CN112650187B

CN112650187B - Workshop scheduling method, device and system

Info

Publication number: CN112650187B
Application number: CN202110088152.1A
Authority: CN
Inventors: 熊辉; 刘检华; 庄存波; 曹远冲; 张雷; 宁伟航
Original assignee: Beijing Institute of Technology BIT
Current assignee: Beijing Institute of Technology BIT
Priority date: 2021-01-22
Filing date: 2021-01-22
Publication date: 2022-04-05
Anticipated expiration: 2041-01-22
Also published as: CN112650187A

Abstract

The invention provides a workshop scheduling method, device and system, and relates to the technical field of computer processing. The workshop scheduling method comprises the following steps: acquiring workshop scheduling data; generating a first scheduling scheme according to the workshop scheduling data; executing the first scheduling scheme; monitoring whether a disturbance condition exists in a scheduling process; if the disturbance condition exists, generating a second scheduling scheme, and updating the first scheduling scheme according to the disturbance condition; judging whether a rescheduling condition is met or not according to the updated first scheduling scheme and the second scheduling scheme; and if the rescheduling condition is met, executing the second scheduling scheme, and if the rescheduling condition is not met, executing the updated first scheduling scheme. The workshop scheduling method, the device and the system improve the accuracy of workshop dynamic scheduling and can well meet the new workshop scheduling requirement.

Description

Workshop scheduling method, device and system

Technical Field

The invention relates to the technical field of computer processing, in particular to a workshop scheduling method, device and system.

Background

Workshop scheduling is a key link in the production process of a workshop, and aims to allocate a group of jobs (usually procedures) to matching machines (or teams) under consideration of a series of constraints so as to optimize given targets, such as total completion time, total delay time, total energy consumption, resource utilization rate and the like. The prior art research on the dynamic scheduling problem mainly focuses on driving the workshop scheduling by real-time or offline production data, that is, researching the dynamic scheduling problem by using the relation between the data-driven information space and the physical space composed of real entities. However, at present, the connection between the information space and the physical space usually needs human intervention, the data of the information space and the physical space are difficult to interact in real time, and the dynamic property, the intelligence and the predictability of the workshop scheduling are insufficient. How to realize continuous and dynamic optimization of workshop scheduling through data interaction of an information physical space is a major challenge in the field of workshop scheduling.

The digital twin technology can realize the interactive fusion of a physical world and an information world of a manufacturing workshop, a user can acquire real-time information of the physical workshop in the virtual workshop and perform simulation prediction on the basis, and a prediction result of the virtual workshop can be received in the physical workshop and adjusted in time on the basis. Thus, the digital twin provides a new approach to achieve more autonomous and proactive plant scheduling. However, at present, scheduling based on the digital twin is mostly oriented to common job shops, the influence of the accuracy of working hours and dynamic changes on the scheduling is rarely considered, and how to realize the dynamic scheduling of the discrete assembly shop of the complex product by using the digital twin under the condition of considering the accuracy of the working hours and the dynamic changes is still a subject worth of research, namely how to accurately acquire various scheduling resource information, and realizing accurate shop scheduling is an important problem facing at present.

Disclosure of Invention

The embodiment of the invention provides a workshop scheduling method, a workshop scheduling device and a workshop scheduling system, which aim to solve the problem of insufficient accuracy of dynamic workshop scheduling in the prior art.

In order to solve the technical problems, the invention adopts the following technical scheme:

the embodiment of the invention provides a workshop scheduling method, which comprises the following steps:

acquiring workshop scheduling data;

generating a first scheduling scheme according to the workshop scheduling data;

executing the first scheduling scheme;

monitoring whether a disturbance condition exists in a scheduling process;

if the disturbance condition exists, generating a second scheduling scheme, and updating the first scheduling scheme according to the disturbance condition;

judging whether a rescheduling condition is met or not according to the updated first scheduling scheme and the second scheduling scheme;

and if the rescheduling condition is met, executing the second scheduling scheme, and if the rescheduling condition is not met, executing the updated first scheduling scheme.

Optionally, the plant scheduling data includes: the process information of the current process, the current scheduling plan information, the historical information of workshop equipment and the real-time information of the workshop equipment.

Optionally, the generating a first scheduling scheme according to the plant scheduling data includes:

preprocessing the workshop scheduling data to obtain a vector which can be input into a prediction model;

inputting the vector to a constructed prediction model, and determining completion time information taking the shortest completion time as an optimization target;

establishing a scheduling model according to the completion time information, wherein the constraint conditions of the scheduling model are as follows: the workshop equipment prohibits processing the workpiece within a preset time;

under the condition of meeting the constraint condition, performing iterative operation on the scheduling model to determine the shortest target completion time;

and generating the first scheduling scheme according to the shortest target completion time.

Optionally, the establishing a scheduling model according to the time completion information includes:

establishing a first objective function with the shortest completion time as follows: MinC_max，

Establishing a second objective function for calculating the completion time as follows:

the third objective function for establishing the time sequence of each assembly process is:

a fourth objective function that establishes uniqueness of the assembly equipment for each assembly process of each workpiece is:

the fifth objective function required to establish the start-up time of each process after the zero time is as follows:

t_jk≥0,

wherein n represents the total number of workpieces, and J represents a workpiece set to be assembled; j represents a workpiece number; s represents the total number of steps; s represents a process set; k represents a process number; m^(k)The number of the optional assembling equipment of the k procedure is shown; l represents the total number of the assembling devices; t is t_jkThe work starting time of the k-th procedure with the work j is shown; c_maxRepresenting a maximum completion time; m represents an assembly equipment number; x is the number of_jkmExpressed in the variable 0-1, the k-th process with the workpiece j is 1 if the k-th process is allocated to the assembly equipment with m, otherwise, the k-th process is 0; p is a radical of_jkmWhich represents the time required for the process of the workpiece j to be assembled by the assembling apparatus m.

Optionally, the constraint condition includes:

each workpiece of the workshop equipment can be processed at zero time, and the preparation time before assembly of the assembly process of each workpiece is 0;

the same workpiece is assembled by an assembling device at the same time of the same process;

the assembling equipment of each assembling process has uniqueness;

the assembly processes of different workpieces are not constrained by the assembly sequence, and the time sequence of the assembly processes of the same workpiece is executed according to the first scheduling scheme.

Optionally, the performing iterative operation on the scheduling model to determine the shortest target completion time includes:

generating an initial population code according to the workshop scheduling data and the parameter information generated by the scheduling model;

performing iterative operation according to a genetic algorithm to obtain the fitness value of each individual in each population, wherein the iterative operation comprises the following steps: performing iterative operation on cross variation and updating the population codes;

and generating the shortest target completion time according to each individual fitness value of the last iterative operation until the iteration times reach a set first threshold.

Optionally, the determining whether the rescheduling condition is met according to the updated first scheduling scheme and the updated second scheduling scheme includes:

detecting whether the second scheduling scheme meets constraint conditions, and determining that rescheduling conditions are met under the condition that the second scheduling scheme does not meet the constraint conditions; alternatively, the first and second electrodes may be,

determining that a rescheduling condition is met if the second scheduling scheme meets a constraint condition and if | C1-C2| is greater than Tmax is met;

wherein C1 represents the maximum completion time of the first scheduling scheme after updating, C2 represents the maximum completion time of the second scheduling scheme, and Tmax represents a preset threshold.

Optionally, if the first scheduling scheme is updated according to the disturbance condition, the updating includes:

determining a disturbance type according to the disturbance condition;

updating the first scheduling scheme according to the disturbance type;

wherein the disturbance types include: at least one of process constraint modification, plant resource constraint modification, and assembly execution time error.

Optionally, in a case that the disturbance type includes a process constraint change, the updating the first scheduling scheme according to the disturbance type includes:

adjusting the time sequence of the changed current process to be the maximum priority;

and sequentially setting the time sequence of other process time affected by the changed current process and the time sequence of other process time affected by the changed current process, and outputting and updating the first scheduling scheme in the current execution process.

Optionally, when the disturbance type includes a change in plant resource constraint, the updating the first scheduling scheme according to the disturbance type includes:

determining whether spare resources exist after the workshop resource constraint is changed;

if the standby resources exist, determining standby replacement time, sequentially postpositing the time sequence of the affected process time to the standby replacement time, and outputting and updating a first scheduling scheme in the current execution process;

and if the standby resources do not exist, determining resource supplement time, sequentially postpositing the time sequence of the affected process time by the resource supplement time, and outputting and updating the first scheduling scheme in the current execution process.

Optionally, when the disturbance type includes an assembly execution time error, the updating the first scheduling scheme according to the disturbance type includes:

if the time error of the current process is the completion time advance, determining whether the advance time is greater than a second threshold value;

if the time sequence is larger than the second threshold, the priority of the time sequence of the affected process time is advanced, and a first scheduling scheme in the current execution process is updated in an output mode; alternatively, the first and second electrodes may be,

if the time error of the current process is the completion time delay, determining whether the delayed time is the time of the previous process or not;

and if the time sequence of the affected process time is greater than the third threshold, setting the time sequence of the affected process time as the priority, and outputting and updating the first scheduling scheme in the current execution process.

Optionally, the method further includes: and when the disturbance condition does not exist, the scheduling task is completed and the completion time information is output.

An embodiment of the present invention further provides a workshop scheduling apparatus, including:

the acquisition module is used for acquiring workshop scheduling data;

the generating module is used for generating a first scheduling scheme according to the workshop scheduling data;

an execution module to execute the first scheduling scheme;

the monitoring module is used for monitoring whether a disturbance condition exists in the scheduling process;

the first processing module is used for generating a second scheduling scheme if a disturbance condition exists, and updating the first scheduling scheme according to the disturbance condition;

the second processing module is used for judging whether a rescheduling condition is met or not according to the updated first scheduling scheme and the second scheduling scheme;

and the third processing module is used for executing the second scheduling scheme if the rescheduling condition is met, and continuing to execute the updated first scheduling scheme if the rescheduling condition is not met.

The embodiment of the present invention further provides a workshop scheduling system, including: a processor, a memory and a program stored on the memory and executable on the processor, the program, when executed by the processor, implementing the steps of the plant scheduling method as described above.

The embodiment of the invention also provides a readable storage medium, wherein the readable storage medium stores a program, and the program realizes the steps of the workshop scheduling method when being executed by a processor.

The invention has the beneficial effects that:

in the technical scheme, a first scheduling scheme is generated through acquired workshop scheduling data, and whether a disturbance condition exists or not is monitored in the scheduling process of executing the first scheduling scheme; if the disturbance condition exists, generating a second scheduling scheme, and updating the first scheduling scheme according to the disturbance condition; judging whether a rescheduling condition is met or not according to the updated first scheduling scheme and the second scheduling scheme; and if the rescheduling condition is met, executing the second scheduling scheme, and if the rescheduling condition is not met, executing the updated first scheduling scheme. The method provided by the invention improves the precedent, initiative and accuracy of workshop scheduling.

Drawings

FIG. 1 is a flow chart of a method for scheduling a plant according to an embodiment of the present invention;

FIG. 2 is a schematic structural diagram of a plant scheduling system according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of a neural network model;

FIG. 4 is a schematic diagram of a chromosome two-layer encoding method according to an embodiment of the present invention;

FIG. 5 is a simplified flowchart of a method for scheduling a plant according to an embodiment of the present invention;

FIG. 6 is a flow chart illustrating an iterative calculation process according to an embodiment of the present invention;

FIG. 7 is a second flowchart illustrating a workshop scheduling method according to an embodiment of the present invention;

FIG. 8 is a flowchart illustrating a method for scheduling a plant under disturbance according to an embodiment of the present invention;

fig. 9 is a schematic block diagram of a plant scheduling apparatus according to an embodiment of the present invention.

Detailed Description

In order to make the technical problems, technical solutions and advantages of the present invention more apparent, the following detailed description is given with reference to the accompanying drawings and specific embodiments. In the following description, specific details such as specific configurations and components are provided only to help the full understanding of the embodiments of the present invention. Thus, it will be apparent to those skilled in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the invention. In addition, descriptions of well-known functions and constructions are omitted for clarity and conciseness.

It should be appreciated that reference throughout this specification to "one embodiment" or "an embodiment" means that a particular feature, structure or characteristic described in connection with the embodiment is included in at least one embodiment of the present invention. Thus, the appearances of the phrases "in one embodiment" or "in an embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

In various embodiments of the present invention, it should be understood that the sequence numbers of the following processes do not mean the execution sequence, and the execution sequence of each process should be determined by its function and inherent logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

As a basis for the production of workshops, shop scheduling plays an important role in the manufacturing industry. The system can make full use of various existing production resources in a workshop and reasonably distribute production and processing tasks, thereby improving the production efficiency of the workshop and ensuring the long-term stable operation of the production process. On one hand, due to the diversity of workshop production resources and the large amount of data related to workshop scheduling, how to accurately acquire various scheduling resource information and realize accurate workshop scheduling is an important problem at present. On the other hand, due to the complexity of the workshop production process, the workshop production scheduling parameters can change continuously, so that the static scheduling plan is difficult to ensure the accuracy of the plan. At the same time, uncertain dynamic disturbance events such as machine faults, orders, delivery date changes, etc. often occur in the plant. These dynamic disturbances will cause the production process to deviate from the plan, affecting production execution efficiency. Therefore, the realization of the interactive fusion of the physical space and the information space of the workshop is the key for realizing the workshop scheduling.

Aiming at the problem that the accuracy of dynamic workshop scheduling in the prior art is not enough, the invention provides a workshop scheduling system based on digital twin spaceflight discrete assembly, researches on-line working hour prediction based on real-time data and a neural network and a dynamic scheduling method oriented to spaceflight discrete digital twin workshops, provides an idea for realizing the dynamic workshop scheduling with precedent and initiative, and provides a workshop scheduling method, device and system.

As shown in fig. 2, the present invention provides a plant scheduling system, which mainly comprises four parts: the method comprises the following steps of physical workshop 1, virtual workshop 2, twin workshop data 3 and workshop service 4, wherein the operation flow is as follows: taking the twin data 3 of the workshop as a center, a production execution system of the physical workshop 2 can sense and acquire scheduling related information in workshop production in real time, such as various scheduling data of equipment operation information, personnel on duty information, processing task information, production state information and the like, the scheduling data is fed back to the corresponding virtual workshop 2 through the center of the twin data 3 of the workshop, the virtual workshop 2 triggers an updating optimization process by combining historical scheduling data on the basis of the real-time sensed scheduling data, simultaneously uploads simulation data generated in the updating process to the center of the twin data 3 of the workshop, the fused twin data can drive various workshop services 4, including production information statistics 41, workshop state monitoring 42, man-hour online prediction 43, real-time scheduling scheme generation 44 and the like, the output of the workshop services is fed back to the physical workshop through the twin data center for execution, and forming a workshop scheduling process of continuously iterative evolution of virtuality and reality.

Under the workshop scheduling system, on one hand, various data required by production scheduling can be accurately obtained, including real-time data and historical data, so that the accuracy of workshop scheduling is ensured; on the other hand, the workshop state and the assembly working hours can be analyzed based on twin data, and scheduling scheme adjustment and production disturbance timely can be responded to through continuous interaction and feedback between the reality and the virtues.

Of course, as shown in fig. 1, the present invention further provides a workshop scheduling method, including:

step 100, acquiring workshop scheduling data;

it should be noted that the plant scheduling data includes: the process information of the current process, the current scheduling plan information, the historical information of workshop equipment and the real-time information of the workshop equipment.

200, generating a first scheduling scheme according to the workshop scheduling data;

here, the first scheduling plan is a plan generated for the first time based on the plant scheduling data, that is, an initial plan generated immediately before the start of the job.

Step 300, executing the first scheduling scheme;

step 400, monitoring whether a disturbance condition exists in a scheduling process;

step 500, if a disturbance condition exists, generating a second scheduling scheme, and updating the first scheduling scheme according to the disturbance condition;

step 600, judging whether a rescheduling condition is met according to the updated first scheduling scheme and the second scheduling scheme;

step 700, if the rescheduling condition is satisfied, executing the second scheduling scheme, and if the rescheduling condition is not satisfied, executing the updated first scheduling scheme.

In this embodiment, in executing the first scheduling scheme, steps 400 to 700 monitor whether a disturbance condition exists in the scheduling process; if no disturbance occurs, continuing to execute the current first scheduling scheme, if abnormal disturbance occurs, generating a service generation real-time scheduling scheme, namely a second scheduling scheme, updating the current actually-executed scheduling scheme according to the disturbance condition, namely generating an updated first scheduling scheme, comparing the generated second scheduling scheme with the updated first scheduling scheme, and judging whether rescheduling is needed; if rescheduling is needed, determining the second scheduling scheme as a real-time scheduling scheme, otherwise, continuously executing the current scheme, namely the updated first scheduling scheme; and the process is continuously circulated until the whole assembly task is finished. The invention considers the disturbance condition in the execution process and improves the precedent, the initiative and the accuracy of workshop scheduling.

It should be appreciated that the assembly object physical attributes, assembly process factors, shop equipment and tooling factors, personnel factors, other factors, and the like. Currently, the influence of a scheduling scheme is less considered in the research of work hour prediction, but for an assembly workshop, particularly a discrete assembly workshop, each process needs different materials, different stations or machines are distributed more dispersedly, and the buffer memory position on each station is limited, so that some workpieces may not be immediately transferred to the station of the next process, and the transfer time of the materials between two adjacent processes of the same workpiece is different, and the work hours of the final process are influenced.

Optionally, the step 200 includes:

step 210, preprocessing the workshop scheduling data to obtain a vector which can be input into a prediction model;

in this embodiment, the purpose of data preprocessing is to convert original input data, that is, plant scheduling data, into vectors that can be input by a prediction model. For the physical attribute data of the assembly objects, the invention is characterized by the types and the number of the assembly objects, the assembly objects related to the invention are divided into 5 types, and the specific types are as follows: plate class, frame class, beam class, rod class, standard class, and others, and then represent the assembly object physical attributes with vector AsmObj ═[ O1, O2, …, Oi ], where Oi represents the number of assembly objects of the ith kind; for the assembling action characteristics, the assembling action characteristics are represented by the types and the subtypes of the assembling actions, the specific types of the assembling actions comprise 7 types of material picking, positioning, clamping, connecting, looseness preventing, sealing and cleaning, and then the assembling action characteristics are represented by vectors (A1, A2, … and Aj), wherein Aj represents the operation times of the jth assembling action; for precision requirement characteristics, the precision requirement characteristics are represented by types and precision grades of precision requirements, the precision types are mainly divided into 6 types of weight, distance, moment, parallelism, flatness and position degree, the precision grades are divided into three types of high, middle and low, then the precision requirement characteristics are represented by vectors AsmPrcs [ P1, P2, … and Pk ], wherein Pk represents the precision grade of the kth precision requirement, and Pk takes values of 3, 2, 1 and 0 and respectively represents high, middle, low and no precision requirements; for equipment and tool characteristics, the equipment and tool characteristics are represented by state parameters of used equipment in the current process, the state parameters comprise static state parameters and dynamic state parameters, then the equipment and tool characteristics are represented by vectors AsmDev [ Dm1, Dm2, … and Dml ], wherein Dml represents the l-th class state parameters of equipment with the number of m; the degree of influence of the proficiency of the operators on the working hours is small, so the influence of the operators on the working hours is not considered temporarily; in addition, the invention additionally considers the influence of the first scheduling scheme on the working hours and is characterized by the two-layer coding of the first scheduling scheme. Since the dimensions of the above features are not the same, the present invention finally opens and splices them into a one-dimensional vector.

Step 220, inputting the vector to the constructed prediction model, and determining completion time information taking the shortest completion time as an optimization target;

in this embodiment, the input of the vector to the constructed prediction model may predict the man-hours in real time, and may determine the completion time information with the shortest completion time as the optimization goal by combining the predicted man-hours and the constraint conditions.

It should be noted that the prediction model is established in advance, and the prediction model used in the present invention is a neural network model. Neural networks are mathematical algorithms which imitate brain nerve synapse working modes in the field of machine learning, and in popular terms, the neural networks are processes which approximate to real values by carrying out distributed function operation on model input and continuously iterating. Because of its excellent fitting ability to nonlinear relations, neural networks have very good applications in many fields.

It should be noted that, as shown in fig. 3, a typical neural network structure generally consists of an input layer x, a hidden layer h, and an output layer y, each layer consisting of several nodes (neurons). The input layer x is an independent variable layer of the model, and is a vector obtained by preprocessing input data in the present document. The hidden layer h refers to a layer for performing iterative computation between input and output, and the number of nodes and the number of layers of the hidden layer h are not limited and are artificially set. Generally, the more the number of hidden layers h is, the more the number of single-layer nodes is, and the stronger the fitting ability of the network is. However, the higher the complexity of the corresponding model, the higher the training cost, and therefore, comprehensive consideration is required. By comprehensively comparing the efficiency and the precision of network training, the number of the hidden layers h of the network is finally selected to be 3, and the number of the nodes is 256 multiplied by 128 multiplied by 64. The output layer y is a target value predicted by the neural network model, and if the researched problem belongs to a classification problem, the number of nodes of the output layer is consistent with the number of classes; if the regression problem, i.e. the predicted target value is a single value, is investigated, the number of output layers is 1. In the invention, the predicted working hours belong to the regression problem, so the dimension of the output layer is 1.

The parameter optimization method used in the training process of the prediction model is an Adam algorithm; the loss function of the training of the prediction model is the root mean square error due to the regression problem; the activation function of the prediction model hidden layer uses a Relu activation function which can avoid gradient diffusion, and the output layer uses a linear activation function. In the training process, 90% of all sample data is taken as a training sample, and 10% of the sample data is used for verifying the fitting accuracy of the model.

The embodiment runs the results on the test data set through step 200 on the trained prediction model, and the results show that the average error rate of the model in the initial test sample is within 10%, and also show that the loss values on the training set and the test set are stable around 10 rounds.

In order to further verify the accuracy of the prediction model, the invention combines the prediction model with the actual case to obtain partial prediction results, and as shown in table 1, the results show that the average error rate of the actual prediction is 14.6%. The invention can also periodically update the prediction model by adding the actual prediction case into the sample library, and the prediction result of the prediction model is far higher than the accuracy of the artificial experience prediction.

Table 1 partial actual prediction results

Step 230, establishing a scheduling model according to the completion time information, wherein the constraint conditions of the scheduling model are as follows: the workshop equipment prohibits processing the workpiece within a preset time;

the scheduling model established in the embodiment can solve the problem of workshop scheduling, namely two sub-problems contained in the problem of workshop assembly scheduling can be solved: determining the assembling equipment of each procedure (assembling equipment selection subproblem) and determining the sequence of the procedures to be assembled on each assembling equipment (procedure sequencing subproblem); by setting constraints on the scheduling model, the accuracy of the data can be further increased.

Step 240, performing iterative operation on the scheduling model under the condition that the constraint condition is met, and determining the shortest target completion time;

it should be noted that, in the field of plant scheduling, the genetic algorithm is widely applied due to its characteristics of simple operation, good convergence effect, and the like, and therefore, the genetic algorithm is used as an optimization algorithm for performing iterative operation on the scheduling model. The genetic algorithm is a random search algorithm simulating natural selection and evolution of organisms, has good global search capability, and can gradually obtain optimal individuals in the evolution process of continuous inheritance and selection, so that the shortest target completion time can be determined through iterative operation decoding.

And 250, generating the first scheduling scheme according to the shortest target completion time.

The first scheduling scheme generated in the embodiment can be applied to a plurality of processes in the assembly process of complex products, including but not limited to assembly of complex products such as satellites and missiles, and can solve the scheduling problem of a mixed assembly flow shop, each process in the assembly process can be regarded as a stage, and for each process, a plurality of assembly devices are provided for selection.

Specifically, the step 230 includes:

establishing a first objective function with the shortest completion time as follows: min C_max，

t_jk≥0,

It should be noted that the second objective function, the third objective function, the fourth objective function and the fifth objective function are all limiting functions of the first objective function, that is, the objective time with the shortest completion time can be determined by constructing the first objective function, and the second objective function, the third objective function, the fourth objective function and the fifth objective function all limit the use range of the first objective function.

In this embodiment, in addition to the original process sequence, the selection of the assembling equipment needs to be considered. Therefore, in the individual chromosome coding method provided by the present invention, the chromosome gene is divided into two layers, which respectively represent an Operation Sequence (OS) gene and a Machine Selection (MS) gene.

Specifically, as shown in fig. 4, in the OS layer, the value j at the position x corresponds to the workpiece j, and the jth time when j appears from left to right is the kth process Ojk of the workpiece j, in the MS layer, the value M at the corresponding position x indicates that the machine equipment selected by the corresponding process Ojk is Mm, the code at the position circled in fig. 4 indicates that the equipment selected by the second process of the workpiece 4 is M4, and the meaning of the codes at other positions can be obtained similarly.

When decoding, the start time and the completion time of each procedure are sequentially determined according to the OS layer sequence, wherein the start time tjk is determined by the later time in the completion time of the previous procedure of the workpiece corresponding to the current procedure (if the current procedure is the first procedure of the corresponding workpiece, the default time is zero time) and the earliest available time of related constraint resources (the constraint of equipment resources is mainly considered in the invention), for the condition that a plurality of procedures are simultaneously distributed to the same equipment, the processing sequence of the procedures on the equipment is determined according to the sequence of the procedures on the OS layer, and the completion time is determined by tjk + pjkm.

Specifically, the constraint conditions include:

the assembling equipment of each assembling process has uniqueness;

In this embodiment, each workpiece of the workshop appliance can be machined at zero time, the preparation time before assembly of the assembly process of each workpiece is 0, that is, all workpieces can be assembled at zero time, the preparation time before assembly of all workpiece assembly processes is 0, each process of each workpiece cannot be finished once assembly is started, and the time interval between adjacent processes of the same workpiece is not limited; the same workpiece is assembled by one assembling device at the same time of the same process, preferably, the same process of the same workpiece can be assembled by only one assembling device at the same time; the assembling equipment of each assembling procedure has uniqueness, namely, the same assembling equipment can only assemble one workpiece at a certain time; the assembling sequence of different workpieces is not restricted, the time sequence of the assembling process of the same workpiece is executed according to the first scheduling scheme, namely the assembling sequence of different workpiece processes is not restricted, the assembling sequence of the same workpiece needs to be executed according to the process flow, and all the processes of all the workpieces determine the time required by assembling after the assembling equipment is determined. The condition constraint of the embodiment provides accuracy guarantee for guaranteeing performance indexes (workpiece lag, equipment utilization rate and equipment energy consumption) of the scheduling scheme, and reduces the error between each process by predefining constraint standards.

Specifically, the assembly process and sequence of each workpiece are fixed on the site of an assembly workshop, optional assembly equipment needs to be distributed to each process according to an assembly process flow, n workpieces need to be assembled through s processes with the same assembly sequence, and the kth stage is provided with M (k) different assembly equipment; each process has one or more different assembling devices for selection; the process assembly time varies depending on the performance and state of the different assembly equipment. The scheduling target is to select the most suitable assembling equipment for each process, determine the optimal processing sequence and the start time of each process of each workpiece, and make the assembling process of the workpieces meet the production expectation.

Specifically, as shown in fig. 5, the process of generating the first scheduling plan can be known through a simple flow diagram, that is, the workshop scheduling data is acquired to include process information of the current process, current scheduling plan information, and history and real-time state information of workshop equipment, the workshop scheduling data is preprocessed, and then a prediction result is output through a prediction model, and then the result is used for generating the first scheduling plan, if the first scheduling plan is finally executed, the obtained actual working hours are used for updating the prediction model, thereby completing the closed-loop flow of the model.

In an alternative embodiment, the step 240 includes:

Specifically, as shown in fig. 6, according to the plant scheduling data and the parameter information generated by the scheduling model, an initial population code, that is, a parameter setting, is generated, which includes a maximum iteration number MI, an algorithm-related parameter, scheduling original data, and the like, and then the population code is initialized, and the iteration number (gen) is set to 1; performing iterative operation according to a genetic algorithm to obtain the fitness value of each individual in each population, wherein the iterative operation comprises the following steps: performing iterative operation to cross variation and update a population code, namely in the first iterative operation, evaluating an adaptive value of an individual in a population, judging whether the iteration frequency reaches a set first threshold value, namely whether gen is greater than the maximum iteration frequency MI, if gen is less than or equal to the maximum iteration frequency MI, performing iterative operation, namely performing cross variation operation, updating the population code, updating the iteration frequency, wherein gen is gen +1, namely after each iterative operation, the iteration frequency is increased once, each individual adaptive value (the adaptive value of the individual in the population is evaluated) is obtained in each iterative operation until gen is greater than MI, outputting the optimal population adaptive value and a scheduling solution, generating the shortest target completion time, and drawing a Gantt chart which is used for displaying a scheme for generating the shortest target completion time.

In an alternative embodiment, the step 600 includes:

In the embodiment, under the condition of disturbance, the second scheduling scheme generated by the real-time scheduling scheme and the updated first scheduling scheme under the condition of disturbance judge whether rescheduling is needed or not by comparison, so that which scheme is more accurate subsequently is determined, the processing on actual data and the processing on the disturbance condition are improved, the idle time of each machine can be reduced, the utilization rate of the machine is improved, and the maximum completion time is reduced. On the other hand, the machine state and the real-time working hour can be predicted, and the disturbance condition is considered in advance, so that the initial scheme is more stable.

Optionally, the method further includes: and step 800, when the disturbance condition does not exist, until the scheduling task is completed and the completion time information is output. Here, the completion time information in the current scheme can be directly acquired, and the completion time information can be updated in real time according to the information in the actual execution process.

Specifically, as shown in the system diagram of fig. 2 and the workshop scheduling flowchart shown in fig. 7, it can be clearly known that, firstly, workshop scheduling data, i.e., scheduling task original data, workshop history, real-time data, etc., are obtained; generating a first scheduling scheme according to the workshop scheduling data, namely calling a real-time generation service to generate an initial scheduling scheme; then, executing the first scheduling scheme, namely executing the current scheduling scheme; monitoring whether a disturbance condition exists in a scheduling process, namely calling a workshop state monitoring service to judge whether the disturbance exists, if the disturbance condition exists, generating a second scheduling scheme, namely calling a real-time scheduling service to generate a real-time scheduling scheme, and updating the first scheduling scheme according to the disturbance condition, namely updating a current actual scheduling scheme according to the disturbance condition; judging whether a rescheduling condition is met or not according to the updated first scheduling scheme and the second scheduling scheme, namely judging whether rescheduling is needed or not, if the rescheduling condition is met, executing the second scheduling scheme, updating the current scheduling scheme to an actual scheduling scheme, namely updating the executing first scheduling scheme to the second scheduling scheme, and if the rescheduling condition is not met, executing the updated first scheduling scheme; and when the disturbance condition does not exist, calling the production information statistical service to judge whether the task is finished or not until the scheduling task is finished and outputting the completion time information.

In an optional embodiment, in the step 500, if the first scheduling scheme is updated according to the disturbance condition, the method includes:

step 510, determining a disturbance type according to the disturbance condition;

step 520, updating the first scheduling scheme according to the disturbance type;

It should be noted that the types include: the method comprises the steps that at least one of process constraint change, workshop resource constraint change and assembly execution time error is achieved, wherein disturbance events of the process constraint change comprise task requirement completion time advance, emergency insertion, task cancellation and the like, disturbance events of the workshop resource constraint change comprise instrument and equipment faults, material loss and the like, disturbance events of the assembly execution time error comprise material distribution delay, execution time error, quality problem troubleshooting and the like, according to each disturbance type, different first scheduling schemes can be updated, and diversity of processing procedures is reflected.

As shown in fig. 8, specifically, in the case that the disturbance type includes a process constraint change, the step 520 includes:

step 521, adjusting the time sequence of the changed current process to be the maximum priority;

and 522, sequentially setting the time sequence of other process time affected by the changed current process and the priority of the other process time, and outputting and updating the first scheduling scheme in the current execution process.

In this embodiment, when the disturbance type includes a process constraint change, and when the process constraint change requirement completion time is advanced, it is determined whether the current time meets a task requirement, if not, the time sequence of the current process after the change is adjusted to the maximum priority to shift the time sequence of the process to be completed in advance on the current equipment, and it is determined again whether the current time meets the task requirement, and if so, the time sequence of the time of the other process affected by the current process after the change is sequentially prioritized and then placed, that is, the process table is traversed, and the time sequence of the time of the other process affected by the current process is sequentially shifted back. When the process constraint is changed into the urgent order, the time sequence of the current process after the change is adjusted to be the maximum priority, namely, the process in the urgent task list is sequentially inserted into the assembly equipment which is idle firstly, then the work order table is traversed, and the time sequence of the time of other processes influenced by the current process is sequentially moved backwards. When the process constraint is changed into task cancellation, the time sequence of the current working procedure after the change is adjusted to be the maximum priority, namely, the working procedure table is traversed, and the working procedures behind the cancelled task are moved forward in sequence.

As shown in fig. 8, specifically, in the case that the disturbance type includes a change of a plant resource constraint, the step 520 includes:

In this embodiment, if a standby resource exists, a standby replacement time is determined, the time sequence of the affected process time is sequentially set back to the standby replacement time, that is, the standby resource is replaced by the current resource, the standby replacement time is obtained, the time sequence of the affected process time is sequentially set back to the standby replacement time, the process table is traversed, the time sequences of the affected other process times of the current process are sequentially set back, and a first scheduling scheme for updating the current execution process is output. And if the standby resources do not exist, determining resource supplement time, sequentially postpositing the time sequence of the affected process time on the resource supplement time, traversing the process table, sequentially retrograding the time sequences of the affected other process times of the current process, and outputting and updating a first scheduling scheme in the current execution process.

As shown in fig. 8, specifically, in the case that the disturbance type includes an assembly execution time error, the step 520 includes:

In this embodiment, if the time error of the current process is the completion time advance, it is determined whether the advance time is greater than a second threshold, and if the advance time is greater than the second threshold, the priority of the time sequence of the affected process time is advanced, that is, the time sequence of the affected process time is advanced in sequence, the process table is traversed, the time sequences of the affected other process times of the current process are advanced in sequence, and the first scheduling scheme in the current execution process is updated by output. Or, if the time error of the current process is the completion time delay, if the time error is greater than a third threshold, the time sequence of the affected process time is sequentially arranged backwards, the process table is traversed, the time sequences of the affected other process time of the current process are sequentially arranged backwards, and the first scheduling scheme in the current execution process is output and updated.

The invention also provides a specific embodiment, in order to verify the effectiveness of the method, a complex product discrete assembly workshop scheduling system is constructed, and a scheduling scheme using the scheme is compared with a scheduling scheme not using the scheme by taking a certain actual assembly scheduling task of a workshop as an example.

The data in table 2 is the empirically estimated assembly time required for each workpiece process to use the corresponding machine, and this time does not take into account the actual machine state and the scheduling scheme.

TABLE 2 task-related parameters

In the actual implementation, the system detects that M5 is out of order at a certain time, and the time required for maintenance is 2 hours. At the moment, the method using the scheme generates a disturbed current scheduling scheme and a current scheduling scheme corresponding to the real-time optimization scheme and not meeting the constraint condition, and automatically triggers rescheduling, namely, the current scheduling scheme is replaced by the real-time optimization scheme, and for the current scheduling scheme meeting the constraint condition, whether rescheduling is required to be judged by combining the maximum completion time difference value and the reference difference value; in addition, in the actual execution process, if rework is required due to defective inspection after the third process M8 of the workpiece 3 is completed, the process needs to be re-executed. The method without the scheme does not consider the disturbance situation, and only judges according to the artificial influence, so that the accuracy is insufficient.

As shown in table 3, the method using the above scheme is compared with the method without the above scheme, wherein the maximum completion time refers to the completion time of the last process in the final scheduling scheme, and the overlap ratio of the final scheme and the initial scheme is obtained by calculating the ratio of the processes with the same completion time of the final scheme and the initial scheme, and is used for evaluating the stability of the initial scheduling scheme.

TABLE 4 comparison of methods

As can be seen from table 3, the method using the above scheme can trigger rescheduling earlier based on the continuous comparison of the physical plant and its dynamic reference virtual plant, which can reduce the idle time of each machine, improve the machine utilization and thus reduce the maximum completion time. On the other hand, the method of the scheme can consider the disturbance in advance, so that the initial scheme is more stable.

In conclusion, the method provided by the invention considers the disturbance situation in the actual operation process, judges the rescheduling condition according to the updated first scheduling scheme and the generated second scheduling scheme, updates the scheme in the operation process in real time, and improves the precedent, the initiative and the accuracy of workshop scheduling.

As shown in fig. 9, an embodiment of the present invention further provides a plant scheduling apparatus, including:

the acquiring module 10 is used for acquiring workshop scheduling data;

a generating module 20, configured to generate a first scheduling scheme according to the workshop scheduling data;

an execution module 30, configured to execute the first scheduling scheme;

the monitoring module 40 is used for monitoring whether a disturbance condition exists in the scheduling process;

a first processing module 50, configured to generate a second scheduling scheme if a disturbance condition exists, and update the first scheduling scheme according to the disturbance condition;

a second processing module 60, configured to determine whether a rescheduling condition is met according to the updated first scheduling scheme and the second scheduling scheme;

a third processing module 70, configured to execute the second scheduling scheme if the rescheduling condition is met, and continue to execute the updated first scheduling scheme if the rescheduling condition is not met.

Optionally, the generating module 20 includes:

the acquisition unit is used for preprocessing the workshop scheduling data and acquiring a vector which can be input into a prediction model;

the first determining unit is used for inputting the vector to the constructed prediction model and determining the completion time information taking the shortest completion time as an optimization target;

the establishing unit is used for establishing a scheduling model according to the completion time information, and the constraint conditions of the scheduling model are as follows: the workshop equipment prohibits processing the workpiece within a preset time;

the second determining unit is used for performing iterative operation on the scheduling model under the condition that the constraint condition is met, and determining the shortest target completion time;

and the generating unit is used for generating the first scheduling scheme according to the shortest target completion time.

Optionally, the establishing unit includes:

a first establishing subunit, configured to establish a first objective function with a shortest completion time as:

min C_max，

a second establishing subunit for establishing a second objective function for calculating the completion time as follows:

a third establishing subunit, configured to establish a third objective function of the time sequence of each assembly process as follows:

a fourth establishing subunit, configured to establish a fourth objective function unique to the assembling apparatus of each assembling process for each workpiece, as follows:

a fifth establishing subunit, configured to establish a fifth objective function that is required to be after the zero time for the start-up time of each process, as follows:

t_jk≥0,

It should be noted that the constraint conditions include:

the assembling equipment of each assembling process has uniqueness;

Optionally, the second determining unit includes:

the first generation subunit is used for generating an initial population code according to the workshop scheduling data and the parameter information generated by the scheduling model;

a first obtaining subunit, configured to perform an iterative operation according to a genetic algorithm to obtain a fitness value of each individual in each population, where the iterative operation includes: performing iterative operation on cross variation and updating the population codes;

and the second generation subunit is used for generating the shortest target completion time according to each individual fitness value of the last iterative operation until the iteration number reaches a set first threshold value.

Optionally, the second processing module 60 includes:

a third determining unit, configured to detect whether the second scheduling scheme satisfies a constraint condition, and determine that a rescheduling condition is satisfied if the second scheduling scheme does not satisfy the constraint condition; alternatively, the first and second electrodes may be,

a fourth determining unit, configured to determine that a rescheduling condition is satisfied if the second scheduling scheme satisfies a constraint condition and if | C1-C2| is greater than Tmax;

Optionally, the first processing module 50 includes:

the determining submodule is used for determining the disturbance type according to the disturbance condition;

the updating submodule is used for updating the first scheduling scheme according to the disturbance type;

Optionally, the update sub-module includes:

a first adjusting unit for adjusting the time sequence of the current process after the change to the maximum priority;

and the first processing unit is used for sequentially setting the time sequence of other process time affected by the changed current process and outputting and updating the first scheduling scheme in the current execution process.

Optionally, the update sub-module further includes:

a fifth determining unit, configured to determine whether there is a standby resource after the workshop resource constraint is changed;

the second processing unit is used for determining the standby replacement time if the standby resources exist, sequentially postpositing the time sequence of the affected process time to the standby replacement time and outputting and updating the first scheduling scheme in the current execution process;

and the third processing unit is used for determining the resource supplementing time if the standby resource does not exist, sequentially postpositing the time sequence of the affected process time on the resource supplementing time, and outputting and updating the first scheduling scheme in the current execution process.

Optionally, the update sub-module further includes:

a sixth determining unit, configured to determine whether the time of the advance is greater than a second threshold if the time error of the current process is the completion time advance;

the fourth processing unit is used for leading the priority of the time sequence of the affected process time and outputting and updating the first scheduling scheme in the current execution process if the priority is larger than the second threshold; alternatively, the first and second electrodes may be,

a seventh determining unit, configured to determine, if the time error of the current process is the completion time delay, whether the delayed time is a time determined to be advanced if not;

and the fifth processing unit is used for setting the time sequence of the affected process time as the rear priority if the time sequence is larger than the third threshold, and outputting and updating the first scheduling scheme in the current execution process.

Optionally, the apparatus further comprises: and the output module is used for finishing the scheduling task and outputting the completion time information when the disturbance condition does not exist.

The embodiment of the present invention further provides a readable storage medium, where a program is stored on the readable storage medium, and when the program is executed by a processor, the program implements each process of the above-described workshop scheduling method embodiment, and can achieve the same technical effect, and in order to avoid repetition, the detailed description is omitted here. The readable storage medium may be a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk.

Finally, it should also be noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or terminal apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.

While the preferred embodiments of the present invention have been described, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention as defined in the following claims.

Claims

1. A method for scheduling a plant, comprising:

acquiring workshop scheduling data;

generating a first scheduling scheme according to the workshop scheduling data;

executing the first scheduling scheme;

monitoring whether a disturbance condition exists in a scheduling process;

if the disturbance condition exists, generating a second scheduling scheme, and determining a disturbance type according to the disturbance condition; updating the first scheduling scheme according to the disturbance type; wherein the disturbance type comprises assembly execution time error and at least one of process constraint change and workshop resource constraint change; the disturbance events generating the assembly execution time errors comprise material distribution delay, execution time errors and quality problem troubleshooting;

if the rescheduling condition is met, executing a second scheduling scheme, and if the rescheduling condition is not met, executing an updated first scheduling scheme;

in the case that the disturbance type includes an assembly execution time error, the step of updating the first scheduling scheme according to the disturbance type includes:

2. The plant scheduling method of claim 1, wherein the plant scheduling data comprises: the process information of the current process, the current scheduling plan information, the historical information of workshop equipment and the real-time information of the workshop equipment.

3. The plant scheduling method of claim 1, wherein generating a first scheduling plan based on the plant scheduling data comprises:

4. The plant scheduling method according to claim 3, wherein the establishing a scheduling model according to the time-out information comprises:

5. The method according to claim 3, wherein the constraint condition comprises:

the assembling equipment of each assembling process has uniqueness;

6. The plant scheduling method according to claim 3, wherein the iterative operation of the scheduling model to determine the shortest target completion time comprises:

7. The method for scheduling a plant according to claim 5, wherein the step of determining whether the rescheduling condition is satisfied according to the updated first scheduling scheme and the updated second scheduling scheme comprises:

8. The plant scheduling method according to claim 1, wherein, in case the disturbance type includes a process constraint change, the updating the first scheduling scheme according to the disturbance type includes:

9. The plant scheduling method according to claim 1, wherein, in the case that the disturbance type includes a plant resource constraint change, the updating the first scheduling scheme according to the disturbance type includes:

10. The plant scheduling method of claim 1, further comprising: and when the disturbance condition does not exist, the scheduling task is completed and the completion time information is output.

11. A plant scheduling apparatus, comprising:

the acquisition module is used for acquiring workshop scheduling data;

an execution module to execute the first scheduling scheme;

the first processing module is used for generating a second scheduling scheme if a disturbance condition exists, and determining a disturbance type according to the disturbance condition; updating the first scheduling scheme according to the disturbance type; wherein the disturbance type comprises assembly execution time error and at least one of process constraint change and workshop resource constraint change; the disturbance events generating the assembly execution time errors comprise material distribution delay, execution time errors and quality problem troubleshooting;

the third processing module is used for executing the second scheduling scheme if the rescheduling condition is met, and continuing to execute the updated first scheduling scheme if the rescheduling condition is not met;

12. A plant room scheduling system, comprising: a processor, a memory and a program stored on the memory and executable on the processor, the program, when executed by the processor, implementing the steps of the plant scheduling method according to any one of claims 1 to 10.

13. A readable storage medium, characterized in that the readable storage medium has stored thereon a program which, when being executed by a processor, carries out the steps of the plant scheduling method according to any one of claims 1 to 10.