CN101592945A - A kind of process and device of handling iron and steel enterprise's color coating unit production disturbance - Google Patents

Abstract

A kind of process of handling iron and steel enterprise's color coating unit production disturbance, it is characterized in that setting up earlier the Matching Model between a time gap and the candidate coiled sheet, viewpoint with global optimization is repaired all time gaps simultaneously, will be to carrying out the feasibility inspection by the resulting new production operation scheme of Matching Model, and repair the infeasible situation that wherein may occur; At last, use intelligent optimization algorithm that feasible new production operation scheme is improved.The invention allows for a kind of device of implementing above-mentioned process that is exclusively used in, it comprises: production disturbance watch-dog 1, TCP/IP network 2, business data server 3, data acquisition unit 4, data processor 5, storer 6, optimized processor 7, based on touch-screen display 8 and the production operation scheme uploader 9 of LCD.Adopt the present invention can respond and effectively handle the production disturbance of iron and steel enterprise's color coating unit fast, thereby guaranteed the continuity of producing better, improve the production operation efficiency of unit, also reduced production cost, possess bigger economic worth.

Description

A kind of process and device of handling iron and steel enterprise's color coating unit production disturbance

Technical field

The invention belongs to the technical field of Ferrous Metallurgy, particularly a kind ofly can effectively handle process and the device that iron and steel enterprise is coated with unit production disturbance.

Background technology

The main production process of Coil Coating Products volume can be represented with Fig. 1 in iron and steel enterprise's cold rolling mill.After arriving cold rolling mill, hot-rolled sheet coil is at first disposed the oxide on surface by pickling process, then be rolled into thin cold rolling coil at cold rolling process.So that the spraying operation in the color coating operation, these cold rolling coils are by the continuous hot galvanizing operation or undertaken zinc-plated by continuous annealing and electrogalvanizing operation for the quality that guarantees the cold rolling coil surface.Last these galvanization rolls are processed to versicolor prepainted coil in the color coating operation according to the requirement of contract under them.Because the color coating unit is worked continuously, galvanization roll at first is launched into the band steel on uncoiler, and these band steel are coupled together one by one by bonding machine then.Through after the surface preparation, their are coated by the rendering machine earlier and just finish pigment, and then are coated with machine by essence and coat essence and finish pigment.The band steel that is coated with paint then enters postprocessing working procedures and carries out operations such as the oven dry of pigment and curing, is rolled into coiled sheet again at coiler at last.

Fig. 2 is the simplification mathematical model of color coating operation.As shown in the drawing, each coiled sheet coating colour and smart coating colour at the beginning of upper and lower surface all has one usually, be called just coat, under the rendering, essence is coated and essence is coated with down.In case the first coating colour of adjacent two coiled sheet i and j or smart coating colour are not simultaneously in the plan unit, between them, will insert a transition volume so aborning, to guarantee under the situation that production can be carried out continuously, time enough being arranged from being coated with the coating that roller washes the coating of coiled sheet color that i is required and changes coiled sheet color that j is required.The change color of two adjacent coiled sheets is called a color and switches in production schedule unit.

For production operation problem complicated in the steel and iron industry, under dynamic environment, be a very difficult and challenging job how for each unit finds an effective production operation scheme.In actual production, be hypothesis static and that determine normally by the resulting initial production operation scheme of manual method or production operation system based on following situation, this production operation scheme can be carried out in actual production under normal conditions smoothly.But because actual production is dynamic and uncertain in itself, some production disturbances (being also referred to as accident) in the production may be in the initial production operation scheme generation time slit (promptly should not having workpiece to be used for processing in the time period), thereby it can not be carried out smoothly.

The typical production disturbance that often occurs in the production of iron and steel enterprise's color coating unit comprises that mainly the coiled sheet of (1) upstream unit processing can not arrive the color coating operation on time, and the quality of the coiled sheet of (2) arrival color coating operation does not reach the standard of contract requirement (CR).Fig. 3 has provided an initial production operation scheme that includes only color coating operation and continuous hot galvanizing operation and has simplified mathematical model.In the figure, the plan deadline of each the coiled sheet j in the continuous hot galvanizing operation is t _j, the time that they arrive the color coating operation is t _j+ Δ t _j, the typical production disturbance in the actual production of color coating unit is broadly divided into following several situation:

(1) coiled sheet does not arrive on time

In the actual production of color coating unit, some production disturbances may cause the coiled sheet of upstream process processing not arrive the color coating operation by the time of original plan.Usually, these production disturbances can be divided into two classes:

1. the machine of upstream unit breaks down;

The color coating operation has a lot of upstreams unit to be its supply coiled sheet.If certain machine in these upstream units breaks down, so original plan could begin processing after will just must waiting until this machine reparation by the coiled sheet of this machining, thus the generation time slit.The example of this production disturbance can be reduced to the model of Fig. 4, among this figure coiled sheet 4,5 and 6 in the original plan in the continuous hot galvanizing operation by machine CG-2 at time period [t ₁, t ₂] in process.But because this machine is at time t ₁In time, break down, and will wait until time t ₂Shi Caineng repairs, and coiled sheet 4,5 and 6 just can not arrive the color coating operation by the time in the original plan so.

2. the speed of production of upstream unit reduce or upstream unit and color coating operation between the movement capacity deficiency.

Except mechanical disorder, other some improper situations also can cause coiled sheet can not arrive the color coating operation on time in the factory, the speed of production of the unit that for example is interfered by control system and causes reduces, or connects the movement capacity deficiency of the loop wheel machine in the coiled sheet storehouse between the adjacent operation.The example that the coiled sheet that is caused by this production disturbance does not arrive on time can be represented with Fig. 5.The deadline of supposing coiled sheet 9 is delayed to t owing to the speed of production of Continuous Hot Dip Galvanizing reduces ₉+ δ, so correspondingly the time of its arrival color coating operation also can be delayed to t ₉+ Δ t ₉+ δ.

(2) coiled sheet quality dissatisfaction

In user's contract of being signed, the quality of each coiled sheet has been made concrete requirement.But because the instability of some equipment in the actual production, the situation of quality dissatisfaction just may appear in some coiled sheets, even and these coiled sheets have arrived the color coating operation on time, can not carry out color coating production by original program.Fig. 6 has provided the mathematical model that causes time of occurrence slit in the initial production operation scheme owing to the coiled sheet quality dissatisfaction.In the figure, because coiled sheet 2 and 8 quality do not reach the desired standard of contract, so they can not be arranged in the color coating operation by the initial production operation scheme and process.

Optimization process method for the production disturbance in the actual production has had a lot of documents to carry out relevant research.People such as Vieira (Vieira GE, Herrmann JW, Lin E.Rescheduling manufacturing systems:A framework ofstrategies, policies, and methods.Journal of Scheduling, 2003,6:39-62) employed strategy in these methods is divided three classes: periodically tactful, the event-driven strategy, mixed strategy.Periodically strategy is periodically the production operation scheme to be adjusted; The event-driven strategy is just just adjusted the production operation scheme when specific accident takes place; Mixed strategy is integrated periodically strategy and event-driven strategy, it can periodically be adjusted the production operation scheme, also can adjust the production operation scheme when particular emergency event takes place.People such as Ouelhadj (Ouelhadj D, Cowling P, Petrovic S.Utility and Stability Measures for Agent-Based Dynamic Scheduling of SteelContinuous Casting, IEEE International Conference on Robotics and Automation ICRA 2003, Taipei, Taiwan) propose several heuristic strategies with the production perturbed problem that solves continuous casting unit in the iron and steel enterprise, and proposed " effectiveness " and " stability " two kinds of evaluation methods for these heuristic strategies.Sabuncuoglu and Bayiz (Sabuncuoglu I, Bayiz M.Analysis of reactive scheduling problems in a job shop environment.European Journal ofOperational Research, 2000,126 (3): 567-586) summed up the conclusion of the calm process of relevant production disturbance, and pointed out that most suggestion is that the performance of these methods depends on concrete problem.

As mentioned above, though proposed certain methods for the production perturbed problem existing literature in the actual production, but these methods are seldom at the relevant unit of iron and steel enterprise, and, they all belong to the heuristic of greedy (Greedy) in itself, promptly at every turn from candidate's slab or coiled sheet set, choose best being filled in the time gap that causes by accident, after a time gap was repaired, the slab or the coiled sheet that are filled in this time gap just can not be considered in the later process of algorithm again.Then other time gap also uses the same method and repairs.Till this process lasts till that always all time gaps all are repaired.The manual method of the reply production disturbance that present most iron and steel enterprise adopted also is to belong to this type.When using this greedy algorithm, obviously after first time gap is repaired, the time gap of back will become more and more difficult and repair, because along with the quantity of the selectable candidate's slab of increase of the quantity of the time gap of being repaired or coiled sheet can be fewer and feweri.Simultaneously, manual type is difficult to accomplish and makes overall plans, and can't respond and handle these disturbances fast with assurance grown place continuity, thereby be difficult to improve usage ratio of equipment.

Therefore, how to seek the successional production disturbance of helping conflicting of actual production situation by coordinating initial production operation scheme and current factory unit the to resume production method of calming and have very important practical significance for the production operation system of each unit of iron and steel enterprise.

Summary of the invention

The purpose of this invention is to provide and a kind ofly can effectively handle the process that iron and steel enterprise is coated with unit production disturbance, it is characterized in that adopting Matching Model to repair all time gaps that caused by the production disturbance in the initial production operation scheme simultaneously with the viewpoint of global optimization, realization is to the quick response of production disturbance, thereby guarantee the stable operation of production, improve the quality of products, improve the unit equipment utilization factor, and then improve the unit production capacity, reduce cost.

The technical solution that the present invention adopts is: at first this problem is established as the Matching Model between a time gap and the candidate's coiled sheet, the viewpoint of the enough global optimizations of this model energy is repaired all time gaps simultaneously, and the coupling expense between time gap and the candidate's coiled sheet mainly comprises the punishment on specification difference to the problem coiled sheet in the time gap and candidate's coiled sheet; Then will be to carrying out the feasibility inspection by the resulting new production operation scheme of Matching Model, and repair the infeasible situation that wherein may occur; At last, use tabu search (Tabu search) algorithm that feasible new production operation scheme is improved.

In the color coating of reality is produced, in case a production operation scheme is issued execution, with regard to unfolded corresponding preliminary work.At first, the coiled sheet information that is discharged into production schedules can send to the upstream unit in the mode of application of materials, and then the staff planners of these upstream units will take into account these application of materials when the production operation scheme of this unit of establishment.Next, the coiled sheet that is transported to the preceding storehouse of color coating from the upstream unit can be put in the coiled sheet storehouse according to they production orders color coating production operation scheme.Therefore, when the production disturbance occurs, it is obviously more reasonable than producing a brand-new production operation scheme again that the initial production operation scheme is carried out local adjustment, all that has been achieved is spoiled because a brand-new production operation scheme may make the preliminary work of being done early stage, causes huge difficulty also can for simultaneously the work in upstream unit and coiled sheet storehouse.Because method proposed by the invention just partly adjusts initial production operation scheme, thereby can minimize the production disturbance to the influence of initial production operation scheme and keep the quality of initial production operation scheme simultaneously.

The process of processing iron and steel enterprise color coating unit production disturbance proposed by the invention comprises following six steps:

The first step: the system when the production disturbance is occurred carries out approximate description

When the production disturbance appears in the color coating unit, the set of supposing the coiled sheet that goes wrong in the initial production operation scheme is M={1,2 ..., m}, this moment, the set of candidate's coiled sheet was N={1,2 ..., n}, notice that transition volumes all in the preceding storehouse of color coating is also contained among the N, and other coiled sheet is called the contract coiled sheet is arranged.Based on Matching Model, the color coating unit production disturbance optimization process problem that the present invention considered can approximate description be:

Minimize

(1)

s.t.

$\underset{j\∈M}{\mathrm{\Σ}}{x}_{\mathrm{ij}}\≤1,$ $\∀i\∈N$ (2)

$\underset{i\∈N}{\mathrm{\Σ}}{x}_{\mathrm{ij}}=1,$ $\∀j\∈M$ (3)

x _ij∈{0，1}， $\∀i\∈N,$ $\∀j\∈M$ (4)

In objective function (1), c _IjBe the coupling expense between candidate's coiled sheet i and the problem coiled sheet j, mainly comprise the weighted sum of the penalty value of the difference on width, thickness, weight, processing time, coating color, delivery date and time of arrival between them, as shown in Figure 7; x _IjIf matching relationship between expression candidate coiled sheet i and the problem coiled sheet j is promptly with candidate's coiled sheet i substitution problem coiled sheet j, then x _Ij=1, otherwise x _Ij=0.Constraint (2) and (3) is the technology and the operation constraint condition of system self, and constraint (2) guarantees that each candidate's coiled sheet is merely able to mate a problem coiled sheet at most.Constraint (3) guarantees that each problem coiled sheet must mate candidate's coiled sheet.Constraint (4) guarantees decision variable x _IjCan only get 0 or 1.Because the quantity of transition volume is always many in the preceding storehouse, can suppose n 〉=m, so this model is a nonequilibrium Matching Model.

Second step: with the model conversation of being set up in the above-mentioned first step is balance model

Thereby obtain new problem coiled sheet set M '=M Y{m+1 by in problem coiled sheet set M, introducing n-m virtual problem coiled sheet, ..., n}, and establish these virtual problem coiled sheets and the coupling expense between candidate's coiled sheet is L (a very big number), then Shang Mian non-equilibrium matching problem model can become an equilibrium problem model (promptly | M ' |=| N|):

Minimize

(5)

s.t.

$\underset{j\∈M^{\′}}{\mathrm{\Σ}}{x}_{\mathrm{ij}}=1,$ $\∀i\∈N$ (6)

$\underset{i\∈N}{\mathrm{\Σ}}{x}_{\mathrm{ij}}=1,$ $\∀j\∈M^{\′}$ (7)

x _ij∈{0，1}， $\∀i\∈N,$ $\∀j\∈M^{\′}$ (8)

The 3rd step: use the shortest increase routing algorithm to find the solution the balance Matching Model;

Use the shortest increase routing algorithm (Shortest augmenting path algorithm) to find the solution this balance matching problem, and the initial production operation scheme is carried out part adjustment according to resulting optimum solution.Generally, be feasible by the resulting new production operation scheme of this model.

The 4th step: adjusted production operation scheme is carried out feasibility inspection and reparation;

The 5th step: use intelligent optimization algorithm that new production operation scheme is improved, color switching times, specification in producing with reduction between the coiled sheet are switched cost, thereby are further improved the quality of products;

The 6th step: carry out last resulting new production operation scheme by the production automatic control system of color coating unit.

In the process in above-mentioned the 4th step, also be appreciated that in resulting adjusted production operation scheme to have following three kinds of infeasible situations by Matching Model:

(1) time gap

First kind of infeasible situation is life period slit still.As shown in Figure 8, problem coiled sheet S _iProcessing time be 50 minutes, and replace its candidate's coiled sheet S ' _iProcessing time be 30 minutes.Because coiled sheet

The process time that may begin the earliest be s _j+ 50, therefore at coiled sheet S ' _iWith Between still have a time gap of 20 minutes.In order to repair this infeasible situation, each time gap is all adopted following three kinds of methods, and therefrom select best a kind of of effect for use.

1. use another and coiled sheet S ' _iThe coupling expense minimum and satisfy processing time p _k〉=p _jCoiled sheet S ' _kReplace coiled sheet S ' _i

2. replace coiled sheet S ' with the combination of two or three best other candidate's coiled sheets that can fill this time gap _i

3. replace coiled sheet S ' with the combination (other candidate's coiled sheet and a transition volume) of two the best coiled sheets that can fill this time gap _i

(2) width knock-on

Because require in the production of color coating unit the width of coiled sheet must be leniently carry out transition to narrow, therefore second kind of infeasible situation is that bouncing appears in width, i.e. the width transition of adjacent coiled sheet is wide from being too narrow to.This situation only appears in the process of the several problem coiled sheets that replace continuous adjacent.Repair this infeasible situation, also will consider two kinds of situations:

If 1. this situation includes only two adjacent coiled sheets, then the method for Xiu Fuing can be replaced in existing two coiled sheets one for the position that exchanges these two coiled sheets or with other best candidate's coiled sheet, as shown in Figure 9;

If 2. this situation comprises the coiled sheet of several successive, then the method for Xiu Fuing can be for rearranging these coiled sheets or replacing some coiled sheet in these coiled sheets with other best candidate's coiled sheet, as shown in figure 10.

(3) width or thickness jump and exceed the scope of permission

Width that the third infeasible situation is adjacent coiled sheet or thickness jump and have exceeded the scope that allows.In order to repair this infeasible situation, adopt following three kinds of methods, and therefrom select best a kind of of effect for use:

1. replace the coiled sheet that this infeasible situation occurs with another best candidate's coiled sheet;

2. replace the coiled sheet that this infeasible situation occurs with the combination of two best in addition candidate's coiled sheets;

3. occurring inserting an only transition volume in the middle of the coiled sheet of this infeasible situation.

Owing to can roll up the quantity of transition volume in replacement problem coiled sheet and the actual production many with transition at least, therefore by obtaining feasible production operation scheme after the stage of feasibility inspection and reparation.

Said intelligent optimization algorithm can be a tabu search algorithm in the 5th step of the present invention, and the step of this algorithm is as follows:

Step1., i=1 is set, initialization maximum iteration time t _Max, the maximum running time T that allows _Max, deep search iterations t _Int, breadth first search's iterations t _Div, a=0.9 empties the taboo table.Current production schedules is preferably separated and current separating as history;

If Step were 2. i＞t _MaxOr running time of algorithm＞T _Max, the resulting best production schedules of output algorithm, and withdraw from; Otherwise enter Step 3;

If Step 3. i are divided exactly (t _Int+ t _Div) remainder＞t _Int, forward Step 4 to; Otherwise use deep search that current separating improved, promptly use all neighborhood type i nner-relocation (certain locational coiled sheet in the plan unit being inserted in this plan unit other position), Swap (exchanging two locational coiled sheets in the plan unit), Outer-relocation (coiled sheet in the plan unit is inserted in other the plan unit) and Exchange (respectively selecting a coiled sheet respectively in two different plan unit exchanges), carry out neighborhood search one time, therefrom determine best feasible moving, judge to move whether can be applied to current separating: (1) if this moves in the taboo table and uses to move and can not improve history and preferably separate, and then gets good feasible this deterministic process that moves and repeat from all feasible neighborhoods move next time; (2) use if this moves in the taboo table to move and to improve history and preferably separate, then current separating used this and moved; (3), then current separating used this and moved if should move not in the taboo table.Applied mobile opposite moving joined in the taboo table, upgrade the adopted number of times of each neighborhood type, upgrade history and preferably separate, and put i=i+1, forward Step2 then to;

Step 4. uses the breadth first search, promptly only being used the maximum a kind of neighborhood type of number of times in the stage with deep search carries out neighborhood search one time, best feasible the moving of therefrom finding out is applied directly to current separating, and with the adopted number of times zero clearing of each neighborhood type, applied mobile opposite moving joined in the taboo table, upgrade history and preferably separate, and put i=i+1.If i is (t _Int+ t _Div) integer multiple, mean that promptly breadth first search's stage carried out t continuously _DivIn generation, just put t _Div=t _Div* a reducing breadth first search's iterations in next circulation, and forwards Step 2 to; Otherwise forward Step 3 to.

Said intelligent optimization algorithm can be simulated annealing in the 5th step of the present invention, and this algorithm adopts multiple neighbour structure, and its flow process is as follows:

Step 1. preferably separates current production schedules S as history; Initial temperature T is set ₀=500, final temperature T _f=20, cooling coefficient a=0.97 is provided with iterations k=1, and makes T _k=T ₀Definition use Inner-relocation move (position that certain locational coiled sheet in the plan unit is inserted in this plan unit other) the neighborhood type be numbered 1, use Swap move (exchange one and plans two locational coiled sheets in the unit) the neighborhood type be numbered 2, use Outer-relocation move (coiled sheet in the plan unit is inserted in other the plan unit) the neighborhood type be numbered 3, use Exchange move (respectively selecting a coiled sheet respectively in two different plan unit exchanges) the neighborhood type be numbered 4;

If Step is 2. T _k＜T _f, the resulting best production schedules of output algorithm, and withdraw from; Otherwise, change Step3;

Step 3. 1,2,3, produce a random number r among the 4}, it is the scale of r kind neighborhood that N (r) is set, round-robin number of times i=0 in being provided with, and the current S of separating is set ₀=S;

Step 4. is for S ₀, in its neighborhood r, produce a neighborhood at random and separate S ' (promptly to S ₀Use moving at random of a r neighborhood type), and i=i+1 is set; The variation delta f=f of calculating target function (S ')-f (S ₀), if Δ f＜0 makes S ₀=S ' changes Step 4; Otherwise in [0,1], produce one and obey equally distributed random number ε, if exp (Δ f/T _k)＞ε then makes S ₀=S ', and change Step 5;

If Step were 5. i＞N (r), would change Step 6; Otherwise, change Step 4;

If Step is 6. f (S ₀)＜f (S) then upgrades S=S ₀Make k=k+1, T _k=T _k* a changes Step 2.

Said intelligent optimization algorithm can also be to become the neighborhood search algorithm in the 5th step of the present invention, and its process flow diagram is as follows:

Step 1. preferably separates current production schedules S as history, iterations i=0 is set, maximum iteration time I _Max=100; Definition use Inner-relocation move (position that certain locational coiled sheet in the plan unit is inserted in this plan unit other) the neighborhood type be numbered 1, use Swap move (exchange one and plans two locational coiled sheets in the unit) the neighborhood type be numbered 2, use Outer-relocation move (coiled sheet in the plan unit is inserted in other the plan unit) the neighborhood type be numbered 3, use Exchange move (respectively selecting a coiled sheet respectively in two different plan unit exchanges) the neighborhood type be numbered 4;

If i＞I Step2. _Max, the resulting best production schedules of output algorithm, and withdraw from; Otherwise, i=i+1 is set, and current neighborhood type k=1 is set, and change Step3;

If Step is 3. k＞and 4, change Step 2; Otherwise, change Step 4;

Step 4. produces a neighborhood at random and separates S ' (promptly S being used moving at random of a k neighborhood type) for S in its k neighborhood type; And the best neighborhood of searching for S ' in being numbered the neighborhood of k separates, and is designated as S "; If f (S ")＜f (S) puts S=S ", and k=1 is set; Otherwise k=k-1 is set.Change Step 3.

The invention allows for a kind of device that is used to implement the process of above-mentioned processing iron and steel enterprise color coating unit production disturbance, comprise following assembly:

(1) production disturbance watch-dog 1 is used for monitoring in real time the production disturbance of color coating unit, gathers the real-time information of production disturbance, and utilizes TCP/IP network 2 that it is uploaded to the data server 3 of enterprise;

(2) TCP/IP network 2 is used for carrying out the data communication between data server 3, data acquisition unit 4 and the production operation scheme uploader 9 of color coating unit, enterprise;

(3) business data server 3 be used for storing the coiled sheet information in color coating unit initial production operation scheme, the preceding storehouse of current time color coating, the real-time information of the production disturbance that collected by production disturbance watch-dog, and the new production scheduling scheme after handling through calm method;

(4) data acquisition unit 4, are used for detecting and downloading the real-time information of production disturbance relevant with the color coating unit in the business data server 3, and are sent to data processor 5; Simultaneously, download the coiled sheet information that is coated with unit initial production operation scheme and the preceding storehouse of current time color coating, and it is saved in storer 6;

(5) data processor 5, be used for to the real-time information of the data collector 4 production disturbance relevant that downloaded with the color coating unit classify, format conversion, and it is saved in the storer 6;

(6) storer 6, are used for preserving color coating unit initial production operation scheme, and the real-time information of the color coating unit production disturbance after 5 processing of process data processor;

(7) optimized processor 7, wherein embedded the calm process that is coated with unit production disturbance based on the iron and steel enterprise of Matching Model;

(8) based on the touch-screen display 8 of LCD, be used for showing the software system interface that embeds in this device, and with form and figure dual mode display optimization processor 7 adjusted new color coating unit production operation schemes;

(9) production operation scheme uploader 9 is used for issuing execution by it then with uploading to the data server 3 of enterprise by the adjusted new production operation scheme of optimized processor by TCP/IP network 2.

Adopt processing iron and steel enterprise color coating unit production disturbance process and device proposed by the invention, can respond and effectively handle the production disturbance of iron and steel enterprise's color coating unit fast, because this method is to come problem analysis for the viewpoint of global optimization, therefore can avoid existing method to be absorbed in the limitation of local optimum easily to a great extent, thereby guarantee the continuity of production better, improve the quality of products, improve the production operation efficiency of unit, reduce production costs.Figure 11 has provided at 60 actual production data, is compared by the resulting result of process proposed by the invention and the result of the manual method processing color coating unit production disturbance of reality.In the figure, utility index (Utilty) refers to by the improvement amount of the resulting new production operation scheme of production disturbance treatment method on product quality and unit production operation efficiency, and stability indicator refers to new production operation scheme and the departure of initial production operation scheme on the execution time.From the result as can be seen, the process of processing color coating unit production disturbance proposed by the invention on quality and stability, all to obviously be better than present steel enterprise the manual method that generally adopts; In addition, use the needed time of this process about a few minutes, be far smaller than the employed time of manual method, therefore can realize quick response to the production disturbance, thereby guarantee the stable operation of production, improve the quality of products, improve the unit equipment utilization factor, and then improve the unit production capacity, reduce production costs.

Description of drawings

Fig. 1 is the general technological process of production of color coating unit;

The simplified model that Fig. 2 switches for color and the transition volume defines;

Fig. 3 is the simplified model of initial production operation scheme;

Fig. 4 causes the time gap exemplary plot for the mechanical disorder of upstream unit;

Fig. 5 does not cause the time gap exemplary plot for coiled sheet arrives on time;

Fig. 6 causes the time gap exemplary plot for the coiled sheet quality dissatisfaction;

Fig. 7 is device and a workflow synoptic diagram thereof proposed by the invention;

Fig. 8 is candidate's coiled sheet proposed by the invention and the matching relationship synoptic diagram between the problem coiled sheet;

Fig. 9 is infeasible situation of time gap and restorative procedure synoptic diagram thereof;

Figure 10 is adjacent two boards base width bounce infeasible situation and restorative procedure synoptic diagram thereof;

Figure 11 is adjacent polylith width of plate slab bounce infeasible situation and restorative procedure synoptic diagram thereof;

Figure 12 compares synoptic diagram for process and the manual method result who uses processing iron and steel enterprise color coating unit production disturbance proposed by the invention;

Figure 13 is by the resulting new production operation scheme of the present invention synoptic diagram at specific embodiment;

Figure 14 is the process flow diagram of tabu search algorithm;

Figure 15 is the process flow diagram of simulated annealing;

Figure 16 is for becoming the process flow diagram of neighborhood search algorithm.

Embodiment

Be further described and replenish below in conjunction with the device of accompanying drawing process of the present invention and this method of enforcement.

Because color coating unit production operation system optimization technological process proposed by the invention realizes with C Plus Plus that for this point, following Example will add explanation no longer in addition in the specific implementation process that provides this method of use.If the specifying information of one section plan will repairing in the initial production operation scheme is as shown in table 1, row PTOP and PBOT represent coiled sheet respectively in this table first coat with rendering under color code, and row FTOP and FBOT represent respectively that the essence of coiled sheet is coated with essence and are coated with down color code.Row PT, AT, ST, FT and DT represent the processing time of coiled sheet respectively, time of arrival, plan begins process time, plan deadline and delivery date.Aborning, coiled sheet m ₄, m ₆And m ₈Owing to the influence of production disturbance becomes and can not utilize.Coiled sheet information in the preceding storehouse of color coating this moment except that the transition volume is as shown in table 2, and the quantity of candidate's coiled sheet is greater than the quantity of problem coiled sheet generally speaking.

Need the part of repairing in the table 1. initial production operation scheme

The details of table 2. candidate coiled sheet (not containing the transition volume)

It is as follows then to use the calm process of processing color coating unit production disturbance proposed by the invention to find the solution the implementation process of this problem:

Embodiment 1 adopts tabu search algorithm

The first step: the system when the production disturbance is occurred carries out approximate description

Utilize the raw data of above-mentioned table 1 and table 2, according to Matching Model proposed by the invention:

Minimize (1)

s.t.

$\underset{j\∈M}{\mathrm{\Σ}}{x}_{\mathrm{ij}}\≤1,$ $\∀i\∈N$ (2)

$\underset{i\∈N}{\mathrm{\Σ}}{x}_{\mathrm{ij}}=1,$ $\∀j\∈M$ (3)

x _ij∈{0，1}， $\∀i\∈N,$ $\∀j\∈M$ (4)

Coupling expense c between initialization problem coiled sheet and the candidate's coiled sheet _Ij, it comprises the weighted sum of the penalty value of the difference on width, thickness, weight, processing time, coating color, delivery date and time of arrival between candidate's coiled sheet i and the problem coiled sheet j.

Second step: with the model conversation of setting up is balance model

Introduce | N|-|M| virtual coiled sheet is in the set of problem coiled sheet, and wherein N is the set of candidate's coiled sheet, and M is the set of problem coiled sheet, obtains new problem coiled sheet set M '={ n ₁..., n ₁₀, and establish these virtual problem coiled sheets and the coupling expense between candidate's coiled sheet is L (a very big number), thereby be the balance Matching Model top model conversation:

Minimize

(5)

s.t.

$\underset{j\∈M^{\′}}{\mathrm{\Σ}}{x}_{\mathrm{ij}}=1,$ $\∀i\∈N$ (6)

$\underset{i\∈N}{\mathrm{\Σ}}{x}_{\mathrm{ij}}=1,$ $\∀j\∈M^{\′}$ (7)

x _ij∈{0，1}， $\∀i\∈N,$ $\∀j\∈M^{\′}$ (8)

The 3rd step: find the solution the balance Matching Model

Find the solution top balance Matching Model, the initial production operation scheme is carried out corresponding local adjustment, resulting new production operation scheme such as table 3 and shown in Figure 12.

Table 3. carries out adjusted production operation scheme according to Matching Model

The 4th step: adjusted production operation scheme is carried out feasibility inspection and reparation

Coiled sheet m in the resulting in the above new production operation scheme ₅And n ₃Between width knock-on appears, because this situation only relates to two adjacent coiled sheets, therefore use one best in following two kinds of restorative procedures: (1) exchanges the position of these two coiled sheets or (2) and replaces in existing two coiled sheets one with other best candidate's coiled sheet.Because first kind of restorative procedure promptly exchanges coiled sheet m ₅With coiled sheet n ₃After remain infeasible situation, therefore use second kind of restorative procedure, promptly from other candidate's coiled sheet, select best coiled sheet n ₆Substitute coiled sheet n ₃

The 5th step: use improved tabu search algorithm (or other intelligent optimization algorithm) to improve feasible new production operation scheme, the process flow diagram of this algorithm as shown in figure 14.

Step 1. beginnings, initialization;

Step 2. is provided with i=1, initialization maximum iteration time t _Max=500, the maximum running time T that allows _Max=500 seconds, deep search iterations t _Int=40, breadth first search's iterations t _Div=10, a=0.9 empties the taboo table.Current production schedules is preferably separated and current separating as history;

If Step were 3. i＞t _MaxOr running time of algorithm＞T _Max, the resulting best production schedules of output algorithm, and withdraw from; Otherwise enter Step4;

If Step 4. i are divided exactly (t _Int+ t _Div) remainder＞t _Int, forward Step 5 to; Otherwise use deep search that current separating improved, promptly use all neighborhood type i nner-relocation (certain locational coiled sheet in the plan unit being inserted in this plan unit other position), Swap (exchanging two locational coiled sheets in the plan unit), Outer-relocation (coiled sheet in the plan unit is inserted in other the plan unit) and Exchange (respectively selecting a coiled sheet respectively in two different plan unit exchanges), carry out neighborhood search one time, therefrom determine best feasible moving, judge to move whether can be applied to current separating: (1) if this moves in the taboo table and uses to move and can not improve history and preferably separate, and then gets good feasible this deterministic process that moves and repeat from all feasible neighborhoods move next time; (2) use if this moves in the taboo table to move and to improve history and preferably separate, then current separating used this and moved; (3), then current separating used this and moved if should move not in the taboo table.Applied mobile opposite moving joined in the taboo table, upgrade the adopted number of times of each neighborhood type, upgrade history and preferably separate, and put i=i+1, forward Step 3 then to;

Step 5. uses the breadth first search, promptly only being used the maximum a kind of neighborhood type of number of times in the stage with deep search carries out neighborhood search one time, best feasible the moving of therefrom finding out is applied directly to current separating, and with the adopted number of times zero clearing of each neighborhood type, applied mobile opposite moving joined in the taboo table, upgrade history and preferably separate, and put i=i+1.If i is (t _Int+ t _Div) integer multiple, mean that promptly breadth first search's stage carried out t continuously _DivIn generation, just put t _Div=t _Div* a reducing breadth first search's iterations in next circulation, and forwards Step 3 to; Otherwise forward Step 4 to;

The 6th step: carry out last resulting new production operation scheme by the production automatic control system of color coating unit.

The inventive system comprises following assembly:

(1) production disturbance watch-dog 1, its sensor are arranged on each reference mark of color coating unit, are used for monitoring in real time the production disturbance of color coating unit, gather the real-time information of production disturbance, and utilize TCP/IP network 2 that it is uploaded in the data server 3 of enterprise;

(2) TCP/IP network 2 is used for carrying out the data communication between data server 3, data acquisition unit 4 and the production operation scheme uploader 9 of color coating unit, enterprise;

(3) business data server 3 be used for storing the coiled sheet information in color coating unit initial production operation scheme, the preceding storehouse of current time color coating, the real-time information of the production disturbance that collected by production disturbance watch-dog, and the new production scheduling scheme after handling through calm method;

(4) data acquisition unit 4, are used for detecting and downloading the real-time information of production disturbance relevant with the color coating unit in the business data server 3, and are sent to data processor 5; Simultaneously, download the coiled sheet information that is coated with unit initial production operation scheme and the preceding storehouse of current time color coating, and it is saved in storer 6;

(5) data processor 5, be used for to the real-time information of the data collector 4 production disturbance relevant that downloaded with the color coating unit classify, format conversion, and it is saved in the storer 6;

(6) storer 6, are used for preserving color coating unit initial production operation scheme, and the real-time information of the color coating unit production disturbance after 5 processing of process data processor;

(7) optimized processor 7, wherein embedded the calm process software that is coated with unit production disturbance based on the iron and steel enterprise of Matching Model, by calculating output production operation scheme;

(8) based on the touch-screen display 8 of LCD, be used for showing the software system interface that embeds in this device, in time use form and figure dual mode display optimization processor 7 adjusted new color coating unit production operation schemes.

(9) production operation scheme uploader 9 is used for issuing execution command by it then with uploading to the data server 3 of enterprise by the adjusted new production operation scheme of optimized processor by TCP/IP network 2.

Embodiment 2 adopts improved simulated annealing

Present embodiment and embodiment 1 are roughly the same, and its difference only is that the processing in time of occurrence slit in the 4th step and the 5th step adopt simulated annealing, thus below in the process of narrating, only introduce this content of 2.

The 4th step: adjusted production operation scheme is carried out feasibility inspection and reparation:

When in the resulting adjusted production operation scheme by Matching Model during the life period slit, as shown in Figure 8, problem coiled sheet S _jProcessing time be 50 minutes, and replace its candidate's coiled sheet S ' _iProcessing time be 30 minutes.Because coiled sheet

The process time that may begin the earliest be s _j+ 50, therefore at coiled sheet S ' _iWith

Between still have a time gap of 20 minutes.In order to repair this infeasible situation, each time gap is all adopted following three kinds of methods, and therefrom select best a kind of of effect for use.

4. use another and coiled sheet S ' _iThe coupling expense minimum and satisfy processing time p _k〉=p _jCoiled sheet S ' _kReplace coiled sheet S ' _i

5. replace coiled sheet S ' with the combination of two or three best other candidate's coiled sheets that can fill this time gap _i

6. replace coiled sheet S ' with the combination (other candidate's coiled sheet and a transition volume) of two the best coiled sheets that can fill this time gap _i

The 5th step: use improved simulated annealing to improve feasible new production operation scheme, this algorithm adopts multiple neighbour structure, its process flow diagram as shown in figure 15:

Beginning, initialization;

Step 1. preferably separates current production schedules S as history.Initial temperature T is set ₀=500, final temperature T _f=20, cooling coefficient a=0.97 is provided with iterations k=1, and makes T _k=T ₀Definition use Inner-relocation move (position that certain locational coiled sheet in the plan unit is inserted in this plan unit other) the neighborhood type be numbered 1, use Swap move (exchange one and plans two locational coiled sheets in the unit) the neighborhood type be numbered 2, use Outer-relocation move (coiled sheet in the plan unit is inserted in other the plan unit) the neighborhood type be numbered 3, use Exchange move (respectively selecting a coiled sheet respectively in two different plan unit exchanges) the neighborhood type be numbered 4;

If Step is 2. T _k＜T _f, the resulting best production schedules of output algorithm, and withdraw from; Otherwise, change Step3;

Step3. 1,2,3, produce a random number r among the 4}, it is the scale of r kind neighborhood that N (r) is set, round-robin number of times i=0 in being provided with, and the current S of separating is set ₀=S;

Step 4. is for S ₀, in its neighborhood r, produce a neighborhood at random and separate S ' (promptly to S ₀Use moving at random of a r neighborhood type), and i=i+1 is set.The variation delta f=f of calculating target function (S ')-f (S ₀), if Δ f＜0 makes S ₀=S ' changes Step 4; Otherwise in [0,1], produce one and obey equally distributed random number ε, if exp (Δ f/T _k)＞ε then makes S ₀=S ', and change Step 5;

If Step5. i＞N (r) changes Step6; Otherwise, change Step4;

If Step is 6. f (S ₀)＜f (S) then upgrades S=S ₀Make k=k+1, T _k=T _k* a changes Step 2.

Embodiment 3 adopts and becomes the neighborhood search algorithm

Roughly the same with embodiment 1,2, its difference is that the 4th walks out of existing width or thickness and jump and exceed the scope of permission, adopt in the 5th step to become the neighborhood search algorithm, for avoiding repetition, thus below narration in only introduce the content of this two aspect.

The 4th step: adjusted production operation scheme is carried out feasibility inspection and reparation:

When checking, find that the width of adjacent coiled sheet or thickness jump and exceeded the scope that allows by the thick production operation scheme of the resulting adjustment of Matching Model.In order to repair this infeasible situation, can adopt following three kinds of methods this moment, and therefrom select best a kind of of effect for use:

4. replace the coiled sheet that this infeasible situation occurs with another best candidate's coiled sheet;

5. replace the coiled sheet that this infeasible situation occurs with the combination of two best in addition candidate's coiled sheets;

6. occurring inserting an only transition volume in the middle of the coiled sheet of this infeasible situation.

The 5th step: use to become the neighborhood search algorithm and improve feasible new production operation scheme, its process flow diagram as shown in figure 16:

Beginning, initialization;

Step 1. preferably separates current production schedules S as history, iterations i=0 is set, maximum iteration time I _Max=100; Definition use Inner-relocation move (position that certain locational coiled sheet in the plan unit is inserted in this plan unit other) the neighborhood type be numbered 1, use Swap move (exchange one and plans two locational coiled sheets in the unit) the neighborhood type be numbered 2, use Outer-relocation move (coiled sheet in the plan unit is inserted in other the plan unit) the neighborhood type be numbered 3, use Exchange move (respectively selecting a coiled sheet respectively in two different plan unit exchanges) the neighborhood type be numbered 4;

If Step were 2. i＞I _Max, the resulting best production schedules of output algorithm, and withdraw from; Otherwise, i=i+1 is set, and current neighborhood type k=1 is set, and change Step3;

If Step is 3. k＞and 4, change Step 2; Otherwise, change Step 4;

Step 4. produces a neighborhood at random and separates S ' (promptly S being used moving at random of a k neighborhood type) for S in its k neighborhood type; And the best neighborhood of searching for S ' in being numbered the neighborhood of k separates, and is designated as S "; If f (S ")＜f (S) puts S=S ", and k=1 is set; Otherwise k=k+1 is set; Change Step 3.

Adopt process and the device of handling iron and steel enterprise's color coating unit production disturbance of the present invention, can use the viewpoint of global optimization, repair all time gaps that caused by the production disturbance in the initial production operation scheme effectively, realization is carried out calm effectively to the every production disturbance that is occurred in the color coated steel sheet production run, thereby guaranteed the stable operation of iron and steel enterprise's color coating unit, both can reduce production costs; Can improve the quality of products again and usage ratio of equipment, have great application value.

Claims (8)

1, a kind ofly handles the process that iron and steel enterprise's color coating is coated with unit production disturbance, it is characterized in that: based on Matching Model, repair all time gaps that caused by the production disturbance in the initial production operation scheme simultaneously with the viewpoint of global optimization, comprise following six steps altogether:

The first step: the system when the production disturbance is occurred carries out approximate description, when the production disturbance appears in the color coating unit, suppose that the set of the coiled sheet that goes wrong in the initial production operation scheme is M={1,2, ..., m}, this moment, the set of candidate's coiled sheet was N={1,2, ..., n} notices that transition volumes all in the preceding storehouse of color coating is also contained among the N, and other coiled sheet is called the contract coiled sheet is arranged; Based on Matching Model, the color coating unit production disturbance optimization process problem that the present invention considered can approximate description be:

$\mathrm{Minimize}\underset{i\∈N}{\mathrm{\Σ}}\underset{j\∈M}{\mathrm{\Σ}}{c}_{\mathrm{ij}}{x}_{\mathrm{ij}}---\left(1\right)$

s.t.

$\underset{j\∈M}{\mathrm{\Σ}}{x}_{\mathrm{ij}}\≤1,$ $\∀i\∈N---\left(2\right)$

$\underset{i\∈N}{\mathrm{\Σ}}{x}_{\mathrm{ij}}=1,$ $\∀j\∈M---\left(3\right)$

x _ij∈{0，1}， $\∀i\∈N,$ $\∀j\∈M---\left(4\right)$

In objective function (1), c _IjBe the coupling expense between candidate's coiled sheet i and the problem coiled sheet j, mainly comprise the weighted sum of the penalty value of the difference on width, thickness, weight, processing time, coating color, delivery date and time of arrival between them; x _IjIf matching relationship between expression candidate coiled sheet i and the problem coiled sheet j is promptly with candidate's coiled sheet i substitution problem coiled sheet j, then x _Ij=1, otherwise x _Ij=0; Constraint (2) and (3) is the technology and the operation constraint condition of system self, and constraint (2) guarantees that each candidate's coiled sheet is merely able to mate a problem coiled sheet at most; Constraint (3) guarantees that each problem coiled sheet must mate candidate's coiled sheet; Constraint (4) guarantees decision variable x _IjCan only get 0 or 1; Because the quantity of transition volume is always many in the preceding storehouse, can suppose n 〉=m, so this model is a nonequilibrium Matching Model;

Second step: with the model conversation of being set up in the above-mentioned first step is balance model, thereby obtain new problem coiled sheet set M '=MY{m+1 by in problem coiled sheet set M, introducing n-m virtual problem coiled sheet, ..., n}, and establish these virtual problem coiled sheets and the coupling expense between candidate's coiled sheet is L (a very big number), then Shang Mian non-equilibrium matching problem model can become an equilibrium problem model (promptly | M ' |=| N|):

$\mathrm{Minimize}\underset{i\∈N}{\mathrm{\Σ}}\underset{j\∈M^{\′}}{\mathrm{\Σ}}{c}_{\mathrm{ij}}{x}_{\mathrm{ij}}---\left(5\right)$

s.t.

$\underset{j\∈M^{\′}}{\mathrm{\Σ}}{x}_{\mathrm{ij}}=1,$ $\∀i\∈N---\left(6\right)$

$\underset{i\∈N}{\mathrm{\Σ}}{x}_{\mathrm{ij}}=1,$ $\∀j\∈M^{\′}---\left(7\right)$

x _ij∈{0，1}， $\∀i\∈N,$ $\∀j\∈M^{\′}---\left(8\right)$

The 3rd step: find the solution the balance Matching Model

Use the shortest increase road algorithm (shortest augmenting path algorithm) to find the solution this balance matching problem, and the initial production operation scheme is carried out part adjustment according to resulting optimum solution; Generally, be feasible by the resulting new production operation scheme of this model;

The 4th step: adjusted production operation scheme is carried out feasibility inspection and reparation;

The 5th step: use intelligent optimization algorithm that new production operation scheme is improved;

The 6th step: carry out last resulting new production operation scheme by the production automatic control system of color coating unit..

2. by the described process of handling iron and steel enterprise's color coating unit production disturbance of claim 1, it is characterized in that existing in said the 4th step first kind of infeasible situation promptly still during the life period slit, can adopt following three kinds of heuristics to each time gap, and therefrom select best a kind of of effect for use:

(1) with another and coiled sheet S ' _iThe coupling expense minimum and satisfy processing time p _k〉=p _jCoiled sheet S ' _kReplace coiled sheet S ' _i

(2) replace coiled sheet S ' with the combination of two or three best other candidate's coiled sheets that can fill this time gap _i

(3) replace coiled sheet S ' with the combination (other candidate's coiled sheet and a transition volume) of two the best coiled sheets that can fill this time gap _i

3. by the process of the described processing of claim 1 iron and steel enterprise color coating unit production disturbance, when it is characterized in that existing second kind of infeasible situation to be the width knock-on in said the 4th step, repair this infeasible situation, also will consider two kinds of situations:

(1) if this situation includes only two adjacent coiled sheets, then the method for Xiu Fuing can be replaced in existing two coiled sheets one for the position that exchanges these two coiled sheets or with other best candidate's coiled sheet;

(2) if this situation comprises the coiled sheet of several successive, then the method for Xiu Fuing can be for rearranging these coiled sheets or replacing some coiled sheet in these coiled sheets with other best candidate's coiled sheet.

4. by the described process of handling iron and steel enterprise's color coating unit production disturbance of claim 1, it is characterized in that having the third infeasible situation in said the 4th step width or the thickness that are adjacent coiled sheet to jump when having exceeded the scope that allows, in order to repair this infeasible situation, can adopt three kinds of heuristics, and therefrom select best a kind of of effect for use:

(1) replaces the coiled sheet that this infeasible situation occurs with another best candidate's coiled sheet;

(2) replace the coiled sheet that this infeasible situation occurs with the combination of two best in addition candidate's coiled sheets;

(3) occurring inserting an only transition volume in the middle of the coiled sheet of this infeasible situation.

5. by the process of the described processing of claim 1 iron and steel enterprise color coating unit production disturbance, it is characterized in that the intelligent optimization algorithm in said the 5th step can be a tabu search algorithm, its computing flow process is as follows:

Step 1. is provided with i=1, initialization maximum iteration time t _Max, the maximum running time T that allows _Max, deep search iterations t _Int, breadth first search's iterations t _Div, a=0.9 empties the taboo table; Current production schedules is preferably separated and current separating as history;

If Step were 2. i＞t _MaxOr running time of algorithm＞T _Max, the resulting best production schedules of output algorithm, and withdraw from; Otherwise enter Step 3;

If Step 3. i are divided exactly (t _Int+ t _Div) remainder＞t _Int, forward Step 4 to; Otherwise use deep search that current separating improved, promptly use all neighborhood type i nner-relocation (certain locational coiled sheet in the plan unit being inserted in this plan unit other position), Swap (exchanging two locational coiled sheets in the plan unit), Outer-relocation (coiled sheet in the plan unit is inserted in other the plan unit) and Exchange (respectively selecting a coiled sheet respectively in two different plan unit exchanges), carry out neighborhood search one time, therefrom determine best feasible moving, judge to move whether can be applied to current separating: (1) if this moves in the taboo table and uses to move and can not improve history and preferably separate, and then gets good feasible this deterministic process that moves and repeat from all feasible neighborhoods move next time; (2) use if this moves in the taboo table to move and to improve history and preferably separate, then current separating used this and moved; (3) if should move not in the taboo table, then current separating used this and moved, applied mobile opposite moving joined in the taboo table, upgrade the adopted number of times of each neighborhood type, upgrade history and preferably separate, and put i=i+1, forward Step2 then to;

Step 4. uses the breadth first search, promptly only being used the maximum a kind of neighborhood type of number of times in the stage with deep search carries out neighborhood search one time, best feasible the moving of therefrom finding out is applied directly to current separating, and with the adopted number of times zero clearing of each neighborhood type, applied mobile opposite moving joined in the taboo table, upgrade history and preferably separate, and put i=i+1; If i is (t _Int+ t _Div) integer multiple, mean that promptly breadth first search's stage carried out t continuously _DivIn generation, just put t _Div=t _Div* a reducing breadth first search's iterations in next circulation, and forwards Step 2 to; Otherwise forward Step 3 to.

6. by the process of the described processing of claim 1 iron and steel enterprise color coating unit production disturbance, it is characterized in that the intelligent optimization algorithm in said the 5th step can be simulated annealing, this algorithm adopts multiple neighbour structure, and its flow process is as follows:

Step 1. preferably separates current production schedules S as history; Initial temperature T is set ₀=500, final temperature Tf=20, cooling coefficient a=0.97 is provided with iterations k=1, and makes T _k=t ₀Definition use Inner-relocation move (position that certain locational coiled sheet in the plan unit is inserted in this plan unit other) the neighborhood type be numbered 1, use Swap move (exchange one and plans two locational coiled sheets in the unit) the neighborhood type be numbered 2, use Outer-relocation move (coiled sheet in the plan unit is inserted in other the plan unit) the neighborhood type be numbered 3, use Exchange move (respectively selecting a coiled sheet respectively in two different plan unit exchanges) the neighborhood type be numbered 4;

If Step is 2. T _k＜T _f, the resulting best production schedules of output algorithm, and withdraw from; Otherwise, change Step3;

Step 3. 1,2,3, produce a random number r among the 4}, it is the scale of r kind neighborhood that N (r) is set, round-robin number of times i=0 in being provided with, and the current S of separating is set ₀=S;

Step 4. is for S ₀, in its neighborhood r, produce a neighborhood at random and separate S ' (promptly to s ₀Use moving at random of a r neighborhood type), and i=i+1 is set; The variation delta f=f of calculating target function (S ')-f (S ₀), if Δ f＜0 makes S ₀=S ' changes Step 4; Otherwise in [0,1], produce one and obey equally distributed random number ε, if exp (Δ f/T _k)＞ε then makes S ₀=S ', and change Step 5;

If Step were 5. i＞N (r), would change Step 6; Otherwise, change Step 4;

If Step is 6. f (S ₀)＜(S) then upgraded S=S ₀Make k=k+1, T _k=T _k* a changes Step 2.

7. by the process of the said processing of claim 1 iron and steel enterprise color coating unit production disturbance, it is characterized in that the intelligent optimization algorithm in said the 5th step can be to become the neighborhood search algorithm, its flow process is as follows:

Step 1. preferably separates current production schedules S as history, iterations i=0 is set, maximum iteration time I _Max=100; Definition use Inner-relocation move (position that certain locational coiled sheet in the plan unit is inserted in this plan unit other) the neighborhood type be numbered 1, use Swap move (exchange one and plans two locational coiled sheets in the unit) the neighborhood type be numbered 2, use Outer-relocation move (coiled sheet in the plan unit is inserted in other the plan unit) the neighborhood type be numbered 3, use Exchange move (respectively selecting a coiled sheet respectively in two different plan unit exchanges) the neighborhood type be numbered 4;

If Step were 2. i＞I _Max, the resulting best production schedules of output algorithm, and withdraw from; Otherwise, i=i+1 is set, and current neighborhood type k=1 is set, and change Step3;

If Step is 3. k＞and 4, change Step 2; Otherwise, change Step 4;

Step 4. produces a neighborhood at random and separates S ' (promptly S being used moving at random of a k neighborhood type) for S in its k neighborhood type; And the best neighborhood of searching for S ' in being numbered the neighborhood of k separates, and is designated as S "; If f (S ")＜f (S) puts S=S ", and k=1 is set; Otherwise k=k+1 is set; Change Step 3.

8. one kind is exclusively used in the described device of handling the process of iron and steel enterprise's color coating unit production disturbance of enforcement claim 1, it is characterized in that it comprises following assembly:

(1) production disturbance watch-dog 1 is used for monitoring in real time the production disturbance of color coating unit, gathers the real-time information of production disturbance, and utilizes TCP/IP network 2 that it is uploaded to the data server 3 of enterprise;

(2) TCP/IP network 2 is used for carrying out the data communication between data server 3, data acquisition unit 4 and the production operation scheme uploader 9 of color coating unit, enterprise;

(3) business data server 3 be used for storing the coiled sheet information in color coating unit initial production operation scheme, the preceding storehouse of current time color coating, the real-time information of the production disturbance that collected by production disturbance watch-dog, and the new production scheduling scheme after handling through calm method;

(4) data acquisition unit 4, are used for detecting and downloading the real-time information of production disturbance relevant with the color coating unit in the business data server 3, and are sent to data processor 5; Simultaneously, download the coiled sheet information that is coated with unit initial production operation scheme and the preceding storehouse of current time color coating, and it is saved in storer 6;

(5) data processor 5, be used for to the real-time information of the data collector 4 production disturbance relevant that downloaded with the color coating unit classify, format conversion, and it is saved in the storer 6;

(6) storer 6, are used for preserving color coating unit initial production operation scheme, and the real-time information of the color coating unit production disturbance after 5 processing of process data processor;

(7) optimized processor 7, are the core components of this device, wherein embedded the calm process that is coated with unit production disturbance based on the iron and steel enterprise of Matching Model;

(8) based on the touch-screen display 8 of LCD, be used for showing the software system interface that embeds in this device, and with form and figure dual mode display optimization processor 7 adjusted new color coating unit production operation schemes;

(9) production operation scheme uploader 9 is used for issuing execution by it then with uploading to the data server 3 of enterprise by the adjusted new production operation scheme of optimized processor by TCP/IP network 2.

Images