CN101441468A - Network coordinative production scheduling system based on Virtual-Hub and self-adapting scheduling method thereof - Google Patents

Abstract

The invention relates to a Virtual-hub-based network coordinated production scheduling system, which comprises an information sharing module, a plan scheduling module and a Virtual-Hub of a production synchronization module, wherein the information sharing module comprises an enterprise basic information module, a product material information module, and a plan scheduling information module; the plan scheduling module comprises plan decomposition, task distribution and dynamic scheduling; and the production synchronization module is used to send latest production information commands to other manufactures on a production chain to ensure the longitudinal synchronization of enterprises in coordinated production plan. A self-adaptive scheduling method based on network coordinated production aims to obtain a better or good plan scheduling policy at cost as low as possible when an uncertain event happens in an enterprise at a time T during coordinated production, so as to readjust internal production plan dynamically.

Description

Network cooperation Production Scheduling System and self-adapting dispatching method thereof based on Virtual-Hub

Technical field

The present invention relates to the method and system of cooperation in production and control, the cooperation in production under especially a kind of manufacturing network production environment and the method and system of control.

Background technology

Networking coordinated production process is a kind of blocking, autonomous, parallel, virtual, distributed manufacture process, and manufacturing enterprise dynamically organizes together in order to finish common task.This multiple enterprises coordinated production pattern is because each enterprise participates in many production chains respectively, complicated production system of crisscross formation.Networking manufacturing respectively participates in enterprise and is in the dynamic changeable environment, and uncertainty and dynamic that enterprise produces inside and outside environment increase, and the generation of uncertainty event then is the incitant that coordinated production enterprise Production Scheduling is dynamically adjusted.According to producing root, uncertainty event in the multiple enterprises coordinated production can be divided into three kinds: an accident for enterprise's manufacturing process itself is meant that mainly enterprises key equipment fault or other irresistible extraneous factors cause the delay and the interruption of manufacture process; The 2nd, down-stream enterprise or project promoter require to make an earlier shipment, rush order is inserted, the cancellation or the modification of former division of labor task; The 3rd, the upstream collaborative enterprise can't provide material or accessory etc. on time according to quantity.Therefore, under networked production environment, need carry out dynamic monitoring to the manufacturing schedule and the production schedule of collaborative enterprise, the adjustment that participates in the indivedual internal pair production plans of enterprise simultaneously will produce that other enterprises are not target on the chain not influence as far as possible.

Through to the patent of invention of prior art retrieval back discovery, Chinese patent application number is: 200710192015.2, and denomination of invention is: a kind of by computer implemented adaptively selected dynamic production scheduling control system.This patented claim has proposed a kind of adaptively selected dynamic production scheduling control system, this technology is introduced the regulation goal function and is proposed a kind of method for optimizing scheduling, on the basis of existing scheduling rule, constantly obtain efficient scheduling knowledge alternately by learner and manufacturing system, be used for the optimization production process.But that this invention only is applicable to is unsettled, the time Workshop Dynamic production scheduling that becomes and do not relate to coordinated production scheduling between enterprise.

It is following several that the production scheduling problems of other network cooperation productions has the research of the Production Scheduling aspect of certain homophylic supply chain and virtual enterprise to mainly contain:

(1) D.F.Kehoe of Liverpool university etc. has proposed a kind of Production Planning and Controlling strategy of supply chain, and has set up towards the production schedule of Web and the C/S model of scheduling.

(2) the Latifa Ouzizi of French MACSI project etc. has proposed based on the architecture of the virtual enterprise of Agent with based on planning model and the operation frame consulted.

(3) Wang Wanshan of Northeastern University etc. has made up a kind of strange land and has produced the cooperative scheduling framework, and carries out dynamic simulation based on this scheduling framework and calculate, and obtains the feasible solution between enterprise's external coordination production and the internal schedule.

The research overview of comprehensive literature as can be seen, the research for the Production Scheduling aspect of supply chain and virtual enterprise at present mainly concentrates on architecture and framework aspect, lack Production Scheduling scheme complete special network-oriented coordinated production, that realize entire system optimization, the work of this part has much room for improvement to satisfy actual needs.

In further retrieving, find as yet so far and subject key words " network cooperation production ", " Virtual-Hub " identical with " adaptive scheduling " or similar bibliographical information and patent of invention.

Summary of the invention

The objective of the invention is at the deficiencies in the prior art, propose a kind of network cooperation Production Scheduling System and self-adapting dispatching method thereof, to solve the deficiency that exists in the background technology based on Virtual-Hub.On the one hand provide a kind of multiple enterprises coordinated production dispatching system under the network environment that is applicable to based on Virtual-Hub, the complicacy that shielding multiple enterprises production information is shared is convenient to the collaborative enterprise production of cooperating; Based on multiobject heuristic search algorithm, provide the self-adapting dispatching method of enterprises under a kind of network cooperation production on the other hand, realize dynamically adjusting the operation plan of this enterprise.

Technical scheme of the present invention is as follows:

A kind of network cooperation Production Scheduling System based on Virtual-Hub.Virtual-Hub is a kind of virtual platform based on Web consensus standard and J2EE technology, be primarily aimed at the various uncertainties that face in the multiple enterprises coordinated production process, in real time production information share, the production schedule synchronously and dynamically demand such as adjustment and " Virtual Space " that multiple enterprises coordinated production process is synchronous of building, with the shared complicacy of shielding multiple enterprises production information, finish the dynamic monitoring of producing node, reach the coordination purpose, promote the extensive coordinated production between enterprise.Virtual-Hub realizes being divided into three parts from design:

Presentation layer mainly is to be responsible for information is represented quickly and accurately to the user.During design structure, pattern and the behavior of presentation layer are finished in different files, and adopted the AJAX technology to support user's asynchronous request, make multiple enterprises in collaborative manufacturing environment, access speed faster be arranged.

Operation layer is mainly handled the various requests that participate in enterprise, calls various workflows, the operation of control total system.Be divided into three during the operation layer design: controller module is responsible for the various requests of process user; Service module is handled complicated workflow, as the distribution of order task, the scheduling of unusual task etc.; Access modules is the visit work of process database then, and main and database is come into contacts with, and the database access details all is encapsulated in this layer, as the type of database-driven etc.

Data Layer is responsible for the business datum of safe and reasonable ground storage and management enterprise, this layer preservation be all kinds of production informations of enterprise, as product information, facility information and design data etc.

Comprise information sharing module, planned dispatching module and produce synchronization module based on Virtual-Hub network cooperation Production Scheduling System.The information sharing module comprises enterprise's Back ground Information model, product material information model and planned dispatching information model; The planned dispatching module comprises plan decomposition, Task Distribution and dynamic dispatching; Produce synchronization module and be used for that all the other manufacturing enterprises send up-to-date production information instruction on the chain to producing, guarantee each enterprise in the coordinated production plan vertically synchronously.

Self-adapting dispatching method under a kind of network cooperation production.In the network cooperation production run, the A of enterprise accepts n different orders in the section between at a time, and the time of delivery of each order is respectively d ₁, d ₂..., d _n, according to the urgency level of task, the back order of different orders all has different penalty coefficients.Certain moment T in the coordinated production process, a certain uncertainty event takes place in the A of enterprise, normal operation in order to ensure many coordinated production chains, the A of enterprise must finish the response to this uncertainty event as early as possible, dynamically adjust the production schedule, to eliminate of the influence of this incident to cooperative enterprise on this enterprise and other many production chains.

The heuristic search algorithm that self-adapting dispatching method of the present invention mainly guides based on constraint, in new production environment, with original production planned dispatching strategy is original state, at producing manufacturing constraint new on the chain, finish minimum duration deviation and the minimum objective function that turns to of cost deviation with task, introduce the deviation penalty as the auto-adaptive parameter adjustment function, the feasible solution of constraint condition is satisfied in searching, obtain the planned dispatching strategy of more excellent or suboptimum as far as possible with lower cost, thereby dynamically adjust the inner production schedule again.

(1) it is as follows to set up production models:

$\mathrm{Min}\underset{i}{\overset{n}{\mathrm{\Σ}}}{R}_i=\mathrm{Min}(\underset{i}{\overset{n}{\mathrm{\Σ}}}\max \{C_{\mathrm{ie}},d_i\}-\underset{i}{\overset{n}{\mathrm{\Σ}}}{d}_i)R_i=\max (0,C_{\mathrm{ie}}-d_i)=\max \{C_{\mathrm{ie}},d_i\}-d_i,i=\mathrm{1,2},...,n$

In the formula, R _IjBe the deviation of product j actual delivery time and estimated time of delivery among the order i, n is the total orders in the cooperation manufacturing enterprise planned time section, d _iBe the time of delivery of the different orders of cooperation manufacturing enterprise, C _IeActual completion date for order i.

(2) definition decision variable γ _Ij(t), when producing in the t period, gets the product j among the order i γ _Ij(t)=1, get γ under other situations _Ij(t)=0.

The actual completion date of some orders is: $C_{\mathrm{ie}}=\underset{j}{\overset{m}{\mathrm{\Σ}}}{C}_{\mathrm{ijs}}+\underset{j}{\overset{m}{\mathrm{\Σ}}}\underset{t}{\overset{T}{\mathrm{\Σ}}}{q}_j\·t\·{\mathrm{\γ}}_{\mathrm{ij}}\left(t\right)$

Wherein, the product category that can provide of cooperation manufacturing enterprise is provided m, and t is the period quantity of dividing in the cooperation plan phase, t=1, and 2 ..., T, q _JtBe the specific productivity of t period enterprise for product j, C _IjsBe the beginning production time of the product j among the order i.

(3) set up the cost minimum model:

$\mathrm{Min}\underset{i}{\overset{n}{\mathrm{\Σ}}}\underset{j}{\overset{m}{\mathrm{\Σ}}}\underset{t}{\overset{T}{\mathrm{\Σ}}}{\mathrm{\δ}}_j\·q_j\·{\mathrm{\γ}}_{\mathrm{ij}}\left(t\right)+\underset{i}{\overset{n}{\mathrm{\Σ}}}\underset{j}{\overset{m}{\mathrm{\Σ}}}\underset{t}{\overset{T}{\mathrm{\Σ}}}{\mathrm{\α}}_j\·(d_i-t)\·q_j\·{\mathrm{\γ}}_{\mathrm{ij}}\left(t\right)+\underset{i}{\overset{n}{\mathrm{\Σ}}}\underset{j}{\overset{m}{\mathrm{\Σ}}}\underset{t}{\overset{T}{\mathrm{\Σ}}}{\mathrm{\β}}_j(t-d_i)\·q_j\·{\mathrm{\γ}}_{\mathrm{ij}}\left(t\right)$

Wherein, δ _jBe the production cost coefficient of product j, α _jFor the stock of product j deposits cost, β _iThe penalty that departs from delivery date for order i.

(4) above-mentioned two targets are combined into the form of goal programming, then its Optimization Model can be expressed as:

$\mathrm{Min}{W}_R\underset{t}{\mathrm{\Σ}}\underset{i}{\mathrm{\Σ}}(C_{\mathrm{it}}^{+}+C_{\mathrm{it}}^{-})+W_F{J}_F^{-}$

S.t. λ _Ijt≤ q _Jt,

(ability constraint)

$\underset{i}{\overset{n}{\mathrm{\Σ}}}\underset{j}{\overset{m}{\mathrm{\Σ}}}\underset{t}{\overset{T}{\mathrm{\Σ}}}{\mathrm{\α}}_j\·(d_i-t)\·q_j\·{\mathrm{\γ}}_{\mathrm{ij}}\left(t\right)\≤Q,\∀i,j,t,$

Wherein, W _R, W _FBe respectively the optimization aim 1 of enterprise plan and 2 weight coefficient, be enough big constant,

With

Be respectively this enterprise and the extension of order i is departed from the time and depart from the time in advance in the t stage,

Be the amount of being unrealized of manufacturing enterprise's cost, λ _IjtWhen producing for product j among the t period order i to the ability need coefficient of enterprise, the inventory cost that Q can provide for the maximum of enterprise.

(5) model in the 4th step is carried out multiobjective Dynamic Optimization, but the present invention proposes a kind of variable weight and self-adapted genetic algorithm, according to the individual instances in the evolutionary process, adjust Different Optimization objective weight coefficient and cross and variation operator automatically, thereby obtain the global optimization result.

1. the weight coefficient of multiple-objection optimization is handled

Suppose that certain function has q optimization aim:

Min{z ₁＝f ₁(x)，z ₂＝f ₂(x)，...，z _q＝f _q(x)}　　s.t.gi(x)≤0，i＝1，2，...，m

Be defined as follows two parameters:

$z^{+}=\{z_1^{\max },z_2^{\max },\·\·\·,z_q^{\max }\}$

$z^{-}=\{z_1^{\min },z_2^{\min },\·\·\·,z_q^{\min }\}$

Wherein,

Be respectively the maximal value and the minimum value of k target in the current population.The weight of target k can adopt following formula to find the solution:

${\mathrm{\ω}}_k=\frac{1}{z_k^{\max }-z_k^{\min }},k=\mathrm{1,2},\·\·\·,q$

Corresponding given individuality, above-mentioned two optimization aim can be determined according to following formula:

$z\left(x\right)=\underset{k=1}{\overset{q}{\mathrm{\Σ}}}{\mathrm{\ω}}_k(z_k^{\max }-z_k)$

$=\underset{k=1}{\overset{q}{\mathrm{\Σ}}}\frac{(z_k^{\max }-z_k)}{(z_k^{\max }-z_k^{\min })}$

In the individuality of next round is evolved, new variation all will take place in weight, constantly upgrade, and optimize direction towards forward and advance.

2. Adaptive Genetic operator operation

During evolution, for guaranteeing not to be absorbed in local optimum, introduce fitness rate of change parameters C R weigh up and down two generation population variation:

$\mathrm{CR}=\frac{{\mathrm{fit}}_t^{\max }-{\stackrel{\‾}{\mathrm{fit}}}_t}{{\mathrm{fit}}_t^{\max }-{\mathrm{fit}}_t^{\min }}/\frac{{\mathrm{fit}}_{(t-1)}^{\max }-{\stackrel{\‾}{\mathrm{fit}}}_{(t-1)}}{{\mathrm{fit}}_{(t-1)}^{\max }-{\mathrm{fit}}_{(t-1)}^{\min }}$

Wherein

With Be respectively t for maximum adaptation degree value, minimum fitness value and average fitness value in the population, CR represent up and down two generation population be illustrated in rate on the fitness value.If CR is excessive, population might enter local optimum, needs to reduce to intersect and mutation operator; If CR is too small, the speed of evolution weakens, and can strengthen in real time to intersect and mutation operator.

The variation of Adaptive Genetic operator can be tried to achieve by following formula:

Can learn by above-mentioned steps, the present invention is in network production scheduling process, but adopting variable weight and self-adapted genetic algorithm is optimization means, according to the individual instances in the evolutionary process, automatically adjust weight coefficient, the cross and variation operator of different multiple-objection optimizations, to obtain the global optimization result, obtain the planned dispatching strategy of more excellent or suboptimum with lower cost as far as possible, thus the inner production schedule of dynamic adjustment again.

The present invention has substantive distinguishing features, has solved problems of the prior art, and a kind of dispatching system and dispatching method that is used under the network cooperation production is provided.1. the network cooperation dispatching system based on Virtual-Hub provides " Virtual Space " that multiple enterprises coordinated production process is synchronous, and dynamic monitoring respectively participates in the production node of enterprise, is convenient to the collaborative enterprise production of cooperating; 2. the production characteristics that existing Workshop Production dispatching method incompatibility network cooperation is produced, self-adapting dispatching method among the present invention is because of the multiple enterprises cooperative scheduling of having considered under the network cooperation production, can solve the network cooperation production scheduling problems according to the production characteristics of uncertainty of producing inside and outside environment and dynamic increase, obtain the planned dispatching strategy of more excellent or suboptimum with lower cost as far as possible, thus the dynamic production schedule of adjusting inside again.

Description of drawings

Fig. 1 is based on Virtual-Hub coordinated production dispatching system structural representation.

Fig. 2 is the network cooperation Production Scheduling System technological frame figure based on Virtual-Hub.

Embodiment

The present invention is further illustrated below in conjunction with content of the present invention and the described embodiment of accompanying drawing.

The invention provides a kind of network cooperation Production Scheduling System based on Virtual-Hub, its structure as shown in Figure 1.Network cooperation Production Scheduling System based on Virtual-Hub is the network virtual platform that just can visit by standard Web agreement (HTTP, WSDL) and Servelet engine.

Native system has adopted the B/S framework on one-piece construction, main portability more intense J2EE technology, HTML, JSP (Java Service Pages), XML and the Web Service technology of adopting in system constructing.Simultaneously, the increase income framework and the JFreeChart drawing assemblies such as Hibernate, Struts, Spring and ibatis of present comparative maturity have been adopted according to the characteristics of system.Server end adopts the JSP technology to realize front-end platform, and background module uses the J2EE technological development, and underlying database adopts MySQL 5.1, and Web server has adopted the Apache Tomcat Server that increases income.In front-end A pplication integrated, according to the different characteristics of each Application, adopted various technology correspondingly, comprise technology such as JavaScript, VBScript, Ajax.In the operation of each member enterprise of terminal user, as long as the employing browser just can normal browsing.

Network cooperation Production Scheduling System and self-adapting dispatching method based on Virtual-Hub realize that as shown in Figure 2 its detailed step is as follows:

(1) information sharing module mainly is responsible for obtaining Enterprise Resource information, product resource information and planned dispatching information etc. and is carried out information interaction with each member enterprise;

(2) planned dispatching module comprises plan decomposition, Task Distribution and dynamic dispatching, and this module has realized the formulation of the order MPS that network cooperation is produced, the thickness distribution and the dynamic dispatching of order;

(3) leader enterprise at first decomposes the order that receives, and formulates the MPS of a certain order.According to the various information that obtain in the information sharing module, order is carried out Task Distribution then;

(4) dispatching and monitoring essence is the dynamic management of production schedule implementation status, and task participates in the regular update user of enterprise manufacturing schedule under the normal condition, feeds back to leader enterprise.Upgrade manufacturing schedule if the leader need participate in enterprise, the request that also can send requires it to upgrade current Task Progress to participating in enterprise.When occur planning execution when unusual monitoring in time handle, send and remind unusually and carry out dynamic dispatching;

(5) certain moment T in the coordinated production process, after a certain uncertainty event takes place in enterprise, if the minimum duration deviation of the dynamic plan adjustment of enterprise satisfies certain condition, the planned dispatching module of Virtual-Hub will be triggered, according to the adaptive scheduling optimization method dynamically re-dispatching is carried out in the execution of the production schedule, generate the collaborative production schedule;

(6) produce synchronization module: the collaborative production schedule after the generation through with the negotiation of each member of an alliance enterprise after, will send up-to-date production schedule instruction after optimizing to producing each manufacturing enterprise on the chain, guarantee each member of an alliance enterprise in the coordinated production plan vertically synchronously.

The present invention has taken into full account the uncertainty under the networked manufacturing environment and the features such as opening of resource, for solving network cooperation Production Scheduling problem exercisable a network cooperation Production Scheduling System and a self-adapting dispatching method based on Virtual-Hub have been proposed, the network interface of system to enterprise is provided, make whole network cooperation production not be subjected to operating system, restrictions such as operating platform, the enterprises heterogeneous system can virtual platform be mutual therewith easily, adapt to the distribution autonomy between the enterprise and between the enterprises manufacturing cell under the networked manufacturing environment well, the characteristic of cooperative cooperating has realized good versatility and reconfigurability.

The above-mentioned description to embodiment is can understand and apply the invention for ease of those skilled in the art.The person skilled in the art obviously can easily make various modifications to these embodiment, and needn't pass through performing creative labour being applied in the General Principle of this explanation among other embodiment.Therefore, the invention is not restricted to the embodiment here, those skilled in the art should be within protection scope of the present invention for improvement and modification that the present invention makes according to announcement of the present invention.

Claims (5)

1. the network cooperation Production Scheduling System based on Virtual-Hub is characterized in that: the Virtual-Hub that comprises information sharing module, planned dispatching module and production synchronization module; Described information sharing module comprises enterprise's Back ground Information model, product material information model and planned dispatching information model; Described planned dispatching module comprises plan decomposition, Task Distribution and dynamic dispatching; Described production synchronization module is used for that all the other manufacturing enterprises send up-to-date production information instruction on the chain to producing, guarantee each enterprise in the coordinated production plan vertically synchronously.

2. the self-adapting dispatching method under the network cooperation production, it is characterized in that: at certain moment T in the coordinated production process, after a certain uncertainty event takes place in the A of enterprise, normal operation in order to ensure many coordinated production chains, heuristic search algorithm based on the constraint guiding, in new production environment, with original production planned dispatching strategy is original state, at producing manufacturing constraint new on the chain, finish minimum duration deviation and the minimum objective function that turns to of cost deviation with task, introduce the deviation penalty as the auto-adaptive parameter adjustment function, the feasible solution of constraint condition is satisfied in searching, obtain the planned dispatching strategy of more excellent or suboptimum as far as possible with lower cost, thereby dynamically adjust the inner production schedule again.

3, method according to claim 2 is characterized in that:

Comprise the steps:

(1) it is as follows to set up production models:

$\mathrm{Min}\underset{i}{\overset{n}{\mathrm{\Σ}}}{R}_i=\mathrm{Min}(\underset{i}{\overset{n}{\mathrm{\Σ}}}\max \{C_{\mathrm{ie}},d_i\}-\underset{i}{\overset{n}{\mathrm{\Σ}}}{d}_i)R_i=\max (0,C_{\mathrm{ie}}-d_i)=\max \{C_{\mathrm{ie}},d_i\}-d_i,i=\mathrm{1,2},...,n$

In the formula, R _IjBe the deviation of product j actual delivery time and estimated time of delivery among the order i, n is the total orders in the cooperation manufacturing enterprise planned time section, d _iBe the time of delivery of the different orders of cooperation manufacturing enterprise, C _IeActual completion date for order i;

(2) definition decision variable γ _Ij(t), when producing in the t period, gets the product j among the order i γ _Ij(t)=1, get γ under other situations _Ij(t)=0;

The actual completion date of some orders is: $C_{\mathrm{ie}}=\underset{j}{\overset{m}{\mathrm{\Σ}}}{C}_{\mathrm{ijs}}+\underset{j}{\overset{m}{\mathrm{\Σ}}}\underset{t}{\overset{T}{\mathrm{\Σ}}}{q}_j\·t\·{\mathrm{\γ}}_{\mathrm{ij}}\left(t\right)$

Wherein, the product category that can provide of cooperation manufacturing enterprise is provided m, and t is the period quantity of dividing in the cooperation plan phase, t=1, and 2 ..., T, q _JtBe the specific productivity of t period enterprise for product j, C _IjsBe the beginning production time of the product j among the order i;

(3) set up the cost minimum model:

$\mathrm{Min}\underset{i}{\overset{n}{\mathrm{\Σ}}}\underset{j}{\overset{m}{\mathrm{\Σ}}}\underset{t}{\overset{T}{\mathrm{\Σ}}}{\mathrm{\δ}}_j\·q_j\·{\mathrm{\γ}}_{\mathrm{ij}}\left(t\right)+\underset{i}{\overset{n}{\mathrm{\Σ}}}\underset{j}{\overset{m}{\mathrm{\Σ}}}\underset{t}{\overset{T}{\mathrm{\Σ}}}{\mathrm{\α}}_j\·(d_i-t)\·q_j\·{\mathrm{\γ}}_{\mathrm{ij}}\left(t\right)+\underset{i}{\overset{n}{\mathrm{\Σ}}}\underset{j}{\overset{m}{\mathrm{\Σ}}}\underset{t}{\overset{T}{\mathrm{\Σ}}}{\mathrm{\β}}_j\·(t-d_i)\·q_j\·{\mathrm{\γ}}_{\mathrm{ij}}\left(t\right)$

Wherein, δ _jBe the production cost coefficient of product j, α _jFor the stock of product j deposits cost, β _iThe penalty that departs from delivery date for order i;

(4) above-mentioned two targets are combined into the form of goal programming, then its Optimization Model can be expressed as:

$\mathrm{Min}{W}_R\underset{t}{\mathrm{\Σ}}\underset{i}{\mathrm{\Σ}}(C_{\mathrm{it}}^{+}+C_{\mathrm{it}}^{-})+W_F{J}_F^{-}$

S.t. λ _Ijt≤ q _Jt (ability constraint)

$\underset{i}{\overset{n}{\mathrm{\Σ}}}\underset{j}{\overset{m}{\mathrm{\Σ}}}\underset{t}{\overset{T}{\mathrm{\Σ}}}{\mathrm{\α}}_j\·(d_i-t)\·q_j\·{\mathrm{\γ}}_{\mathrm{ij}}\left(t\right)\≤Q,\∀i,j,t,$

Wherein, W _R, W _FBe respectively the optimization aim 1 of enterprise plan and 2 weight coefficient, be enough big constant, span is: 1 * 10 ⁶～1 * 10 ⁸,

With

Be respectively this enterprise and the extension of order i is departed from the time and depart from the time in advance in the t stage,

Be the amount of being unrealized of manufacturing enterprise's cost, λ _IjtWhen producing for product j among the t period order i to the ability need coefficient of enterprise, the inventory cost that Q can provide for the maximum of enterprise;

(5) model in the 4th step is carried out multiobjective Dynamic Optimization, but a kind of variable weight and self-adapted genetic algorithm are proposed, according to the individual instances in the evolutionary process, automatically adjust Different Optimization objective weight coefficient and cross and variation operator, thereby obtain the global optimization result, obtain the planned dispatching strategy of more excellent or suboptimum with lower cost as far as possible, thus the dynamic production schedule of adjusting inside again.

4, method according to claim 3 is characterized in that: in the described step 5, comprising: 1. the weight coefficient of multiple-objection optimization is handled

Suppose that certain function has q optimization aim:

Min{z ₁＝f ₁(x)，z ₂＝f ₂(x)，...，z _q＝f _q(x)}s.t.gi(x)≤0，i＝1，2，...，m

Be defined as follows two parameters:

$z^{+}=\{z_1^{\max },z_2^{\max },\·\·\·,z_q^{\max }\}$

$z^{-}=\{z_1^{\min },z_2^{\min },\·\·\·,z_q^{\min }\}$

Wherein,

Be respectively the maximal value and the minimum value of k target in the current population; The weight of target k can adopt following formula to find the solution:

${\mathrm{\ω}}_k=\frac{1}{z_k^{\max }-z_k^{\min }},k=\mathrm{1,2},\·\·\·,q$

Corresponding given individuality, above-mentioned two optimization aim can be determined according to following formula:

$z\left(x\right)=\underset{k=1}{\overset{q}{\mathrm{\Σ}}}{\mathrm{\ω}}_k(z_k^{\max }-z_k)$

$=\underset{k=1}{\overset{q}{\mathrm{\Σ}}}\frac{(z_k^{\max }-z_k)}{(z_k^{\max }-z_k^{\min })}$

In the individuality of next round is evolved, new variation all will take place in weight, constantly upgrade, and optimize direction towards forward and advance;

2. Adaptive Genetic operator operation

During evolution, for guaranteeing not to be absorbed in local optimum, introduce fitness rate of change parameters C R weigh up and down two generation population variation:

$\mathrm{CR}=\frac{{\mathrm{fit}}_t^{\max }-{\stackrel{\‾}{\mathrm{fit}}}_t}{{\mathrm{fit}}_t^{\max }-{\mathrm{fit}}_t^{\min }}/\frac{{\mathrm{fit}}_{(t-1)}^{\max }-{\stackrel{\‾}{\mathrm{fit}}}_{(t-1)}}{{\mathrm{fit}}_{(t-1)}^{\max }-{\mathrm{fit}}_{(t-1)}^{\min }}$

Wherein

With

Be respectively t for maximum adaptation degree value, minimum fitness value and average fitness value in the population, CR represent up and down two generation population be illustrated in rate on the fitness value; If CR is excessive, population might enter local optimum, needs to reduce to intersect and mutation operator; If CR is too small, the speed of evolution weakens, and can strengthen in real time to intersect and mutation operator;

The variation of Adaptive Genetic operator can be tried to achieve by following formula:

5. system according to claim 1, it is characterized in that: described Virtual-Hub is a kind of virtual platform based on Web consensus standard and J2EE technology, at the various uncertainties that face in the multiple enterprises coordinated production process, in real time production information share, the production schedule synchronously and dynamically demand such as adjustment and " Virtual Space " that multiple enterprises coordinated production process is synchronous of building, with the shared complicacy of shielding multiple enterprises production information, finish the dynamic monitoring of producing node, reach the coordination purpose, promote the extensive coordinated production between enterprise.

Images