CN112861371A - Steel industry cloud production scheduling method based on edge computing - Google Patents
Steel industry cloud production scheduling method based on edge computing Download PDFInfo
- Publication number
- CN112861371A CN112861371A CN202110230369.1A CN202110230369A CN112861371A CN 112861371 A CN112861371 A CN 112861371A CN 202110230369 A CN202110230369 A CN 202110230369A CN 112861371 A CN112861371 A CN 112861371A
- Authority
- CN
- China
- Prior art keywords
- scheduling
- cloud
- production
- slab
- edge
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/60—Protecting data
- G06F21/62—Protecting access to data via a platform, e.g. using keys or access control rules
- G06F21/6218—Protecting access to data via a platform, e.g. using keys or access control rules to a system of files or objects, e.g. local or distributed file system or database
- G06F21/6245—Protecting personal data, e.g. for financial or medical purposes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Bioethics (AREA)
- Strategic Management (AREA)
- General Health & Medical Sciences (AREA)
- Computer Hardware Design (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- General Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- Medical Informatics (AREA)
- Databases & Information Systems (AREA)
- Computer Security & Cryptography (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an edge computing-based cloud production scheduling method for the steel industry, which is used for converting the split of a traditional cloud production scheduling mode into a cloud-edge cooperative production scheduling mode and comprises two modules of edge preprocessing and cloud platform universal solution; firstly, establishing an intelligent scheduling framework of cloud edge cooperation; then, the satisfying degree of the production process is characterized as the comprehensive attribute difference between the plate blanks; process requirements are met to the greatest extent by minimizing composite attribute differences between slabs in a production plan sequence; determining an optimization target needing to be considered such as minimizing the production time on the basis of the data; and finally, according to the optimization target and the obtained functional relation, determining two parts of calculation tasks of an edge processing module of the edge server and a universal solving module of the cloud production scheduling platform.
Description
Technical Field
The invention relates to the field of hot rolling of steel and high-level planning and scheduling, in particular to a cloud production scheduling method for the steel industry based on edge computing.
Background
In the scheduling task of the steel rolling process, the enterprise production process execution management system transmits the slab data to be scheduled to a high-level planning and scheduling system, which contains a large amount of attribute information such as slab length, width, tapping temperature and the like. And the advanced planning and scheduling system automatically formulates a slab processing sequence according to the rolling flow requirement after acquiring the data. The difficulty of steel rolling production scheduling is that the setting parameters of workshop production equipment (such as a heating furnace, a rolling mill and the like) are related to the properties of a produced slab. For example, when rolling slabs having different target widths, it is necessary to set the rolling mill differently. In order to improve production efficiency and reduce setup latency and machine wear, it is desirable to achieve a relatively smooth transition in various attributes between slabs in a production plan. The method requires that the slabs to be rolled in the slab library are comprehensively considered and evaluated according to the length, the width, the tapping temperature and other properties of the slabs, and then are scheduled to make a reasonable production plan. The advanced plan scheduling system realizes automatic scheduling, greatly reduces scheduling time compared with manual scheduling, enables the production plan to be more efficient and reasonable to be formulated, and avoids errors caused by artificial fatigue and misjudgment. However, the traditional advanced planning and scheduling system is expensive, complex to deploy and difficult to maintain, so that the traditional advanced planning and scheduling system is not cost-effective for small and medium-sized enterprises and is not suitable for independently deploying and maintaining the system.
In recent years, a cloud platform mode appears in an advanced planning and scheduling system, and the core concept is to convert an application system of the advanced planning and scheduling system into a mode for providing scheduling services. The method is divided into an enterprise/factory platform and a cloud platform, and the working process is as follows: in the enterprise/factory platform, an enterprise user uploads related industrial data related to scheduling to a server of the cloud scheduling platform, then waits for the return of a scheduling result and analyzes the result; and on the cloud scheduling platform, according to the received industrial data and scheduling logic, performing scheduling of a production plan by using an advanced plan scheduling system which is deployed in the cloud server in advance, and returning a result. The cloud scheduling system not only ensures the advantages and convenience brought by the high-level planning and scheduling, but also solves the problems of high price, complex deployment, difficult maintenance and the like caused by the high-level planning and scheduling system. However, cloud scheduling also faces scheduling efficiency and enterprise production information leakage problems. Since the physical location of the cloud server is often far from the user, if the amount of uploaded data required for providing the service is large, the situation that the transmission speed is slow and a large amount of bandwidth is occupied is likely to occur. In addition, in order to support the mode of cloud scheduling, the enterprise needs to upload all production data to a cloud platform of a third party, so that the risk of data leakage after cloud uploading is avoided.
Therefore, the invention introduces the edge computing mode into the cloud production scheduling system, and transfers part of data processing and analyzing tasks in the production scheduling process to the edge of the network for computing by giving certain computing power and storage capacity to the network edge equipment so as to achieve the effects of reducing time delay and improving service quality. And the original data is processed at the edge server, and the uploaded data can not reveal the production information of enterprises on the premise of not influencing the quality of the scheduling result.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides an edge-computing-based cloud production scheduling method for the steel industry, which is used for solving the transmission efficiency problem and the potential enterprise production information leakage problem in the traditional cloud production scheduling mode of the steel industry. The cloud scheduling task is divided into the edge preprocessing module and the cloud platform general solving module on the basis of establishing the cloud-edge cooperative intelligent scheduling framework, partial scheduling tasks are transferred to the edge of a network to be calculated so as to achieve the effects of reducing time delay and improving service quality, and after the edge server processes original data, the uploaded data are prevented from being leaked by enterprise production information on the premise of not influencing the quality of scheduling results.
In order to realize the purpose of the invention, the technical scheme adopted by the invention is as follows: an edge computing-based cloud production scheduling method for the steel industry comprises the following steps:
the method comprises the following steps of (1) firstly changing a computing mode of traditional cloud production scheduling, providing a highly decoupled cloud-edge cooperative cloud production scheduling framework in the steel industry, dividing the cloud production scheduling framework into an edge preprocessing module and a universal solution module, respectively deploying the edge preprocessing module and the universal solution module on an edge server and a cloud production scheduling platform, and enabling the two modules to independently operate.
Step (2) in the actual production scene, the production processDifferent from constraint conditions, the common solving method and the common flow need to be extracted after the common solving flow of the scheduling problem is deduced, namely, the satisfying degree of the production process is characterized as the comprehensive attribute difference cost between any two plate blanks i, ji,j;
And (3) converting the process satisfaction degree rationality representation problem of the production plan into a minimum comprehensive attribute difference cost in the production plan sequence, wherein the comprehensive attribute difference between any two slabs in the production plan sequence is obtained in the step (2). Determining a final optimization target by combining the process satisfaction degree, wherein the optimization target is a measurement standard for the quality of a production plan formulated result;
and (4) designing and optimizing a preprocessing module g (X, alpha) of the edge server according to the optimization target and the function relation, and finishing the evaluation and analysis of the satisfaction degree of the plate blank process, wherein X represents input data of the edge server, and alpha represents a parameter vector for extracting industrial information. Meanwhile, a reasonable cloud production scheduling platform universal solving module h (G) is designed, a universal production scheduling algorithm is solved according to the received intermediate result G, and a production scheduling result is returned;
and (5) combining the edge preprocessing module g (X, alpha) obtained in the step (4) with the cloud production scheduling platform universal solving module h (G) to perform cloud production scheduling task calculation on the input industrial production scheduling data in a cloud edge collaborative mode.
Further, in the step (1), the paradigm of the cloud scheduling framework of the steel industry with cloud-edge cooperation is as follows:
establishing a general paradigm of a scheduling framework of cloud edge cooperation, wherein f (X, alpha) represents a scheduling task for a certain process, g (X, alpha) represents an edge personalized preprocessing module, and h (G) represents a cloud scheduling platform solving module:
f(X,α)=h(g(X,α))
for the edge preprocessing module, the industrial original data X and the parameter vector α are input and starting points of the whole scheduling task, G (X, α) is a personalized preprocessing module customized by the edge server according to specific process requirements, production logic and constraint conditions, the input scheduling task is converted into an intermediate result G in a standard form after passing through the personalized preprocessing module at the edge server, and the form is as follows:
G=g(X,α)
for the cloud platform universal solving module, the intermediate result G is data which is extracted from the original industrial data and can be directly used for solving the scheduling algorithm and does not directly contain the standard form of the real attribute information of the slab. The cloud scheduling platform can perform universal solution only through the received standard intermediate result G, wherein f represents the result of the cloud scheduling:
f=h(G)
in a scheduling framework of cloud-side cooperation, an edge server mainly has the task of designing a reasonable personalized preprocessing function module G (X, alpha), and obtaining an intermediate result G interacted with a cloud platform according to input industrial data and parameter vectors; the cloud scheduling platform mainly has the task of designing a reasonable function module h (G), solving a general scheduling algorithm according to a received intermediate result G and returning a scheduling result.
Further, in the step (2), making the comprehensive attribute difference cost between any two slabs i, ji,jThe general flow of the calculation method and the scheduling problem is as follows:
the essence of the problem is to minimize the overall attribute differences between adjacent slabs, while two slabs with identical attributes are best suited to be arranged in adjacent locations for production, the greater the overall attribute differences the lower the process satisfaction. Wherein the composite attribute difference cost between slabs i, ji,jCan be expressed in the form of a weighted sum of differences of properties between slabs, wherein attrkRepresenting the kth attribute of the slab, a representing the set of all attributes of the slab,indicating that the slab i, j is at attribute attrkDegree of difference of (a)kRepresenting the current attribute k weight. The specific formula is as follows:
to this end, the process satisfaction degree can be expressed as the composite attribute difference cost between slabs i, ji,j. The essential goal of such scheduling problems is to minimize the sum of costs between slabs in the production planning sequence. The general flow of such advanced planning scheduling problems can be derived as follows:
(a) representing the satisfying degree of the depicting process as the comprehensive attribute difference between the slabs;
(b) minimizing the composite attribute differences between slabs in the production plan sequence to meet process requirements to the greatest extent;
(c) on the basis, optimization solution is carried out by combining conventional optimization objectives such as minimized working hours, minimized delay quantity and the like.
Further, in the step (3), a final optimization target is determined according to the comprehensive property difference between any two slabs in the production plan sequence and the process satisfaction degree.
Further, in the step (4), a preprocessing module g (X, α) is designed for the edge server and a general solving module h (g) is designed for the cloud scheduling platform, and the specific method is as follows:
(4.1) for the edge server, the evaluation score of the process satisfaction degree between the slabs i and j can be expressed as a composite attribute difference value costi,jThen only the comprehensive attribute difference values between all the slabs need to be costi,jThe expected purpose can be achieved by storing the data in a data structure and uploading the data to the cloud production scheduling platform. To this end, the input form of the intermediate result G may be represented in the form of a weighted adjacency matrix. Any row and column represents a certain slab in the slab warehouse, and the array [ i, j]The numerical value of (a) represents the comprehensive attribute difference value cost of the slab j of the rear slab ii,jThe slab is not reusable, so the distance from the slab to the slab is infinite, and the specific data structure is as follows:
(4.2) for the cloud production scheduling platform, the main task is to designAnd the reasonable universal solving module h (G) is used for solving the universal production scheduling algorithm according to the received intermediate result G and returning a production scheduling result. According to the received scheduling task and the intermediate result data, a weighted directed complete graph G (V, E), V (Node) can be constructed1、Node2、Node3、……、Nodei},E={Pair1,2、Pair1,3、……、Pair1,j、Pair2,j……、Pairi,j}. The set of points V represents a set of slabs, each point representing a block of slab. The edge set E represents the sequential set between two adjacent slabs. The weight of each point is represented by the property of the slab itself and the weight of each edge is represented by the combined property difference between adjacent slabs. From the constructed graph, it can be modeled as a vehicle path problem.
Further, in step (4.2), a modeling manner of the vehicle path problem on the cloud scheduling platform is as follows:
for simplicity, when the model is solved, only the sum of the comprehensive attribute difference values of the minimized slab pair needs to be considered, so that the final production optimization target is obtained as follows. Wherein xi,jIndicating whether the path is selected, if so, 1, otherwise, 0, costi,jThe comprehensive property difference between slabs i and j is as follows:
in order to ensure the uniqueness of each slab in the production plan, each slab needs to be only a precursor node of any slab in the slab library except for the slab, and the specific constraints are as follows:
similarly to the above formula, in order to ensure the uniqueness of each slab in the production plan, each slab needs to be only a successor node of any slab in the slab library except for itself, and the specific constraints are as follows:
in order to ensure that the number of production lines is the same as the actual number, a super source point s and a super sink point d are arranged, and the number of paths from the super source point is equal to the number of production lines | K |, x |s,iIndicating whether the path from s to i is selected, if so, it is 1, otherwise it is 0. The specific constraints are as follows:
to date, a general vehicle path problem model has been created, but slight adjustments are required for a particular manufacturing process. Such as: the capacity factor may be incorporated as a capacity-limited Vehicle Routing Problem (CVRP) when limiting the total volume of batch production of production slabs, the time factor may be incorporated as a time-windowed Vehicle Routing Problem (VRPTW) when the time horizon for slab production is limited, and so on.
Compared with the prior art, the technical scheme of the invention has the following beneficial technical effects:
(1) according to the technical scheme, the data transmission time in the scheduling process is shortened, and partial data processing and analyzing tasks in the scheduling process are transferred to the edge of the network for calculation by giving certain calculation capacity and storage capacity to the network edge equipment so as to achieve the effect of reducing time delay. According to the experimental result of the comparison experiment by using the simple cloud production scheduling algorithm of a certain steel group, the time length of the scheme adopting the edge calculation is reduced by about 10% compared with the simple cloud production scheduling mode in the production scheduling time consumption;
(2) and the occupation situation of the bandwidth in the scheduling process is reduced. The original production data are processed at the edge server, so that the data volume which needs to be uploaded to the cloud platform for the next calculation is greatly reduced, and the occupation condition of the bandwidth is reduced. According to the experimental result of the comparison experiment by using the simple cloud production scheduling algorithm of a certain steel group, the scheme of adopting edge calculation reduces the transmission data amount by about 30 percent compared with the mode of simple cloud production scheduling;
(3) the privacy disclosure problem caused by uploading real original production data to a cloud platform of a third party is avoided. The original data are processed at the edge server, and the uploaded data can not reveal the production information of enterprises on the premise of not influencing the quality of the scheduling result;
(4) the highly-coupled scheduling process is decoupled, the universality of a cloud-side cooperation scheduling mode based on the steel industry is improved, and the cloud-side cooperation scheduling framework can be more conveniently deployed and expanded to be applied to different fields.
Drawings
FIG. 1 is an advanced planning scheduling framework based on edge computation;
FIG. 2 is a flow chart of a cloud scheduling method for the steel industry based on edge computing.
Detailed Description
For the purposes of promoting an understanding and understanding of the invention, reference will now be made to the following descriptions taken in conjunction with the accompanying drawings and specific examples.
Example 1: the invention is operated cooperatively between an edge server and a cloud production scheduling platform in the steel industry, an integral production scheduling framework is shown in figure 1, factory sensor equipment is responsible for collecting industrial data related to production scheduling, the edge server is responsible for carrying out personalized preprocessing on the data, an obtained intermediate result is transmitted to the cloud production scheduling platform after the data are processed, the intermediate result is used as the input of an algorithm model deployed on the cloud production scheduling platform, and the cloud production scheduling platform obtains an operation result of a production scheduling algorithm under the input of the intermediate result and returns the operation result to the edge server at a factory end. The invention provides a cloud production scheduling method for the steel industry based on edge computing, and the flow is shown in figure 2. The specific execution steps are as follows:
the method comprises the following steps of (1) firstly changing a computing mode of traditional cloud production scheduling, providing a highly decoupled cloud-edge cooperative cloud production scheduling framework in the steel industry, dividing the cloud production scheduling framework into an edge preprocessing module and a universal solution module, respectively deploying the edge preprocessing module and the universal solution module on an edge server and a cloud production scheduling platform, and enabling the two modules to independently operate.
In the actual production scene, the production process and the constraint conditions are different, and a common solving method and a common flow need to be extracted after a common solving flow of the scheduling problem is deduced, namely, the satisfaction degree of the production process is characterized as the comprehensive attribute difference cost between any two slabs i and ji,j;
Step (3) converting the process satisfaction degree rationality representation problem of the production plan into a comprehensive attribute difference cost in a minimized production plan sequence, obtaining the comprehensive attribute difference between any two slabs in the production plan sequence from the step (2), and determining a final optimization target by combining the process satisfaction degree, wherein the optimization target is a measurement standard for the quality of a production plan formulation result;
and (4) designing and optimizing a preprocessing module g (X, alpha) of the edge server according to the optimization target and the function relation, and finishing the evaluation and analysis of the satisfaction degree of the plate blank process, wherein X represents input data of the edge server, and alpha represents a parameter vector for extracting industrial information. Meanwhile, a reasonable cloud production scheduling platform universal solving module h (G) is designed, a universal production scheduling algorithm is solved according to the received intermediate result G, and a production scheduling result is returned;
and (5) combining the edge preprocessing module g (X, alpha) obtained in the step (4) with the cloud production scheduling platform universal solving module h (G) to perform cloud production scheduling task calculation on the input industrial production scheduling data in a cloud edge collaborative mode.
Further, in the step (1), the paradigm of the cloud scheduling framework of the steel industry with cloud-edge cooperation is as follows:
establishing a general paradigm of a scheduling framework of cloud edge cooperation, wherein f (X, alpha) represents a scheduling task for a certain process, g (X, alpha) represents an edge personalized preprocessing module, and h (G) represents a cloud scheduling platform solving module:
f(X,α)=h(g(X,α))
for the edge preprocessing module, the industrial original data X and the parameter vector α are input and starting points of the whole scheduling task, G (X, α) is a personalized preprocessing module customized by the edge server according to specific process requirements, production logic and constraint conditions, the input scheduling task is converted into an intermediate result G in a standard form after passing through the personalized preprocessing module at the edge server, and the form is as follows:
G=g(X,α)
for the cloud platform universal solving module, the intermediate result G is data which is extracted from the original industrial data and can be directly used for solving the scheduling algorithm and does not directly contain the standard form of the real attribute information of the slab. The cloud scheduling platform can perform universal solution only through the received standard intermediate result G, wherein f represents the result of the cloud scheduling:
f=h(G)
in a scheduling framework of cloud-side cooperation, an edge server mainly has the task of designing a reasonable personalized preprocessing function module G (X, alpha), and obtaining an intermediate result G interacted with a cloud platform according to input industrial data and parameter vectors; the cloud scheduling platform mainly aims to design a reasonable function module h (G), solve a general scheduling algorithm according to a received intermediate result G and return a scheduling result
Further, in the step (2), making the comprehensive attribute difference cost between any two slabs i, ji,jThe general flow of the calculation method and the scheduling problem is as follows:
the essence of the problem is to minimize the overall attribute differences between adjacent slabs, while two slabs with identical attributes are best suited to be arranged in adjacent locations for production, the greater the overall attribute differences the lower the process satisfaction. Wherein the composite attribute difference cost between slabs i, ji,jCan be expressed in the form of a weighted sum of differences of properties between slabs, wherein attrkRepresenting the kth attribute of the slab, a representing the set of all attributes of the slab,indicating that the slab i, j is at attribute attrkDegree of difference of (a)kRepresenting the current attribute k weight. The specific formula is as follows:
to this end, the process satisfaction degree can be expressed as the composite attribute difference cost between slabs i, ji,j. The essential goal of such scheduling problems is to minimize the sum of costs between slabs in the production planning sequence. The general flow of such advanced planning scheduling problems can be derived as follows:
(a) representing the satisfying degree of the depicting process as the comprehensive attribute difference between the slabs;
(b) minimizing the composite attribute differences between slabs in the production plan sequence to meet process requirements to the greatest extent;
(c) on the basis, optimization solution is carried out by combining conventional optimization objectives such as minimized working hours, minimized delay quantity and the like.
Further, in the step (3), a final optimization target is determined according to the comprehensive property difference between any two slabs in the production plan sequence and the process satisfaction degree.
Further, in the step (4), a preprocessing module g (X, α) is designed for the edge server and a general solving module h (g) is designed for the cloud scheduling platform, and the specific method is as follows:
(4.1) for the edge server, the evaluation score of the process satisfaction degree between the slabs i and j can be expressed as a composite attribute difference value costi,jThen only the comprehensive attribute difference values between all the slabs need to be costi,jThe expected purpose can be achieved by storing the data in a data structure and uploading the data to the cloud production scheduling platform. To this end, the input form of the intermediate result G may be represented in the form of a weighted adjacency matrix. Any row and column represents a certain slab in the slab warehouse, and the array [ i, j]The numerical value of (a) represents the comprehensive attribute difference value cost of the slab j of the rear slab ii,jThe slab is not reusable, so the distance from the slab to the slab is infinite, and the specific data structure is as follows:
(4.2) for the cloud production scheduling platform, the main task is to design a reasonable universal type solving module h (G), carry out universal type production scheduling algorithm solving according to the received intermediate result G and return a production scheduling result. According to the received scheduling task and the intermediate result data, a weighted directed complete graph G (V, E), V (Node) can be constructed1、Node2、Node3、……、Nodei},E={Pair1,2、Pair1,3、……、Pair1,j、Pair2,j……、Pairi,j}. The set of points V represents a set of slabs, each point representing a block of slab. The edge set E represents the sequential set between two adjacent slabs. The weight of each point is represented by the property of the slab itself and the weight of each edge is represented by the combined property difference between adjacent slabs. From the constructed graph, it can be modeled as a vehicle path problem.
Further, in step (4.2), a modeling manner of the vehicle path problem on the cloud scheduling platform is as follows:
for simplicity, when the model is solved, only the sum of the comprehensive attribute difference values of the minimized slab pair needs to be considered, so that the final production optimization target is obtained as follows. Wherein xi,jIndicating whether the path is selected, if so, 1, otherwise, 0, costi,jThe comprehensive property difference between slabs i and j is as follows:
in order to ensure the uniqueness of each slab in the production plan, each slab needs to be only a precursor node of any slab in the slab library except for the slab, and the specific constraints are as follows:
similarly to the above formula, in order to ensure the uniqueness of each slab in the production plan, each slab needs to be only a successor node of any slab in the slab library except for itself, and the specific constraints are as follows:
in order to ensure that the number of production lines is the same as the actual number, a super source point s and a super sink point d are arranged, and the number of paths from the super source point is equal to the number of production lines | K |, x |s,iIndicating whether the path from s to i is selected, if so, it is 1, otherwise it is 0. The specific constraints are as follows:
to date, a general vehicle path problem model has been created, but slight adjustments are required for a particular manufacturing process. Such as: the capacity factor may be incorporated as a capacity-limited Vehicle Routing Problem (CVRP) when limiting the total volume of batch production of production slabs, the time factor may be incorporated as a time-windowed Vehicle Routing Problem (VRPTW) when the time horizon for slab production is limited, and so on.
It should be noted that the above-mentioned embodiments are only preferred embodiments of the present invention, and are not intended to limit the scope of the present invention, and all equivalent substitutions or substitutions made on the basis of the above-mentioned technical solutions belong to the scope of the present invention.
Claims (5)
1. A cloud production scheduling method for the steel industry based on edge computing is characterized by comprising the following steps:
establishing a highly decoupled cloud-edge cooperative steel industry cloud production scheduling framework, wherein the framework is divided into an edge preprocessing module and a universal solving module, and the two modules are respectively deployed on an edge server and a cloud production scheduling platform and can independently operate;
step (2) is in practiceIn the production scene, the production process and the constraint conditions are different, and a common solving method and a common flow need to be extracted after a common solving flow of the scheduling problem is deduced, namely, the satisfaction degree of the production process is characterized as the comprehensive attribute difference cost between any two slabs i and ji,j;
Step (3) converting the process satisfaction degree rationality representation problem of the production plan into a comprehensive attribute difference cost in a minimized production plan sequence, obtaining the comprehensive attribute difference between any two slabs in the production plan sequence from the step (2), and determining a final optimization target by combining the process satisfaction degree, wherein the optimization target is a measurement standard for the quality of a production plan formulation result;
step (4) designing and optimizing a preprocessing module G (X, alpha) of the edge server according to the optimized target and the function relation, finishing evaluation and analysis of the satisfaction degree of the slab process, wherein X represents input data of the edge server, alpha represents a parameter vector for extracting industrial information, and simultaneously designing a reasonable cloud production scheduling platform universal solving module h (G), carrying out universal production scheduling algorithm solving according to the received intermediate result G, and returning a production scheduling result;
and (5) combining the edge preprocessing module g (X, alpha) obtained in the step (4) with the cloud production scheduling platform universal solving module h (G) to perform cloud production scheduling task calculation on the input industrial production scheduling data in a cloud edge collaborative mode.
2. The cloud scheduling method for the steel industry based on edge computing as claimed in claim 1, wherein in the step (1), the scheduling framework of cloud-edge collaboration is specifically designed as follows:
(1.1) establishing a scheduling framework overall paradigm of cloud edge cooperation, wherein f (X, alpha) represents a scheduling task for a certain process, g (X, alpha) represents an edge personalized preprocessing module, and h (G) represents a cloud scheduling platform solving module:
f(X,α)=h(g(X,α))
(1.2) for the edge preprocessing module, wherein the industrial original data X and the parameter vector alpha are input and starting points of the whole scheduling task, G (X, alpha) is a personalized preprocessing module which is customized by an edge server according to specific process requirements, production logic and constraint conditions, the input scheduling task is converted into an intermediate result G in a standard form after passing through the personalized preprocessing module at the edge server, and the form is as follows:
G=g(X,α)
(1.3) for the cloud platform universal solution module, the intermediate result G is data in a standard form which is extracted from original industrial data and can be directly used for solving the scheduling algorithm and does not directly contain real attribute information of the slab, the cloud scheduling platform can carry out universal solution only through the received standard intermediate result G, and f represents the result of cloud scheduling:
f=h(G)
(1.4) in a scheduling framework of cloud-side cooperation, the edge server mainly has the task of designing a reasonable personalized preprocessing function module G (X, alpha), and obtaining an intermediate result G interacted with a cloud platform according to input industrial data and parameter vectors; the cloud scheduling platform mainly has the task of designing a reasonable function module h (G), solving a general scheduling algorithm according to a received intermediate result G and returning a scheduling result.
3. The cloud scheduling method for steel industry based on edge computing as claimed in claim 1, wherein in the step (2), the satisfaction degree of the production process is characterized by being drawn as the comprehensive attribute difference cost between any two slabs i, ji,jAnd the general flow of the scheduling problem, the specific method is as follows:
(2.1) wherein the composite attribute difference cost between slabs i, ji,jExpressed as a weighted sum of the differences of the properties between slabs, wherein attrkRepresenting the kth attribute of the slab, a representing the set of all attributes of the slab,indicating that the slab i, j is at attribute attrkDegree of difference of (a)kRepresenting the weight of the current attribute k, and the specific formula is as follows:
(2.2) Up to this point, the degree of process satisfaction is expressed as the composite attribute difference cost between slabs i, ji,jThe essential goal of such scheduling problems is to minimize the sum of costs between slabs in the production planning sequence, so the flow of such advanced scheduling problems can be derived as follows:
(a) representing the satisfying degree of the depicting process as the comprehensive attribute difference between the slabs;
(b) minimizing the composite attribute differences between slabs in the production plan sequence to meet process requirements to the greatest extent;
(c) on the basis, optimization solution is carried out by combining conventional optimization objectives such as minimized working hours, minimized delay quantity and the like.
4. The cloud scheduling method for the steel industry based on edge computing as claimed in claim 1, wherein in step (4), a preprocessing module g (X, α) is designed for an edge server, and a general solving module h (g) is designed for a cloud scheduling platform, and the specific method is as follows:
(4.1) for the edge server, the evaluation score of the process satisfaction degree between the slabs i and j can be expressed as a composite attribute difference value costi,jThen only the comprehensive attribute difference values between all the slabs need to be costi,jThe expected purpose can be achieved by storing the intermediate result in a data structure and uploading the intermediate result to a cloud production scheduling platform, for this reason, the input form of the intermediate result G is expressed in the form of a weighted adjacent matrix, any row and column represent a certain slab in a slab warehouse, and the array [ i, j ] is]The numerical value of (a) represents the comprehensive attribute difference value cost of the slab j of the rear slab ii,jThe slab is not reusable, so the distance from the slab to the slab is infinite, and the specific data structure is as follows:
(4.2) for the cloud scheduling platform, the main task is to design a reasonable universal solution module h (G), perform universal scheduling algorithm solution according to the received intermediate result G and return scheduling results, and construct a weighted directed complete graph G ═ V, E ═ V ═ Node { Node ═ n according to the received scheduling tasks and intermediate result data1、Node2、Node3、……、Nodei},E={Pair1,2、Pair1,3、……、Pair1,j、Pair2,j……、Pairi,jA point set V represents a set of slabs, each point represents one slab, and an edge set E represents a set of sequences between two adjacent slabs; the weight of each point is represented by the property of the slab itself, the weight of each edge is represented by the composite property difference value between adjacent slabs, and the graph is modeled as a vehicle path problem according to the constructed graph.
5. The cloud scheduling method for the steel industry based on edge computing as claimed in claim 4, wherein in the step (4.2), the concrete constraint formula of the modeling mode of the vehicle path problem is as follows:
(4.21) for simplicity, when solving the model, only the sum of the comprehensive attribute difference values of the minimized slab pair needs to be considered, so that the final production optimization target is obtained as follows, wherein xi,jIndicating whether the path is selected, if so, 1, otherwise, 0, costi,jThe comprehensive property difference between slabs i and j is as follows:
(4.22) in order to ensure the uniqueness of each slab in the production plan, each slab only needs to be a precursor node of any slab in the slab library except for the slab, and the specific constraints are as follows:
(4.23) and (4.22) similarly, in order to ensure the uniqueness of each slab in the production plan, each slab needs to be only a successor node of any slab in the slab library except for the slab, and specific constraints are as follows:
(4.24) in order to ensure that the number of production lines is the same as the actual number, a super source point s and a super sink point d are arranged, and the number of paths from the super source point is equal to the number of production lines | K |, xs,iWhether a path from s to i is selected or not is represented, if so, the path is 1, otherwise, the path is 0, and specific constraints are as follows:
(4.25) thus far, a general vehicle path problem model was established.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110230369.1A CN112861371B (en) | 2021-03-02 | 2021-03-02 | Steel industry cloud production scheduling method based on edge computing |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110230369.1A CN112861371B (en) | 2021-03-02 | 2021-03-02 | Steel industry cloud production scheduling method based on edge computing |
Publications (2)
Publication Number | Publication Date |
---|---|
CN112861371A true CN112861371A (en) | 2021-05-28 |
CN112861371B CN112861371B (en) | 2022-11-18 |
Family
ID=75990958
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110230369.1A Active CN112861371B (en) | 2021-03-02 | 2021-03-02 | Steel industry cloud production scheduling method based on edge computing |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112861371B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117575216A (en) * | 2023-11-15 | 2024-02-20 | 淄博京科电气有限公司 | Intelligent factory management method and system based on Internet of things and industrial big data |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101097617A (en) * | 2006-06-30 | 2008-01-02 | 东北大学 | Metal hot rolling optimizing scheduling method and system thereof |
CN108053047A (en) * | 2017-10-26 | 2018-05-18 | 北京航天智造科技发展有限公司 | Cloud resources of production scheduling methods, devices and systems |
CN111585916A (en) * | 2019-12-26 | 2020-08-25 | 国网辽宁省电力有限公司电力科学研究院 | LTE electric power wireless private network task unloading and resource allocation method based on cloud edge cooperation |
-
2021
- 2021-03-02 CN CN202110230369.1A patent/CN112861371B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101097617A (en) * | 2006-06-30 | 2008-01-02 | 东北大学 | Metal hot rolling optimizing scheduling method and system thereof |
CN108053047A (en) * | 2017-10-26 | 2018-05-18 | 北京航天智造科技发展有限公司 | Cloud resources of production scheduling methods, devices and systems |
CN111585916A (en) * | 2019-12-26 | 2020-08-25 | 国网辽宁省电力有限公司电力科学研究院 | LTE electric power wireless private network task unloading and resource allocation method based on cloud edge cooperation |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117575216A (en) * | 2023-11-15 | 2024-02-20 | 淄博京科电气有限公司 | Intelligent factory management method and system based on Internet of things and industrial big data |
Also Published As
Publication number | Publication date |
---|---|
CN112861371B (en) | 2022-11-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Tang et al. | Flexible job-shop scheduling with tolerated time interval and limited starting time interval based on hybrid discrete PSO-SA: An application from a casting workshop | |
CN101097617B (en) | Metal hot rolling optimizing scheduling method and system thereof | |
CN102637272A (en) | Method and device for distribution of residual plate blanks in hot-rolled plate blank front storeroom | |
CN102136104A (en) | Load balance and Lin-Kernighan (LK) algorithm based vehicle route planning method | |
CN107807623A (en) | Flexible machine adds the Dynamic Scheduling method, system and the application in high ferro motor-car parts group flexible machine adds production line of production line | |
CN103065207B (en) | Based on the engine cylinder-body process route optimization method of processing unit | |
CN107993052A (en) | More investment subject My Perspective On The Co-building project Dynamic Schedule control methods | |
CN106097055A (en) | Enterprise order processing method under personalized customization demand | |
CN112861371B (en) | Steel industry cloud production scheduling method based on edge computing | |
Wang et al. | Optimized skill configuration for the seru production system under an uncertain demand | |
CN101727622A (en) | layering modeling and optimizing method targeting complicated manufacture system | |
CN115796328A (en) | Product process route simulation system and product performance prediction method | |
CN117132181B (en) | Distributed flexible production and transportation cooperative scheduling method | |
Barlatt et al. | Ford motor company implements integrated planning and scheduling in a complex automotive manufacturing environment | |
CN115456268A (en) | Guide roller manufacturing resource optimal allocation method, device, equipment and medium | |
Illgen et al. | Simulation based production support system in the field of steel construction for large offshore structures | |
Chen et al. | A research on construction project based on BIM | |
Tzeng et al. | New idea of multi-objective programming with changeable spaces for improving the unmanned factory planning | |
CN117391423B (en) | Multi-constraint automatic scheduling method for chip high multilayer ceramic package substrate production line | |
Zaritskyi et al. | Technology in the Industrial Revolutions and its Impact on the Key Performance Indicators of Organizations. | |
Zhou et al. | Simulation based capacity optimization of a general assembly line with extremely unbalanced station process time | |
CN107918805A (en) | A kind of intelligent aero-engine assembly shop scheduling system | |
Lijun et al. | AN OPTIMIZATION ALGORITHM FOR SUPPLY-DEMAND MATCHING AND PRODUCTION SCHEDULING IN FLEXIBLE WORKSHOPS BASED ON DYNAMIC NETWORK. | |
CN113434926B (en) | BIM-based method and system for optimizing building implementation | |
Toliušienė et al. | The study on implementation of agile manufacturing system in Lithuanian industry |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |