CN115808909A - Dynamic batch scheduling method for on-time and energy-saving production of die heat treatment - Google Patents

Abstract

A dynamic batch scheduling method for on-time and energy-saving production of die heat treatment comprises the following steps: periodically acquiring information of a workpiece to be heat-treated, and taking part of the workpiece in the rolling period as an external cooperation workpiece according to the rolling period in the information of the workpiece to be heat-treated; and performing on-line dispatching on the rest workpieces on a hot workshop. The invention provides a proactive periodical rolling scheduling, which decomposes a dynamic scheduling problem into a series of deterministic subproblems according to a time axis in sequence, and formulates a workpiece external cooperation plan and a pre-scheduling scheme in a period, thereby balancing the production load of a short-term heat treatment workshop; therefore, a heat treatment production plan in a future period is obtained, a reliable outsourcing decision is made, and the proactive of the production plan is realized.

Description

Dynamic batch scheduling method for on-time and energy-saving production of die heat treatment

Technical Field

The invention relates to the technical field of production scheduling, in particular to a dynamic batch scheduling method for on-time and energy-saving production of die heat treatment.

Background

Heat treatment is also a typical batch process, which is an important process in the manufacture of molds. The heat treatment production of the die is a key process which seriously influences the production progress and the manufacturing cost of the die. Most of the dies need heat treatment between rough machining and finish machining, which is an important process for ensuring the performance of the dies and directly influences the precision, the strength, the service life and the manufacturing cost of the dies. In actual production, the die manufacturing is often delayed in the heat treatment process, so the heat treatment often becomes the bottleneck of die manufacturing, and the punctuality and energy conservation of die production are greatly influenced. From the perspective of production management control (different from improving the heat treatment process or purchasing new equipment), it is of great practical significance to improve the operation efficiency of the mold production line by optimizing the batch scheduling scheme of mold heat treatment.

However, the mold heat treatment production is difficult to control in practice, and there are various problems: 1) There is a large variation between processing tasks. During the heat treatment, parts of different materials need to be heated to different critical temperatures, and the cooling time is different. This difference in the requirements of the heat treatment process results in an inability to batch at will between workpieces, and thus the mold heat treatment tasks are of the type typically incompatible with multiple workpiece families, and the tasks within the same workpiece family also differ in weight, lead time, and importance.

2) The arrival time of the processing task has high dynamic uncertainty. Different from the static batch scheduling problem, the processing tasks do not arrive at the same time at zero time, so that the processing tasks of the die heat treatment unit have dynamic arrival, therefore, the die heat treatment production control is also a dynamic batch scheduling problem, and not only the existing tasks but also the tasks to be arrived need to be considered at the scheduling time.

3) Enterprises often use an outsourcing processing mode to relieve the contradiction between production and demand. Because the order of the die enterprises often does not follow the uniform distribution, in order to ensure that the delivery date of the die is met as much as possible, when the actual production of a workshop exceeds the load, an outsourcing decision is often adopted, and part of workpieces which are possibly delayed are delivered to other enterprises for production, and the complexity of heat treatment management and control is further increased by the outsourcing decision.

4) The heat treatment production of the die is a key process for restricting delivery date and a high-energy consumption process, so that the scheduling optimization target is not only required to meet the optimization completion time, but also required to meet the energy-saving production target, and the heat treatment batch scheduling research of the die is required to be oriented to energy-saving and on-time manufacturing. It is because of these four main realistic features that the mold heat treatment production is often difficult to control effectively.

Disclosure of Invention

In view of the above-mentioned drawbacks, the present invention provides a method for dynamic batch scheduling of on-time and energy-saving production of mold heat treatment, so as to solve the problems of difficult on-schedule delivery, difficult external processing and difficult internal processing balance in the existing heat treatment workshop.

In order to achieve the purpose, the invention adopts the following technical scheme: a dynamic batch scheduling method for on-time and energy-saving production of die heat treatment comprises the following steps:

periodically acquiring information of a workpiece to be heat treated;

according to the rolling period in the information of the workpiece to be heat-treated, taking part of the workpiece in the rolling period as an external cooperation workpiece;

and performing on-line dispatching on the rest workpieces on a hot workshop.

Preferably, the step of obtaining the outside cooperation workpiece is as follows:

establishing a rolling scheduling mathematical model through the information of the workpiece to be heat-treated;

and solving the rolling scheduling mathematical model through an M-GEP algorithm to obtain a scheduling scheme of the outsourced workpieces.

Preferably, the specific steps of constructing the rolling scheduling mathematical model are as follows:

obtaining the production cost PI of the die workpiece j in the workpiece information to be heat-treated in the enterprise from the erp of the enterprise _j Exterior cost PO _j Calculating the total production cost in the rolling period and the extra cost for the outside cooperation;

setting the drag penalty of the workpiece according to the drag amount and the drag penalty coefficient of the workpiece;

predicting a stall penalty according to the total production cost, and establishing an objective function of a minimum stall penalty coefficient and an extra cost coefficient;

obtaining a scheduling scheme of the outside cooperation workpiece by analyzing the objective function;

the total production cost is as follows:

the additional cost in the scroll cycle is as follows:

where Nb is the number of workpieces in the rolling period, n _o Number n of outer cooperation workpieces _Em For number of emergency outsourcing workpieces, PI _j For the cost of producing the die work j inside the enterprise, em _ PO _j Is a processing time margin Tm _j Less than outside preparation time T _pre The outside cooperation cost of the die part, ord _ PO, is urgent _j The outside cooperation cost is met for the workpiece j under the condition of meeting the outside cooperation preparation time;

the hangover penalty is as follows:

wherein J _b For all the work sets contained in lot b,

The hold-off amount of the workpiece j in the batch b,

A drag penalty factor for the workpiece j in batch b;

the objective function is as follows:

Extra-Cost is the Extra Cost in the roll cycle, and Extra-Cost is the total production Cost, max-Tw _k The maximum batch hold penalty, bn, is the number of batches processed by the furnace in a production cycle.

Preferably, the solving of the rolling scheduling mathematical model by the M-GEP algorithm is as follows:

step S1: initializing individuals and populations in the M-GEP, taking the scheduling schemes of the outsource plan and the hot workshop as the individuals in the M-GEP, and taking the scheduling schemes of the multiple outsource plans and the multiple hot workshops as the populations in the M-GEP;

wherein each individual in the M-GEP algorithm represents a scheduling plan, wherein decision information about the scheduling plans of the outside cooperation plan and the hot shop is stored in a chromosome, wherein the chromosome comprises two types of genes: the gene coding of the batch determines the batch rule of the workpieces, and the structural gene coding determines the loading sequence of the batch;

step S2: decoding chromosomes of an individual, wherein the decoding comprises batch gene decoding and structural gene decoding;

wherein the batching genes are decoded to calculate the priority of the workpieces, an outsource task set is selected and the rest task sets are batched;

decoding the structural gene into adjusting batch loading sequence;

and step S3: obtaining the reciprocal of each individual in the target function as the fitness value of the individual;

and step S4: judging whether the current evolution algebra reaches the maximum evolution algebra, if so, outputting an individual with the highest fitness value as a production scheduling scheme of the external cooperation workpiece; if not, evolving the algebra +1, selecting the individual with the highest response value to enter the next generation of population, and selecting N-2 individuals from the father generation of population to enter the next generation of population by adopting a roulette algorithm, wherein N is the number of the father generation of population;

step S5: and (3) carrying out genetic operation on individual chromosomes in the population, wherein the genetic operation comprises gene variation, insertion and chromosome cross recombination, and repeating the steps S2-S5.

Preferably, the chromosome is composed of a head segment and a tail segment, wherein the head segment is composed of a function and a terminator together or only of a function, and the tail segment is composed of a terminator;

wherein the function symbol is four operation symbols of { +, -, - ×,% }, wherein% is protective division, when the divisor is 0, the return value is 1;

the terminator is the attribute of the workpiece;

the head fragment and the tail fragment have the following relationship: lt = lh +1,lt is the tail fragment length and lh is the head fragment length.

Preferably, the online scheduling process of the workpieces on the hot shop is as follows:

a1, when a heat treatment machine of a heat treatment workshop is switched into an idle state or a new workpiece reaches the idle heat treatment machine, recording the current time as tnow, and establishing a prediction time window [ tnow, tnow + LAW ] of the current workpiece task i;

step A2: forming a plurality of workpiece tasks which reach the predicted time window at the moment of tnow into a task set J _NT And selecting any workpiece task in a task set JNT to construct an LAB scheme, wherein the total mass of any workpiece task after combination is smaller than the rated load of the heat treatment machine, and each workpiece task correspondingly constructs an LAB squareCase;

step A3: judging whether the task time length of the next workpiece task I +1 exceeds the time length of the predicted time window, if not, updating the current time tnow by using the time of the next workpiece I +1 reaching the heat treatment machine, if so, selecting the optimal LAB scheme of a single workpiece in the workpiece task from the constructed LAB schemes, and recording the collection of the optimal LAB schemes of all the single workpieces as a set I ^* ；

Step A4: in set I ^* And selecting an optimal LAB scheme, recording the optimal LAB scheme as a final LAB scheme, using the final LAB scheme as a computer-on scheme of the heat treatment machine, judging whether the final LAB scheme is operated on the computer at the current moment, if so, immediately performing heat treatment processing on the workpiece of the final LAB scheme by the heat treatment machine, and if not, immediately processing the workpiece of the final LAB scheme after the heat treatment machine enters an idle state.

Preferably, the LAB protocol constructed in step A2 is specifically as follows:

wherein gamma is a target weight coefficient, D _j Is a task set J _NT The delivery date of the middle work piece j,

Is a task set J _NT Minimum lead time of all workpieces,

Is a task set J _NT Maximum lead time, w, in all workpieces _j Is a task set J _NT The weight of the middle workpiece j,

Is a task set J _NT Average weight of the workpiece,

Is a task set J _NT Middle workpieceMaximum difference in weight from the average weight.

Preferably, the specific steps of obtaining the optimal LAB scheme for a single workpiece in step A3 are as follows:

substituting the LAB scheme into an evaluation function, and acquiring a workpiece LAB scheme with the minimum evaluation function value in the LAB scheme as an optimal LAB scheme of a single workpiece;

wherein the merit function is as follows:

wherein gamma is the target weight coefficient, I _WTI Weighted pull-off index, min (I) for LAB version I _WTI ) Is the smallest weighted lag index among all LAB solutions; max (I) _WTI ) Is the largest weighted lag index in all LAB solutions; i is _EEI Energy consumption index, min (I) for LAB protocol I _EEI ) Is the smallest energy consumption index in all LAB schemes; max (I) _EEI ) Is the largest weighted pull finger in all LAB schemes.

Preferably, the formula for obtaining the final LAB solution is as follows:

wherein I is set I ^* Middle LAB protocol, J _i* Representation set I ^* Set of all workpieces in, J _i ′ _* Representing non-collections I ^* The collection of all the workpieces in the group,

is shown in the set J _i* The pull-off penalty factor for the jth workpiece,

is shown in the set J _i* The amount of the pull-out period of the jth workpiece,

is shown in the set J _i ′ _* The amount of pull-out for the jth workpiece in (j'),

for the heat treatment completion time of the workpiece j',

represents a set J _i ′ _* Heat treatment delivery date, n, of the jth workpiece _i* Represents a set J _i* Number of middle workpieces, n _i ′ _* Representing non-aggregate J _i ′ _* The number of workpieces;

wherein

Represents the average heat treatment time of the workpiece in j', W represents the payload of the heat treatment machine,

represents a set J _i ′ _* The weight of the middle work piece j',

represents a set J _i ′ _* The number of batches that can be made up,

represents a set J _i ′ _* Average processing time of the medium workpiece task;

and acquiring an LAB scheme with the minimum M (i) value as a final LAB scheme.

One of the above technical solutions has the following advantages or beneficial effects: the invention provides a proactive periodical rolling scheduling, which decomposes a dynamic scheduling problem into a series of deterministic subproblems according to a time axis in sequence, and formulates a workpiece external cooperation plan and a pre-scheduling scheme in a period, thereby balancing the production load of a short-term heat treatment workshop; therefore, a heat treatment production plan in a future period is obtained, a reliable outsourcing decision is made, and the proactive of the production plan is realized.

Drawings

Fig. 1 is a schematic structural diagram of an embodiment of the present invention.

FIG. 2 is a flowchart of batch gene decoding according to one embodiment of the present invention.

FIG. 3 is a flow chart of gene decoding according to an embodiment of the present invention.

FIG. 4 is a schematic diagram of gene mutation according to an embodiment of the present invention.

FIG. 5 is a schematic cross-regrouping diagram of one embodiment of the present invention.

FIG. 6 is a LAB scheme build flow diagram of one embodiment of the present invention.

Detailed Description

Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the accompanying drawings are illustrative only for the purpose of explaining the present invention, and are not to be construed as limiting the present invention.

In the description of the embodiments of the present invention, the terms "first", "second", and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or to implicitly indicate the number of technical features indicated. Thus, features defined as "first", "second", may explicitly or implicitly include one or more of the described features. In the description of the embodiments of the present invention, "a plurality" means two or more unless specifically limited otherwise.

Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of the present invention, "a plurality" means two or more unless otherwise specified. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.

As shown in fig. 1 to 6, a method for dynamic batch scheduling for on-time and energy-saving production of mold heat treatment comprises the following steps:

periodically acquiring information of a workpiece to be thermally treated;

according to the rolling period in the information of the workpiece to be heat-treated, taking part of the workpiece in the rolling period as an external cooperation workpiece;

and performing on-line dispatching on the rest workpieces on a hot workshop.

Aiming at the management and control problem of heat treatment production, the invention makes two researches: firstly, selecting an external cooperation workpiece, and secondly, carrying out real-time on-machine decision in a heat treatment workshop. The researches in the two aspects are not independent, the selection of the outsourced workpiece has the effect of balancing the production load of a workshop, and the decision of real-time computer operation of the workpiece is directly influenced. In view of the uncertainty and multiple optimization goals of the mold thermal processing production environment, a hierarchical scheduling mechanism is proposed herein. Firstly, acquiring part of workpieces needing outsourcing, interfering the workpiece processing amount of a rolling period, and then carrying out online scheduling on the workpieces based on the rest, wherein the workpieces in an outsourcing plan are not added with a task set of online scheduling, and the rest of the workpieces realize real-time control through online scheduling, so that batching and on-machine decisions of a heat treatment workshop are completed.

The invention provides a proactive periodical rolling scheduling, which decomposes a dynamic scheduling problem into a series of deterministic subproblems according to a time axis in sequence, and formulates a workpiece external cooperation plan and a pre-scheduling scheme in a period, thereby balancing the production load of a short-term heat treatment workshop; therefore, a heat treatment production plan in a future period is obtained, a reliable outsourcing decision is made, and the proactive of the production plan is realized.

Preferably, the step of obtaining the outsourced workpiece is as follows:

establishing a rolling scheduling mathematical model through the information of the workpiece to be heat-treated;

and solving the rolling scheduling mathematical model through an M-GEP algorithm to obtain a scheduling scheme of the outsourced workpieces.

The heat treatment production process of the die is shown in figure 1, and most workpieces dynamically reach a heat treatment workshop from an upstream process, are subjected to heat treatment and then are transferred out to a downstream processing process. The heat treatment workshop mainly comprises two parts, namely a task pool without capacity limitation and a heat treatment furnace. After dynamically reaching a heat treatment workshop, the workpieces firstly enter a task pool, are batched and put on the machine in batches when the time of putting on the machine arrives, and enter a downstream procedure after heat treatment; and (3) carrying out outside processing on a small part of workpieces after the upstream process is finished without entering a heat treatment workshop, returning to the downstream process after the outside processing is finished, and finishing the production of the die together.

Preferably, the specific steps of constructing the rolling scheduling mathematical model are as follows:

obtaining the production cost PI of the die workpiece j in the workpiece information to be heat-treated in the enterprise from the erp of the enterprise _j Exterior cost PO _j Calculating the total production cost in the rolling period and the extra cost for the outside cooperation;

when the external cooperation workpiece is selected, the extra cost, the pull-in penalty and the internal manufacturing cost of the external cooperation are required to be compromised to be the production cost, an optimization model (and the objective function) is established with the aim of minimizing the production cost, and an external cooperation plan and a pre-scheduling scheme are obtained through calculation;

the rolling scheduling decision selected by the outsourced workpieces on the upper layer can act on the on-machine decision of the rest workpieces on the lower layer, and the fundamental reason is that the upper layer and the lower layer both meet the delivery date and reduce the delay penalty as the optimization target. The upper-layer rolling scheduling mechanism considers the extra production cost brought by the outside cooperation plan while reducing the possibility of generating the drag period of the future workpiece; and the lower-layer online scheduling reduces the production energy consumption as much as possible while considering the reduction of the work-piece stall penalty. The two have a common target, so the upper layer rolling decision has a direct effect on the lower layer real-time decision.

The choice of the external cooperation workpiece aims at balancing the production load of the workshop. If the production load of the workshop is too large, the delivery delay of the order of the enterprise is often caused, and the enterprise bears corresponding delay punishment (paying default money). The effective outsourcing strategy is a method for reducing the production load of a workshop and also plays a main role in scheduling.

Setting the drag penalty of the workpiece according to the drag amount and the drag penalty coefficient of the workpiece;

with n _o The production load of the workshop is reduced, so that the risk of order pull-off of the enterprise is reduced. But also increases Extra _ Cost, which causes the loss of production profit of enterprises. Therefore, the development of the outsourcing scheme is to balance the profit and the risk of order pull-off sufficiently.

Predicting a stall penalty according to the total production cost, and establishing an objective function of a minimum stall penalty coefficient and an extra cost coefficient;

obtaining a scheduling scheme of the outside cooperation workpiece by analyzing the objective function;

in actual production, a contradiction relation exists between the extra cost and the pull-out penalty, and the addition of the external cooperation workpiece can reduce the pull-out penalty on one hand and increase the extra cost on the other hand. The need for upper layer scheduling finds the optimum between the two, for which an objective function is established that minimizes the stall penalty factor and the extra cost factor.

The total production cost is as follows:

the additional cost in the scroll cycle is as follows:

where Nb is the number of workpieces in the rolling period, n _o Number n of outer cooperation workpieces _Em For number of emergency outsourcing workpieces, PI _j For die work j in the enterpriseCost of internal production, em _ PO _j Is a processing time margin Tm _j Less than outside preparation time T _pre The outside cooperation cost of the die part, ord _ PO, is urgent _j The outside cooperation cost is met for the workpiece j under the condition of meeting the outside cooperation preparation time;

the hangover penalty is as follows:

wherein J _b For all the work sets contained in lot b,

The hold-off amount of the workpiece j in the batch b,

A drag penalty factor for the workpiece j in batch b;

the objective function is as follows:

Extra-Cost is the Extra Cost in the roll cycle, and Extra-Cost is the total production Cost, max-Tw _k The maximum batch hold penalty, bn, is the number of batches processed by the furnace in a production cycle.

Scheduling decision is carried out based on predicted data of the workshop production condition in the whole period, and low robustness is an important bottleneck faced by the decision mode. In order to improve the robustness of the decision, the optimistic arrival time and the pessimistic arrival time of the workpiece can be respectively calculated to obtain two objective functions, and two outsourcing schemes are formulated according to the two objective functions. Finally, two outsourced schemes are given to the management layer for decision.

Preferably, the solving of the rolling scheduling mathematical model by the M-GEP algorithm is as follows:

the traditional GEP algorithm flow is as follows:

step1: initializing individuals and populations;

step2: calculating an individual fitness value;

step3: and (5) iteration times are plus 1, and whether the maximum iteration times are reached is judged. If the optimal solution is reached, outputting the optimal solution, and finishing the algorithm; otherwise, continuing to execute Step4;

step4: selecting excellent individuals to form a next generation population according to the fitness value of the individuals;

step5: random genetic manipulation was performed on the selected superior individuals. The genetic manipulation comprises: chromosomal gene variation, insertion, chromosomal cross recombination. Returning to Step2.

And the M-GEP is a GEP algorithm which is used with the aim of minimizing the lag time cost coefficient and the additional cost coefficient.

Step S1: initializing individuals and populations in the M-GEP, using the scheduling schemes of the external cooperation plan and the hot workshop as the individuals in the M-GEP, and using the scheduling schemes of the multiple external cooperation plans and the multiple hot workshops as the populations in the M-GEP;

wherein each individual in the M-GEP algorithm represents a scheduling plan, wherein decision information about the scheduling plans of the outside cooperation plan and the hot shop is stored in a chromosome, wherein the chromosome comprises two types of genes: the gene coding of the batch determines the batch rule of the workpieces, and the structural gene coding determines the loading sequence of the batch;

preferably, the chromosome is composed of a head segment and a tail segment, wherein the head segment is composed of a function and a terminator together or only of a function, and the tail segment is composed of a terminator;

the function symbol is four operation symbols of { +, - ×,% }, wherein% is protective division, and when the divisor is 0, the return value is 1;

the terminator is the attribute of the workpiece;

for example: the terminator may be

Sequentially comprises the following steps: arrival time, processing time allowance, weight, delivery date,A lingering penalty coefficient, an internal production profit, and an external profit. Chromosomes require the use of multiple batching genes to batch workpieces from different workpiece families, respectively. After the batching is finished, initializing a batch loading scheme according to the batching finishing sequence, and then adjusting the batch loading sequence by the same structural gene, wherein the batching gene is shown as follows in one embodiment:

in the above-mentioned batch genes, the workpieces are from 3 workpiece groups in total, and thus the chromosome contains 3 batch genes. Wherein the length of the head segment is 4 and the length of the tail segment is 5. In the context of the present invention,

the head segment and the tail segment are separated for convenient understanding.

In addition, the structural genes in this example are shown below:

the head fragment and the tail fragment have the following relationship: lt = lh +1,lt is the tail fragment length and lh is the head fragment length.

Step S2: decoding chromosomes of an individual, wherein the decoding comprises batch gene decoding and structural gene decoding;

wherein the batching genes are decoded to calculate the priority of the workpieces, an outsource task set is selected and the rest task sets are batched;

decoding the structural gene into adjusting batch loading sequence;

as shown in fig. 2, the batch genes of the workpiece family are converted into K-expression trees one by one, and the K-expression trees are decoded into a priority calculation formula by a width-first traversal method, that is, in the order from top to bottom and from left to right. And substituting the attribute value of the workpiece into a priority calculation formula corresponding to the workpiece family to calculate the priority value of the workpiece. The workpieces in each workpiece group are arranged from large to small according to the priority, and the smaller the priority is, the lower the priority of the workpiece is, and sufficient machining time allowance is provided for carrying out external coordination.

As shown in FIG. 3, after all the workpieces are batched, the batching order is initialized according to the batching completion time. The lot completion time is defined as the time of arrival at the heat treatment shop of the latest arriving workpiece in the lot. Without loss of generality, the processing time margin of some batches is smaller although the batch completion time is later, so that the batch processing sequence can be adjusted through the structural genes, the production scheduling plan is optimized, and the outsourcing scheme is more robust. In addition, although the empty lot is allowed to exist in the structural gene, the empty lot can be actually regarded as a lot with a heat treatment time of 0, and therefore the calculated completion time of the heat treatment of the lot is the same as the actually calculated value.

And step S3: obtaining the reciprocal of each individual in the target function as the fitness value of the individual;

and step S4: judging whether the current evolution algebra reaches the maximum evolution algebra, if so, outputting an individual with the highest fitness value as a production scheduling scheme of the external cooperation workpiece; if not, evolving the algebra +1, selecting the individual with the highest response value to enter the next generation of population, and selecting N-2 individuals from the father generation of population to enter the next generation of population by adopting a roulette algorithm, wherein N is the number of the father generation of population;

step S5: and (3) carrying out genetic operation on individual chromosomes in the population, wherein the genetic operation comprises genetic variation, insertion and chromosome cross recombination, and repeating the steps S2-S5.

The genetic manipulation of the M-GEP specifically comprises gene variation, insertion and chromosome cross recombination;

and carrying out genetic operation on the group genes and the structural genes simultaneously by the M-GEP. The individual batch scheme, the outsource scheme and the batch computer-on scheme are adjusted through the parallel evolution of the two genes, so that the evolution capability of the M-GEP is greatly improved, and the convergence speed is accelerated.

1) Genetic variation:

as shown in fig. 4: in M-GEP, the group genes and the structural genes adopt variant genetic operators.

For the head fragment of the set of genes, the first element can only be mutated into a function, the other elements of the head can be mutated into a function or terminator, and the elements on the tail fragment can only be mutated into a terminator.

For the structural gene, the head segment can only be mutated into a functional character, and the tail segment does not do any operation.

2) Inserting strings:

in the invention, only the group genes adopt genetic operators. Since the head segment of the structural gene cannot have a terminator, the structural gene does not employ an insertion operator.

The insertion string has two operations: IS insertion string and RIS root insertion string.

IS insertion string: a gene fragment, which may be 2 or 3 or 4 in length, is selected and copied from the tail fragment of the bulk gene and inserted at any position after the first element of the head fragment. After the head segment is inserted into the gene segment, elements behind the insertion point are moved backwards, redundant segments of the head are deleted, and the length of the head is kept unchanged.

The RIS inserts the cluster: selecting and copying a gene segment from any position behind the first element of the head of the group gene, wherein the gene segment should contain a complete expression, inserting the gene segment in front of the first element of the head, deleting the redundant segment of the head, and keeping the length of the head unchanged. If no gene fragment containing the complete formula can be found, the procedure is not performed.

3) Chromosome cross recombination:

as shown in fig. 5: both the batch genes and the structural genes adopt chromosome single-point crossing genetic operators.

Chromosome cross recombination is the most destructive genetic operator. Its destructive power is reflected in the possibility of generating entirely new chromosome structures by simple crossover operations.

Preferably, the online scheduling process of the workpieces on the hot shop is as follows:

as shown in fig. 6: a1, when a heat treatment machine of a heat treatment workshop is switched into an idle state or a new workpiece reaches the idle heat treatment machine, recording the current time as tnow, and establishing a prediction time window [ tnow, tnow + LAW ] of the current workpiece task i (a plurality of workpieces need to perform heat treatment in one workpiece task);

step A2: forming a plurality of workpiece tasks which reach the predicted time window at the moment of tnow into a task set J _NT Selecting any workpiece task in a task set JNT to construct an LAB scheme, wherein the total mass of any workpiece task after combination is smaller than the rated load of a heat treatment machine, and each workpiece task correspondingly constructs an LAB scheme;

step A3: judging whether the task time length of the next workpiece task I +1 exceeds the time length of the predicted time window, if not, updating the current time tnow by using the time of the next workpiece I +1 reaching the heat treatment machine, if so, selecting the optimal LAB scheme of a single workpiece in the workpiece task from the constructed LAB schemes, and recording the collection of the optimal LAB schemes of all the single workpieces as a set I ^* ；

Step A4: in set I ^* And selecting an optimal LAB scheme, recording the optimal LAB scheme as a final LAB scheme, using the final LAB scheme as a computer-on scheme of the heat treatment machine, judging whether the final LAB scheme is operated on the computer at the current moment, if so, immediately performing heat treatment processing on the workpiece of the final LAB scheme by the heat treatment machine, and if not, immediately processing the workpiece of the final LAB scheme after the heat treatment machine enters an idle state.

Preferably, the LAB protocol constructed in step A2 is specifically as follows:

where γ is the target weight coefficient, D _j Is a task set J _NT The delivery date of the middle work piece j,

Is a task set J _NT Minimum lead time of all workpieces,

Is a task set J _NT Maximum lead time, w, in all workpieces _j Is a task set J _NT The weight of the middle workpiece j,

Is a task set J _NT Average weight of the workpiece,

Is a task set J _NT The maximum difference between the average weight and the median workpiece weight.

Preferably, the specific steps of obtaining the optimal LAB scheme for a single workpiece in step A3 are as follows:

substituting the LAB scheme into an evaluation function, and acquiring a workpiece LAB scheme with the minimum evaluation function value in the LAB scheme as an optimal LAB scheme of a single workpiece;

wherein the merit function is as follows:

wherein I _WTI Weighted pull-off index, min (I) for LAB version I _WTI ) Is the smallest weighted lag index among all LAB solutions; max (I) _WTI ) Is the largest weighted lag index in all LAB solutions; i is _EEI Energy consumption index, min (I) for LAB protocol I _EEI ) Is the smallest energy consumption index in all LAB schemes; max (I) _EEI ) Is the largest weighted pull finger in all LAB schemes.

Preferably, the formula for obtaining the final LAB solution is as follows:

wherein I is set I ^* Middle LAB protocol, J _i* Representation set I ^* All work set in, J' _i* Representing non-collections I ^* The collection of all the workpieces in the group,

is shown in the set J _i* The pull-off penalty factor for the jth workpiece,

is shown in the set J _i* The amount of the pull-out period of the jth workpiece,

is represented by the aggregate J' _i* The amount of pull-out for the jth workpiece in (j'),

for the time to complete the heat treatment of the workpiece j',

represents a set J' _i* Heat treatment delivery date, n, of the jth workpiece _i* Represents a set J _i* Number of workpieces, n' _i* Denotes non-collective J' _i* The number of workpieces;

wherein

Represents the average heat treatment time of the workpiece in j', W represents the payload of the heat treatment machine,

represents a set J' _i* The weight of the middle work piece j',

represents a set J' _i* The number of batches that can be made up,

represents a set J' _i* Average processing time of the medium workpiece task;

and acquiring an LAB scheme with the minimum M (i) value as a final LAB scheme.

In the description of the present specification, reference to the description of "one embodiment," "some embodiments," "an illustrative embodiment," "an example," "a specific example," or "some examples" or the like means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and not to be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made in the above embodiments by those of ordinary skill in the art within the scope of the present invention.

Claims (9)

1. A dynamic batch scheduling method for on-time and energy-saving production of die heat treatment is characterized by comprising the following steps:

periodically acquiring information of a workpiece to be thermally treated;

according to the rolling period in the information of the workpiece to be heat-treated, taking part of the workpiece in the rolling period as an external cooperation workpiece;

and performing on-line dispatching on the rest workpieces on a hot workshop.

2. The method for dynamic batch scheduling for on-time and energy-saving production of die heat treatment according to claim 1, wherein the step of obtaining the outsourced workpiece is as follows:

establishing a rolling scheduling mathematical model through the information of the workpiece to be heat-treated;

and solving the rolling scheduling mathematical model through an M-GEP algorithm to obtain a scheduling scheme of the outsourced workpieces.

3. The method for dynamically scheduling the on-time and energy-saving production of the heat treatment of the mold according to claim 2, wherein the specific steps for constructing the mathematical model of the rolling scheduling are as follows:

obtaining the production cost PI of the die workpiece j in the workpiece information to be heat-treated in the enterprise from the erp of the enterprise _j Exterior cost PO _j Calculating the total production cost in the rolling period and the extra cost for the outside cooperation;

setting the drag penalty of the workpiece according to the drag amount and the drag penalty coefficient of the workpiece;

predicting a stall penalty according to the total production cost, and establishing an objective function of a minimum stall penalty coefficient and an extra cost coefficient;

obtaining a scheduling scheme of the outside cooperation workpiece by analyzing the target function;

the total production cost is as follows:

the additional cost in the scroll cycle is as follows:

where Nb is the number of workpieces in the rolling period, n _o Number n of outer cooperation workpieces _Em For number of emergency outsourcing workpieces, PI _j Cost of production of the die work j inside the enterprise, em _ PO _j Tm as a margin of processing time _j Less than outside preparation time T _pre The outside cooperation cost of the die part, ord _ PO, is urgent _j The outside cooperation cost is met for the workpiece j under the condition of meeting the outside cooperation preparation time;

the hangover penalty is as follows:

wherein J _b For all the work sets contained in lot b,

The hold-off amount of the workpiece j in the batch b,

A drag penalty factor for the workpiece j in batch b;

the objective function is as follows:

Extra-Cost is the Extra Cost in the roll cycle, and Extra-Cost is the total production Cost, max-Tw _k The maximum batch hold penalty, bn, is the number of batches processed by the furnace in a production cycle.

4. The method for dynamically scheduling the batch for the on-time and energy-saving production of the die heat treatment according to claim 2, wherein the mathematical model of the rolling scheduling is solved by an M-GEP algorithm as follows:

step S1: initializing individuals and populations in the M-GEP, using the scheduling schemes of the external cooperation plan and the hot workshop as the individuals in the M-GEP, and using the scheduling schemes of the multiple external cooperation plans and the multiple hot workshops as the populations in the M-GEP;

wherein each individual in the M-GEP algorithm represents a scheduling plan, wherein decision information about the scheduling plans of the outside cooperation plan and the hot shop is stored in a chromosome, wherein the chromosome comprises two types of genes: the gene coding of the batch determines the batch rule of the workpieces, and the structural gene coding determines the loading sequence of the batch;

step S2: decoding chromosomes of an individual, wherein the decoding comprises batch gene decoding and structural gene decoding;

wherein the batching genes are decoded to calculate the priority of the workpieces, an outsource task set is selected and the rest task sets are batched;

decoding the structural gene into adjusting batch loading sequence;

and step S3: obtaining the reciprocal of each individual in the target function as the fitness value of the individual;

and step S4: judging whether the current evolution algebra reaches the maximum evolution algebra, if so, outputting an individual with the highest fitness value as a production scheduling scheme of the external cooperation workpiece; if not, evolving the algebra +1, selecting the individual with the highest response value to enter the next generation of population, and selecting N-2 individuals from the father generation of population to enter the next generation of population by adopting a roulette algorithm, wherein N is the number of the father generation of population;

step S5: and (3) carrying out genetic operation on individual chromosomes in the population, wherein the genetic operation comprises genetic variation, insertion and chromosome cross recombination, and repeating the steps S2-S5.

5. The dynamic batch scheduling method for the on-time and energy-saving production of the mold-oriented heat treatment, according to claim 4, wherein the chromosome is composed of a head segment and a tail segment, wherein the head segment is composed of a function symbol and a terminator together or only of a function symbol, and the tail segment is composed of a terminator;

the function symbol is four operation symbols of { +, - ×,% }, wherein% is protective division, and when the divisor is 0, the return value is 1;

the terminator is the attribute of the workpiece;

the head fragment and the tail fragment have the following relationship: lt = lh +1,lt is the tail fragment length and lh is the head fragment length.

6. The method for dynamically scheduling the on-time and energy-saving production of the heat treatment of the die as claimed in claim 1, wherein the on-line scheduling process of the workpieces on the hot workshop is as follows:

a1, when a heat treatment machine of a heat treatment workshop is switched into an idle state or a new workpiece reaches the idle heat treatment machine, recording the current time as tnow, and establishing a prediction time window [ tnow, tnow + LAW ] of the current workpiece task i;

step A2: forming a plurality of workpiece tasks which reach the predicted time window at the moment of tnow into a task set J _NT Selecting any workpiece task in a task set JNT to construct an LAB scheme, wherein the total mass of any workpiece task after combination is smaller than the rated load of the heat treatment machine, and each workpiece task correspondingly constructs an LAB scheme;

step A3: judging whether the task time length of the next workpiece task I +1 exceeds the time length of the predicted time window, if not, updating the current time tnow by using the time of the next workpiece I +1 reaching the heat treatment machine, if so, selecting the optimal LAB scheme of a single workpiece in the workpiece task from the constructed LAB schemes, and recording the collection of the optimal LAB schemes of all the single workpieces as a set I ^* ；

Step A4: in set I ^* And selecting an optimal LAB scheme, recording the optimal LAB scheme as a final LAB scheme, using the final LAB scheme as a computer-on scheme of the heat treatment machine, judging whether the final LAB scheme is operated on the computer at the current moment, if so, immediately performing heat treatment processing on the workpiece of the final LAB scheme by the heat treatment machine, and if not, immediately processing the workpiece of the final LAB scheme after the heat treatment machine enters an idle state.

7. The method for dynamic batch scheduling for on-time and energy-saving production of mold heat treatment according to claim 6, wherein the LAB scheme constructed in step A2 is specifically as follows:

where γ is the target weight coefficient, D _j As a set of tasks J _NT The delivery date of the middle work piece j,

Is a task set J _NT Minimum lead time of all workpieces,

Is a task set J _NT Maximum lead time, w, in all workpieces _j Is a task set J _NT The weight of the middle workpiece j,

Is a task set J _NT Average weight of workpiece,

Is a task set J _NT The maximum difference between the average weight and the median workpiece weight.

8. The method for dynamically scheduling the on-time and energy-saving production of the die heat treatment according to claim 6, wherein the specific steps for obtaining the optimal LAB scheme of a single workpiece in the step A3 are as follows:

substituting the LAB scheme into an evaluation function, and acquiring a workpiece LAB scheme with the minimum evaluation function value in the LAB scheme as an optimal LAB scheme of a single workpiece;

wherein the merit function is as follows:

wherein I _WTI Weighted pull-off index, min (I) for LAB version I _WTI ) Is the smallest weighted lag index among all LAB solutions; max (I) _WTI ) Is the largest weighted lag index in all LAB solutions; i is _EEI Energy consumption index, min (I) for LAB protocol I _EEI ) Is the minimum energy consumption index in all LAB schemes; max (I) _EEI ) Is the largest weighted pull finger among all LAB schemes.

9. The method for dynamic batch scheduling for on-time and energy-saving production of mold heat treatment according to claim 6, wherein the formula for obtaining the final LAB scheme is as follows:

wherein I is set I ^* Middle LAB protocol, J _i* Representation set I ^* Set of all workpieces in, J _i ′ _* Represents a non-aggregate I ^* The collection of all the workpieces in the group,

is shown in the set J _i* The pull-off penalty factor for the jth workpiece,

is shown in the set J _i* The amount of the pull-out period of the jth workpiece,

is shown in the set J _i ′ _* The amount of pull-out for the jth workpiece in (j'),

for the time to complete the heat treatment of the workpiece j',

represents a set J _i ′ _* Heat treatment delivery date, n, of the jth workpiece _i* Represents a set J _i* Number of middle workpieces, n _i ′ _* Representing non-aggregate J _i ′ _* The number of middle workpieces;

wherein

Represents the average heat treatment time of the workpiece in j', W represents the payload of the heat treatment machine,

represents a set J _i ′ _* The weight of the middle work piece j',

represents a set J _i ′ _* The number of batches that can be made up,

represents a set J _i ′ _* Average processing time of the medium workpiece task;

and acquiring an LAB scheme with the minimum M (i) value as a final LAB scheme.

Images