakened threads enter the critical section in turn to submit the competition score, which is, ~M, and the second door is closed, and the thread that has accessed the critical section will be intercepted and enter the sleep state;

[0033] S4, after time t0, the system decides based on the voting scores of all threads in the critical section, notifies the winning thread that its application is accepted, and resources are about to be scheduled;

[0034] S5, the system closes the first door D1 and prepares to intercept the thread that is about to enter the first door;

[0035] S6, the system opens the second door D2 and wakes up all threads to repeat the next cycle;

[0036] S7. The awakened thread continues to loop and enter step S2 in sequence.

[0037] The request of step S2 includes the resource information produced by the producer, the score assessment of the producer's urgency for the transportation of the produced resources, and the specific unloading location of the resources.

[0038] The request in step S3 includes the consumer information, the consumer's evaluation score of the urgency of the resource, and the loading location type of the requested resource.

[0039] The beneficial effects of the present invention are as follows: the present invention divides the equipment of the entire production line according to producers, consumers, dispatching centers and transfer control functions, utilizes a system architecture based on competition to synchronize and manage the production and consumption relationships of multiple devices, processes a dual-gate algorithm for distributed concurrent information synchronization, and designs virtual devices to adapt to the robustness of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0041] Figure 1 It is a block diagram of the virtual device structure of the present invention;

[0042] Figure 2 It is a structural block diagram of the scheduling system of the present invention;

[0043] Figure 3 It is a block diagram of the abstract process competition structure of the present invention;

[0044] Figure 4 It is a structural block diagram of the double-gate algorithm of the present invention;

[0045] Figure 5 It is a flow chart of the scheduling system of the present invention. DETAILED DESCRIPTION

[0046] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further described below.

[0047] like Figures 1 to 5 As shown, the competitive distributed scheduling system includes a host computer for distributing tasks, including:

[0048] Consumer set, a collection of multiple consumers, used to apply for resources;

[0049] Producer set: a set of multiple producers, used to produce resources, apply for and wait for resources to be scheduled;

[0050] The dispatch center is responsible for communicating with all consumers and producers and making decisions to dispatch producer resources to consumers;

[0051] Resource transfer control is responsible for transferring resources between consumers and producers according to the instructions of the dispatch center.

[0052] The concurrent scheduling system has made the following improvements and innovations:

[0053] 1) Equipment identity classification: Equipment is divided into producers, consumers and controlled equipment. Producers can actively send requests to the scheduling system to send the prepared resources to the transmission equipment. Once approved by the system, the producer will immediately hand over the resources to the transmission equipment either actively or passively. Consumers can actively request the system to send resources to themselves. Once approved, the system will immediately dispatch the current resources to the consumer. Controlled equipment provides a protocol interface and is called by other identity-type devices;

[0054] 2) Design of the resource competition system of devices: Producers and consumers compete for production and consumption resources in the form of scoring according to their own urgency and conditions. The system determines the current producers and consumers in an orderly manner according to the scores submitted by each identity device and the system environment status. In essence, the system decision-making process is initially completed on each device concurrently, reducing the complexity of system decision-making;

[0055] 3) Virtual device establishment: Figure 1 As shown, the virtual device is responsible for communicating with the scheduling system, managing its own data information, synchronizing the real hardware equipment and equipment safety detection. The establishment of virtual devices is conducive to improving the robustness of the system; when any device has a problem, the virtual system can continue to work, generate alarms and submit records.

[0056] The scheduling system is an abstract process of competition. The abstract process is as follows:

[0057] Through competition, the scheduling system finds the only optimal producer in the producer set;

[0058] The scheduling system broadcasts the production resource information of the unique optimal producer to the consumer set;

[0059] The consumer set competes for the resource, and the scheduling system determines the only optimal consumer;

[0060] The scheduling system transfers resources from the current optimal producer to the current optimal consumer to complete a complete competition.

[0061] The competition method of the scheduling system is completely reflected in the form of scores. According to the rules, the producer / consumer with the highest competition score has priority to obtain the configuration rights of the scheduling system.

[0062] Competition rules are determined based on the current production environment and production goals.

[0063] The design of competition rules determines the operating efficiency of the current system.

[0064] The competition is completely quantified in the form of scores, and the scheduling system determines the winner of the competition based on the scores.

[0065] Consumers and producers that do not participate in the competition do not submit scores.

[0066] The dispatch center communicates with consumers and producers in a one-to-many manner. Since producers and consumers are not unique and each device operates independently of each other, the dispatch center often communicates with consumers and producers in a one-to-many manner.

[0067] The dispatch center needs to synchronize producer information first and then consumer information in order. During the synchronization process with producer aggregate information, consumers must be kept silent to avoid information confusion, and vice versa.

[0068] The dispatch center needs to maintain a high responsiveness to all consumers and producers at all times. When the dispatch center opens up consumer competition, it needs to obtain competition responses from all consumers in a short period of time.

[0069] The present invention divides the equipment of the entire production line according to producers, consumers, dispatching centers and transfer control functions, utilizes a system architecture based on competition to synchronize and manage the production and consumption relationships of multiple devices, processes a dual-gate algorithm for distributed concurrent information synchronization, and designs virtual devices to adapt to the robustness of the system.

[0070] The present invention enables the subsystems in the system to participate in decision-making based on the local information they have mastered, thereby optimizing the efficiency of overall decision-making. In a distributed system, each subnode and subsystem, such as a single device, has the most comprehensive information about itself. Producers and consumers in each device can actively calculate and give the scheduling system the demand level of the subsystem based on their own identity and status, just like people. The central scheduling system then optimizes resource allocation based on the demand level of the subsystem, thereby making resource scheduling more optimized and decision-making more efficient.

[0071] The specific steps of the scheduling method based on the competitive distributed scheduling system are as follows:

[0072] S1. Dynamic allocation of workstation tasks:

[0073] a: First, the host computer sends tasks to the scheduling system from time to time. After the scheduling system receives the tasks, it broadcasts the tasks to all workstations. Idle workstations decide whether to apply for the tasks based on their own conditions and calculate the scores based on the degree of demand.

[0074] b: The scheduling system obtains the demand scores of all workstations that apply for the task through the double-gate algorithm, and makes decisions based on the scores, selects the best workstation, and notifies the accepted workstation to get the task, so that the tasks of the workstation can be dynamically allocated;

[0075] S2. Multiple producers compete for production:

[0076] a: The scheduling system obtains resource transfer requests from all applicant producers through a double-gate algorithm;

[0077] b: Notify the producer whose request is accepted. The scheduling system determines the best producer based on the producer's request information. The notified producer immediately prepares and waits for the resources to be transported away by the transport equipment.

[0078] S3, multiple consumers compete for resources:

[0079] a: The scheduling system broadcasts the resource information to be received by the producer currently accepted to all consumers. Consumers request the resource through the double-gate algorithm based on their own status;

[0080] b: The scheduling system decides the best consumer of the current resources based on the consumer's request information and notifies the accepted consumer that the resources are available;

[0081] c: The consumer immediately prepares and waits for the resources to be transported to its loading port by the transport equipment. Then, the scheduling system notifies the conveying equipment to transport the resources from the corresponding producer's unloading port to the corresponding loading port of the consumer.

[0082] S4. The system's current resources are allocated to the optimal location.

[0083] The dual-gate algorithm of step S1 is to realize the one-to-many information exchange, information synchronization and high information response requirements. The system synchronization information is submitted by the producer or consumer to the system's mutually exclusive memory, that is, the critical area.

[0084] The scheduling system is more robust and scalable. When a part of the equipment of the producer and consumer fails to work, the subsystem will no longer participate in decision-making and request resources, and the subsystem will not participate in production. Then the faulty system is automatically excluded from the system, and the overall system operates abnormally and is not affected by the faulty system. When the subsystem fault is eliminated, the subsystem continues to participate in the decision-making and resource application process. Under the condition that the system does not stop, any subsystem can participate in and exit the system production process at will. Ultimately, the overall system has high robustness and scalability.

[0085] The present invention has excellent concurrent processing capability. By adding a double-gate algorithm design, the scheduling system can quickly and effectively respond to the concurrent needs of subsystems and subthreads, and solve the problem of concurrent communication resource preemption from an algorithmic perspective.

[0086] The dual-gate algorithm includes:

[0087] Definition of critical section: It is an accessible shared resource that cannot be accessed by multiple threads or processes at the same time. The critical section is represented by M.

[0088] Thread: The smallest unit in the system that can independently calculate and schedule. It can independently apply for and access resources in the scheduling system. Producers and consumers are independent threads. The set of all threads is represented by P.

[0089] System intercept gate: The operating system is used to intercept, sleep and wake up thread signals. There are two groups of signals in the scheduling system, represented as D1 and D2;

[0090] Entering thread sleep: the entering thread enters the sleep state, no longer works, and waits to be awakened;

[0091] Incoming thread wake-up: The incoming thread is awakened by the operating system and continues to work;

[0092] When threads enter the resource critical section in turn, it is represented as ~M.

[0093] The double-gate algorithm is used to solve the problem of duplicate consumption and information synchronization of each thread in the process of concurrent communication in the production system. The double-gate algorithm controls thread information synchronization and solves the problem of duplicate resource consumption in an orderly manner by setting two "gates" to synchronize the rhythm of thread access to the critical section of resources.

[0094] The specific operation steps of the dual-gate algorithm are as follows:

[0095] S1, first initialize all threads, i.e. p0, p1, p2...pn∈P, initialize the system synchronization interception gates D1, D2 and critical section M;

[0096] S2, all threads are intercepted by the first gate D1 and enter the sleep state;

[0097] S3, the system opens the first door D1, and the awakened threads enter the critical section in turn to submit the competition score, which is, ~M, and the second door is closed, and the thread that has accessed the critical section will be intercepted and enter the sleep state;

[0098] S4, after time t0, the system decides based on the voting scores of all threads in the critical section, notifies the winning thread that its application is accepted, and resources are about to be scheduled;

[0099] S5, the system closes the first door D1 and prepares to intercept the thread that is about to enter the first door;

[0100] S6, the system opens the second door D2 and wakes up all threads to repeat the next cycle;

[0101] S7. The awakened thread continues to loop and enter step S2 in sequence.

[0102] The request of step S2 includes the resource information produced by the producer, the score assessment of the producer's urgency for the transportation of the produced resources, and the specific unloading location of the resources.

[0103] The request in step S3 includes the consumer information, the consumer's evaluation score of the urgency of the resource, and the loading location type of the requested resource.

[0104] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A competitive distributed scheduling system, including a host computer for distributing tasks, characterized in that: include: Consumer set, a collection of multiple consumers, used to apply for resources; Producer set: a set of multiple producers, used to produce resources, apply for and wait for resources to be scheduled; The dispatch center is responsible for communicating with all consumers and producers and making decisions to dispatch producer resources to consumers; Resource transfer control, responsible for transferring resources between consumers and producers according to the instructions of the dispatch center; A system architecture that uses a competition-based approach to synchronize and manage the production and consumption relationships of multiple devices, a dual-gate algorithm for processing distributed concurrent information synchronization, and a virtual device design that adapts to system robustness; The dual-gate algorithm includes: Definition of critical section: It is an accessible shared resource that cannot be accessed by multiple threads or processes at the same time. The critical section is represented by M. Thread: The smallest unit in the system that can independently calculate and schedule. It can independently apply for and access resources in the scheduling system. Producers and consumers are independent threads. The set of all threads is represented by P. System intercept gate: The operating system is used to intercept, sleep and wake up thread signals. There are two groups of signals in the scheduling system, represented as D1 and D2; Entering thread sleep: the entering thread enters the sleep state, no longer works, and waits to be awakened; Incoming thread wake-up: The incoming thread is awakened by the operating system and continues to work; When threads enter the resource critical section in turn, it is represented as ~M.

2. The scheduling method of the competitive distributed scheduling system according to claim 1, characterized in that: The specific steps are as follows: S1. Dynamic allocation of workstation tasks: a: First, the host computer sends tasks to the scheduling system from time to time. After the scheduling system receives the tasks, it broadcasts the tasks to all workstations. Idle workstations decide whether to apply for the tasks based on their own conditions and calculate the scores based on the degree of demand. b: The scheduling system obtains the demand scores of all workstations that apply for the task through the double-gate algorithm, and makes decisions based on the scores, selects the best workstation, and notifies the accepted workstation to get the task, so that the tasks of the workstation can be dynamically allocated; S2. Multiple producers compete for production: a: The scheduling system obtains resource transfer requests from all applicant producers through a double-gate algorithm; b: Notify the producer whose request is accepted. The scheduling system determines the best producer based on the producer's request information. The notified producer immediately prepares and waits for the resources to be transported away by the transport equipment. S3, multiple consumers compete for resources: a: The scheduling system broadcasts the resource information to be received by the producer currently accepted to all consumers. Consumers request the resource through the double-gate algorithm based on their own status; b: The scheduling system decides the best consumer of the current resources based on the consumer's request information and notifies the accepted consumer that the resources are available; c: The consumer immediately prepares and waits for the resources to be transported to its loading port by the transport equipment. Then, the scheduling system notifies the transport equipment to transport the resources from the corresponding producer unloading port to the corresponding loading port of the consumer. S4. The system's current resources are allocated to the optimal location.

3. The scheduling method of the competitive distributed scheduling system according to claim 2, characterized in that: The dual-gate algorithm of step S1 is to realize the one-to-many information exchange, information synchronization and high information response requirements. The system synchronization information is submitted by the producer or consumer to the system's mutually exclusive memory, that is, the critical area.

4. The scheduling method of the competitive distributed scheduling system according to claim 2, characterized in that: The specific operation steps of the dual-gate algorithm are as follows: S1. Initialize all threads first, namely p0, p1, p2 ...pn ∈P, initialize the system synchronization interception gates D1, D2 and critical section M; S2, all threads are intercepted by the first gate D1 and enter the sleep state; S3, the system opens the first door D1, and the awakened threads enter the critical section in turn to submit the competition score, which is, ~M, and the second door is closed, and the thread that has accessed the critical section will be intercepted and enter the sleep state; S4, after time t0, the system decides based on the voting scores of all threads in the critical section, notifies the winning thread that its application is accepted, and resources are about to be scheduled; S5, the system closes the first door D1 and prepares to intercept the thread that is about to enter the first door; S6, the system opens the second door D2 and wakes up all threads to repeat the next cycle; S7. The awakened thread continues to loop and enter step S2 in sequence.

5. The scheduling method of the competitive distributed scheduling system according to claim 2, characterized in that: The request of step S2 includes the resource information produced by the producer, the score assessment of the producer's urgency for the transportation of the produced resources, and the specific unloading location of the resources.

6. The scheduling method of the competitive distributed scheduling system according to claim 2, characterized in that: The request in step S3 includes the consumer information, the consumer's evaluation score of the urgency of the resource, and the loading location type of the requested resource.