CN101079121A - Metallurgical industry integrative plan scheduling system and method - Google Patents

Abstract

The invention discloses an integrated project program system and the method in the metallurgical industry, which comprises the following steps: the first step: dividing the original integrated project optimization problem into a plurality of son problems; the second step: establishing the linear programming model and solving a son problem; the third step: deducting the energy and the material resource occupied by the son problem in the second step and rebuilding the linear programming model and solving the next son problem until all the problems are solved; the fourth step: getting the integrated project result by synthesizing every son problem result. The invention enhances the program integration.

Description

Metallurgical industry integrative plan scheduling system and method

Technical field

The present invention relates to planning scheduling system and method, especially a kind of metallurgical industry integrative plan scheduling system and method.

Background technology

Overall planning is based on advanced requirement and productive capacity and the senior production schedule that provides, can be used for instructing each department to carry out equilibrium of stock and order is replied.The overall planning system need consider unit production capacity, material supply, product processing route, safety inventory, maximum stock restriction, and other constraint condition, and produce one and comprise and comprise the production schedule satisfying on the basis of these constraint conditions, resource planning, inventory planning, and income is planned the overall planning in interior optimization.

Because in overall planning, the object of processing all disperses, under the more prerequisite of technological procedure constraint, realize the multiple-objection optimization more complicated.The method of existing integrative plan scheduling system, overall planning optimization problem with reality, comprise production logic, constraint condition and optimization aim etc., be mapped as continuous linear programming model, ask the optimum solution of linear programming model then, to separate at last and be converted into the production schedule, resource planning, inventory planning, cost budgeting and income plan or the like.

In the finding the solution of linear programming model, mature methods is simplicial method and interior point method at present, and the complexity of algorithm is polynomial expression level (O (n at least ^x), x＞3), along with the model scale increase time of finding the solution increases rapidly.In the prior art based on the various business softwares of these algorithms develop comparative maturity, CPLEX of ILOG etc. for example.But even best algorithm and software, for finding the solution of very large linear programming model, its used time still is difficult to satisfactory.For the huge supply chain of metallurgy industry model, can not satisfy the requirement of integration of production and marketing operations in the prior art, there is following defective: 1, the plan of needs and the production schedule disconnect.Existing integrative plan scheduling system and method conclusion plan belong to the pattern of basing sales on production, and can't come the demand of quick customer in response; 2, each produces the calculated bottleneck information of line and all changes every month, causes recovery prediction inaccurate, and the production schedule can not reach real balance, and existing integrative plan scheduling system can not be considered this point; 3, the logistics demand amount can not be clear and definite between the last next procedure of existing program system, influenced the balance of capacity planning; 4, calculated hourly output is obtained according to historical experience, and it is inaccurate to divide level, causes overall planning to do the production schedule and resource planning on thinner levels such as the specification of product, the trade mark; 5, each season, the moon and week plan all are to be formulated by different people, and the parameter of every kind of plan is provided with inconsistently, and interdependence is not tight between several plans; 6, the plan unit is disunity in each plan and system, and each plan is disconnected.The problems referred to above are difficult to obtain the result of a global optimization.

Summary of the invention

Technical matters to be solved by this invention provides a kind of metallurgical industry integrative plan scheduling system and method, can make the plan based on all unfilled orders and demand forecast accurately in future, the target that realization is fixed output quota by sales, to improve the speed to respond to the market of Baosteel, strengthen competitive power, promote production procedure in calculated embodiment, and can promote the integrated of plan.

For solving the problems of the technologies described above, the technical scheme of a kind of metallurgical industry integrative plan scheduling system of the present invention and method is, may further comprise the steps, and the first step is divided into several subproblems with original overall planning optimization problem; In second step, set up linear programming model and find the solution a subproblem; In the 3rd step, resources such as the deduction production capacity that subproblem takies in second step, material rebulid linear programming model and find the solution next subproblem, up to solving all subproblems; In the 4th step, comprehensively separating of each subproblem obtains the overall planning result.

The present invention is decomposed into objects such as material, operation, unit, demand by the Supply Chain Model with Baosteel, each subproblem is found the solution, and comprehensively these constraints and target, provide optimum plan, can not only make the plan based on all unfilled orders and demand forecast accurately in future, realization and plan of needs be synchronous, production procedure can also embodied in the works, promoted the integrated of plan.

Description of drawings

Below in conjunction with drawings and Examples the present invention is further described:

Fig. 1 is a kind of metallurgical industry integrative plan scheduling system of the present invention and method flow synoptic diagram;

Fig. 2 is an embodiment of the invention resolution problem synoptic diagram;

Fig. 3 is a product-feed chain model synoptic diagram in the embodiment of the invention;

Fig. 4 is a production capacity balance model synoptic diagram in the embodiment of the invention.

Embodiment

As shown in Figure 1, the technical scheme of a kind of metallurgical industry integrative plan scheduling system of the present invention and method is, may further comprise the steps, and the first step is divided into several subproblems with original overall planning optimization problem; In second step, set up linear programming model and find the solution a subproblem; In the 3rd step, resources such as the deduction production capacity that subproblem takies in second step, material rebulid linear programming model and find the solution next subproblem, up to solving all subproblems; In the 4th step, comprehensively separating of each subproblem obtains the overall planning result.

Moon flowsheet synthesis plan with Baosteel is an example.At first, resolution problem is divided into several subproblems with original overall planning optimization problem.The time span of the moon flowsheet synthesis plan of Baosteel is three months, and all kinds of demands that relate to reach several ten thousand, and requiring to provide with the sky is the planned outcome of chronomere.Wherein demand can be divided in goods, order and prediction, and forward more priority is high more.Satisfied in this month at goods and general requirement of order, prediction is divided into the second month again and trimestral prediction, satisfies in the moon at place respectively.In the utilization of production capacity, require the high forward production capacity of demand holding time of priority, the production capacity after the demand that priority is low takies and leans on.Therefore, priority and take the time period partition problem of production capacity according to demand, as shown in Figure 2, production capacity with two weeks satisfies at goods as problem one, with first month production capacity satisfy order as problem two, with the second month prediction as problem three, the trimestral predictions as problem four.With the several problems of the obscure one-tenth of overall planning, one by one each subproblem is found the solution then.

Secondly, set up linear programming model and find the solution a subproblem." continuous linear programming optimization is applied to the supply chain of network flow pattern, and this subproblem model as shown in Figure 3 in the model employing of problem.The logistics of supply chain of each product of flowing through at first is arranged, and is used for describing the production operation method of product, and makes necessary raw material, turnout, resource and the inventory in two operations in each operation.The supply chain of product A and product B is illustrated on the top of figure.The supply and demand of mating various products can be described by the network flow chart of throwing the net.The flow through a network chart of a product has been described in the lower part of figure.With the flow through a network model conversation is linear programming model, and the various constraints and the target that are about to the flow through a network model are converted into linear equation.

At first, satisfy the material node balance.For any material node b, its upstream is a production process, and the downstream is to consume operation.On the node t, material the leave over amount of b when t finishes equals influx and deducts discharge at any time, and influx is b in the amount of leaving over+production process of (t-1) time quantum of output at t; And discharge is the amount that is used for satisfying the demands (if on b demand)+consume the expend amount of operation at t.

The amount that satisfies the demands (if on b demand) again+consume the expend amount of operation at t), being write as linear equation is:

$\mathrm{Inventory}(b,t)=\mathrm{Inventory}(b,t-1)+\underset{O\&Element;\mathrm{produceop}}{\mathrm{\&Sigma;}}\underset{\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="A20061002682400102.png"\; state="\{\{state\}\}">t</figure-callout>}1}{\mathrm{\&Sigma;}}\mathrm{OpProduceBuffer}(o,b,t1)$

$-\underset{d\&Element;\mathrm{Demand}}{\mathrm{\&Sigma;}}\mathrm{DemandSat}(d,b,t)-\underset{O\&Element;\mathrm{consumeop}}{\mathrm{\&Sigma;}}\underset{\mathrm{<figure-callout\; id="2"\; label="t"\; filenames="A20061002682400102.png"\; state="\{\{state\}\}">t</figure-callout>}2}{\mathrm{\&Sigma;}}\mathrm{OpConsumeBuffer}(o,b,t1)$

Wherein t1 is all can produce the operation of b in timing node t start-up time, and t2 is that all can be in the start-up time of the operation of timing node t internal consumption b.

Secondly, ask the production capacity balance.Production capacity generally is regularly to replenish, and for any production capacity node r, its consumption on t can not be less than its magnitude of recruitment.The model of this problem can be released material nodal equilibrium equation and production capacity nodal equilibrium equation by Fig. 4 as shown in Figure 4.Corresponding linear equation is:

$\mathrm{ResourceCapacity}(r,t)\&GreaterEqual;\underset{O\&Element;\mathrm{consumeRes}}{\mathrm{\&Sigma;}}\underset{\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="A20061002682400102.png"\; state="\{\{state\}\}">t</figure-callout>}1}{\mathrm{\&Sigma;}}\mathrm{OpConsumeRes}(o,\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="A20061002682400102.png"\; state="\{\{state\}\}">t</figure-callout>}1,r)$

Wherein t1 is that all can be in the start-up time of the operation of timing node t internal consumption r

The optimization aim correspondence of integrative plan scheduling system and the objective function of linear programming model, main target has:

(1) maximum reaches the plan of meeting customer need the most timely

$\mathrm{Max}:\underset{d\&Element;\mathrm{DEMAND}}{\mathrm{\&Sigma;}}\mathrm{DemandSatisfed}(d,...)$

(2) the shortest average time order production cycle

At first there is constraint condition:

$\&ForAll;d,\mathrm{DemandSatisfied}(d,b,\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="A20061002682400102.png"\; state="\{\{state\}\}">t</figure-callout>}1,t2)=\mathrm{Inventory}(b,t1)+\underset{\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="A20061002682400102.png"\; state="\{\{state\}\}">t</figure-callout>}1<t\&le;\mathrm{<figure-callout\; id="2"\; label="t"\; filenames="A20061002682400102.png"\; state="\{\{state\}\}">t</figure-callout>}2}{\mathrm{\&Sigma;}}\mathrm{Delay}(d,t)$

Wherein, t1 is the intersection of ideals phase of demand d; T2 is that the maximum of demand d allows the friendship phase.

D satisfy by t1 satisfy and the expiration foot that drags from (t1+1) to t2 every day is formed, the shortest average order production cycle target be equivalent to dragging of as far as possible demaning reduction expire sufficient, promptly

$\mathrm{Min}:\underset{d\&Element;\mathrm{DEMAND}}{\mathrm{\&Sigma;}}\underset{t}{\mathrm{\&Sigma;}}\mathrm{Selay}(d,t)$

(3) meet customer need by client's priority level

$\mathrm{Max}:\underset{d\&Element;\mathrm{DEMAND}}{\mathrm{\&Sigma;}}\mathrm{DemandSatisfed}(d,...)\&times;\mathrm{rank}\left(d\right),$

Wherein, rank (d) is the client priority level (PRI) of order d

(4) the only product line of decision product and minimizing intersection production path

If before certain operation crossedpath is arranged, then there is constraint:

$\&ForAll;o\&Element;\mathrm{operation},t\&Element;\mathrm{bucket\; OpPlan}(o,t)=\mathrm{PrimaryFlow}(o,t)+\underset{t}{\mathrm{\&Sigma;}}\mathrm{AlternateFlow}(o,f,t)$

Wherein, OpPlan is the amount of operation at t; PrimaryFlow is the amount of predominating path contribution; AlternateFlowe is the amount that less important crossedpath is contributed.

Target is

$\mathrm{Min}:\underset{o}{\mathrm{\&Sigma;}}\underset{f}{\mathrm{\&Sigma;}}\underset{t}{\mathrm{\&Sigma;}}\mathrm{AlternateFlow}(o,f,t)$

(5) under maximum and the most timely prerequisite of meeting customer need, be reduced in goods and warehouse for finished product storage and reduce tank farm stock, must reduce unnecessary production as far as possible, the target of this moment is:

$\mathrm{Min}:\underset{o}{\mathrm{\&Sigma;}}\underset{t}{\mathrm{\&Sigma;}}\mathrm{OpPlan}(o,t)$

(6) the maximum overall benefit is that income subtracts cost and is:

$\mathrm{Max}:\underset{d\&Element;\mathrm{DEMAND}}{\mathrm{\&Sigma;}}\mathrm{DemandSatisfed}(d,...)\&times;(\mathrm{Incomde}\left(d\right)-\mathrm{Cost}\left(d\right))$

The overall planning system considers multiple-objection optimization, and in progressively the finding the solution of each linear programming problem, single target according to priority successively is considered, and is guaranteeing to go to ask preferably separating of current goal under the optimum solution situation of previous target.

In the 3rd step, resources such as the deduction production capacity that subproblem takies in second step, material rebulid linear programming model and find the solution next subproblem, up to solving all subproblems.Finish in case progressively find the solution, scheme is produced, and target will be used in during the next one progressively finds the solution, and can be maintained from the last good target of optimization that obtains progressively finding the solution simultaneously.At last, finish all corresponding optimal schemes that just obtained to reach the business goal of all expectations after progressively finding the solution.

Claims (4)

1. metallurgical industry integrative plan scheduling system and method is characterized in that may further comprise the steps, the first step is divided into several subproblems with original overall planning optimization problem; In second step, set up linear programming model and find the solution a subproblem; In the 3rd step, resources such as the deduction production capacity that subproblem takies in second step, material rebulid linear programming model and find the solution next subproblem, up to solving all subproblems; In the 4th step, comprehensively separating of each subproblem obtains the overall planning result.

2. metallurgical industry integrative plan scheduling system according to claim 1 and method is characterized in that, the subproblem in described second step and the 3rd step comprises objects such as material, operation, unit, demand.

3. metallurgical industry integrative plan scheduling system according to claim 1 and method, it is characterized in that the Supply Chain Model that adopts continuous linear programming optimization to be applied to the network flow pattern in described second step and the 3rd step is mapped as linear programming model with subproblem.

4. metallurgical industry integrative plan scheduling system according to claim 1 and method is characterized in that, set up linear programming model in described second step and the 3rd step and find the solution subproblem employing Software tool CPLEX and find the solution.

Images