CN100495434C - Bulk goods ship stowage method for iron and steel product - Google Patents
Bulk goods ship stowage method for iron and steel product Download PDFInfo
- Publication number
- CN100495434C CN100495434C CNB2005100283764A CN200510028376A CN100495434C CN 100495434 C CN100495434 C CN 100495434C CN B2005100283764 A CNB2005100283764 A CN B2005100283764A CN 200510028376 A CN200510028376 A CN 200510028376A CN 100495434 C CN100495434 C CN 100495434C
- Authority
- CN
- China
- Prior art keywords
- goods
- cabin
- prestowage
- ship
- genetic algorithm
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Abstract
3 The present invention relates to an iron and steel product bulk cargo ship stowage method. It is characterized by that it utilizes mixed genetic algorithm, and adopts a heuristic method to define concrete arrangement mode of cargo in every cargo hold so as to meet the cargo loading quality requirement. Said genetic algorithm can use heuristic method to calculate stowage index and define the distribution of cargo in every cargo hold so as to meet the capacity limitation of every cargo hold, so that a reasonable cargo stowage scheme can be made up.
Description
Technical field
The present invention relates to shipping boats and ships cargo distribution method, be applicable to the prestowage planning of iron and steel enterprise's finished product, be used to calculate distribution and the concrete arrangement method of products such as volume, plate, determine the prestowage scheme at each cabin of shipping boats and ships in bulk.
Background technology
The general mode that adopts large-scale shipping boats and ships transportation in bulk of dispatching from the factory of steel products, in order to guarantee navigation safety and smooth and easy at each harbour handling goods, consider performance index such as vessel stability, longitudinal strength and trim, the ship will according to the concrete condition of goods and boats and ships determine each hold cargo thing to port and permission weight; From the angle of port side,, the concrete disposing way of goods in cabin there is strict requirement for guaranteeing the freight transportation quality.
The quantity of goods that ship of large-scale iron and steel enterprise is criticized is huge, of a great variety, and the prestowage of goods exist above-mentioned to the port, cabin allow weight and goods put many constraints such as requirement, therefore the prestowage method of finishing the cargo thing is the combinatorial optimization problem of a complexity, belongs to NP-hard problem (nondeterministic polynomial time hard problem).
Mostly by the generally manual by rule of thumb prestowage scheme of formulating of prestowage person, workload is big in the prestowage of present shipping boats and ships transportation in bulk, and work efficiency is lower, and specifically puts order, freight transportation difficult quality guarantee owing to can't consider goods in cabin.
Summary of the invention
The object of the present invention is to provide a kind of bulk goods ship stowage method of iron and steel product, each hold cargo thing of determining according to the ship to the port and allow weight, and different goods put requirement, determine the distribution of goods in each cabin and concrete arrangement method, to guarantee the freight transportation quality, improve shipment efficient.
The core of the inventive method is a Calculation of stowage on board, since the ship specified each hold cargo thing to the port, therefore can be to the goods prestowage respectively at each harbour, after promptly the goods at a selected harbour closes single and corresponding cabin, determine that these close singly in the distribution in each cabin and specifically put in proper order.Because steel products prestowage planning problem belongs to the NP-hard problem,, used genetic algorithm: determine the concrete disposing way of each hold cargo thing to satisfy the cargo loading quality requirements with heuristic for finding the solution this problem; Genetic algorithm is called heuristic, calculates the prestowage index, determines the distribution of goods in each cabin, to satisfy the capacity restriction in each cabin, formulates rational cargo distribution scheme.
Introduce the relevant design of genetic algorithm below earlier, put off until some time later the detailed process of bright Calculation of stowage on board, the design of heuristic is described then.
Genetic algorithm is to find the solution complex combination optimization problem local searching method commonly used, has formed more complete theory, has many tool software directly to use, as the GALib that uses in the method.Genetic algorithm mainly comprises 4 parts: coding, decoding, fitness function and genetic operator.
Coding refers to separating of finding the solution problem is described as the form that genetic algorithm requires, and is about to separating of problem and is compiled into chromosome.
Decoding refers to that the chromosome that will obtain after the evolution of genetic algorithm process several times is converted to separating of problem, and this is the inverse process of coding;
Fitness function is used for describing chromosomal adaptability, the quality of promptly separating;
Genetic operator is used for realizing chromosomal evolution in the iterative process, and the improvement of promptly separating comprises and selects operator, crossover operator and mutation operator etc.
During utilization genetic algorithm for solving practical problems, mainly be characteristics design coding, decoding process and fitness function according to problem, and various genetic operators all have several different methods to realize, and can directly use in tool software, only need to select suitable method according to the characteristics of problem.
The characteristics of genetic algorithm:
We know that traditional optimization method mainly contains three kinds: enumerative technique, heuritic approach and searching algorithm:
(1) enumerative technique enumerates all feasible solutions in the feasible solution set, to obtain accurate optimum solution.For continuous function, this method requires earlier it to be carried out discretize and handles, and so just may not reach optimum solution forever because of discrete processes.In addition, when enumerating the space when bigger, this method to find the solution efficiency ratio lower, can't find the solution sometimes even on present advanced computational tool.
(2) heuritic approach is sought a kind of heuristic rule that can produce feasible solution, to find an optimum solution or approximate optimal solution.This method to find the solution efficiency ratio higher, but must find out its distinctive heuristic rule to the problem that each demand is separated, this heuristic rule does not generally have versatility, is not suitable for other problems.
(3) searching algorithm is sought a kind of searching algorithm, and this algorithm carries out search operation in a subclass of feasible solution set, with optimum solution or the approximate optimal solution that finds problem.Though this method does not guarantee that one obtains the optimum solution of problem surely,, just can on the quality of approximate solution and efficient, reach a kind of balance preferably if suitably utilize some heuristic knowledges.
Genetic algorithm is from representing problem assortment of genes decision.Therefore, be coding work needing to realize at the beginning to genotypic mapping from phenotype.Owing to copy the work of gene code very complicated, we often simplify, as binary coding.After just producing,, produce the approximate solution of becoming better and better by generation (generation) evolution according to the principle of the survival of the fittest and the survival of the fittest for population.In each generation, select (selection) individuality according to fitness (fitness) size individual in the Problem Areas, and make up intersection (crossover) and variation (mutatton) by means of the genetic operator (geneticoperators) of natural genetics, produce the population of the new disaggregation of representative.This process will cause the same back life of evolving naturally of kind of images of a group of characters more to be adapted to environment for population than former generation, and the optimum individual in the last reign of a dynasty population can be used as the problem approximate optimal solution through decoding (decoding).
Genetic algorithm has been adopted natural evolution Model, as selection, intersection, variation, migration, local and neighborhood etc.Fig. 1 has represented the process of basic genetic algorithmic.When calculating beginning, some N individual (father's individuality 1, father's individuality 2, father's individuality 3, father's individuality 4 ...) be population initialization randomly, and calculate each individual fitness function, also promptly in initial generation, just produced the first generation.Do not optimize criterion if do not satisfy, begin to produce the calculating of a new generation.In order to produce the next generation, select individuality according to fitness, parent requires genetic recombination (intersection) and produces filial generation.All filial generations are by certain probability variation.The fitness of filial generation is recomputated again then, and filial generation is inserted in the population parent is replaced, and constitutes a new generation (son individuality 1, son individual 2, sub individual 3, sub individual 4 ...).This process circulation is carried out, till satisfying the optimization criterion.
The single population genetic algorithm of even now is very powerful, can solve problem quite widely well.But adopting on multiple populationsly promptly has the algorithm of sub-population to tend to obtain better result.After each son is planted that images of a group of characters list Population Genetic Algorithm is the same and calculated some generations independently, between sub-population, carry out individuality and exchange.This genetic algorithm on multiple populations is close to the evolution of race in the nature more, is called paralleling genetic algorithm.
Along with the difference of problem kind and the expansion of problem scale, seek a kind of universal method that can solve search with limited cost and optimize, the valid approach that genetic algorithm provides for us just, it is different from traditional search and optimization method.The key distinction is:
1. self-organization, self-adaptation and self-study habit (intelligent).When using the genetic algorithm for solving problem, after encoding scheme, fitness function and genetic operator are determined, algorithm will utilize the information that obtains in the evolutionary process to organize search voluntarily.Because the selection strategy based on nature is " survival of the fittest, incompatibility person is eliminated ", thereby the big individuality of fitness has higher survival probability.Usually, the individuality that fitness is big has the gene structure that more conforms, and again by genetic manipulations such as genetic recombination and gene mutations, just may produce the offspring who more conforms.This self-organization of evolution algorithm, self-adaptive features have it simultaneously and can find the characteristic of environment and the ability of rule automatically according to environmental change.A biggest obstacle in the algorithm design process has been eliminated in natural selection, promptly needs to describe in advance whole characteristics of problem, and the measure that should take at the different characteristics algorithm of problem will be described.Therefore, utilize the method for genetic algorithm, we can solve those complicated unstructured problems.
2. the essential concurrency of genetic algorithm.Genetic algorithm is by population number point of destination of parallel mode search, rather than single-point.Its concurrency shows two aspects, the one, and genetic algorithm is inherent parallel (inherent parallelism), promptly genetic algorithm itself is fit to large-scale parallel very much.The simplest parallel mode is to allow hundreds of even thousands of computing machines carry out the EVOLUTIONARY COMPUTATION of independent population separately, even do not carry out any communication (if independently between the population have a spot of communication generally can bring better result) in the operational process, by the time optimized individual is chosen in the computing comparison of just communicating by letter when finishing.This parallel processing mode does not have what restriction and requirement to parallel system organization, we can say, genetic algorithm is adapted at carrying out parallel processing at present all parallel machines or the distributed system, and parallel efficiency is not had much affect.The 2nd, genetic algorithm include concurrency (implicit parallelismm).Because the mode of genetic algorithms use population is organized search, thereby can search for a plurality of zones in the solution space simultaneously, and mutual exchange of information.Use this way of search, though each only the execution and the proportional calculating of population scale n carried out about O (n in fact
3) inferior efficient search being arranged, this just makes genetic algorithm obtain bigger income with less calculating.
3. genetic algorithm does not need differentiate or other supplementary knowledges, and only need influence the objective function and the corresponding fitness function of the direction of search.
4. genetic algorithm is emphasized the probability transformation rule, rather than the transformation rule of determining.
5. genetic algorithm can more directly be used.
6. genetic algorithm can produce many potential separating to given problem, and final selection can be determined (at some in particular cases, to separate existence as more than one of multi-objective optimization question, one group of pareto optimum solution is arranged by the user.This genetic algorithm is specially suitable for confirming alternative disaggregation).
The basic operation of genetic algorithm
Genetic algorithm comprises three basic operations: select, intersect and variation.These basic operations have many diverse ways again, are introduced one by one below.
1, selects (selection)
Selection is to be used for determining the reorganization or the individuality that intersects, and what filial generation individualities selected individuality will produce.At first calculate fitness:
1. pro rata fitness calculates (proportional fitness assignment);
2. the fitness based on ordering calculates (rank-based fitness asslgranent).
Be actual selection after fitness calculates, carry out the selection of parent individuality according to fitness.
1. roulette is selected (roulette wheel selection);
2. random ergodic sampling (stochastic universal sampling);
3. local select (local selection);
4. block selection (truncation selection);
5. algorithm of tournament selection (tournament selection).
2, intersection or genetic recombination (crossover/recombination)
Genetic recombination is that combination produces new individuality together from the information in the parent mating population, and following algorithm can be arranged:
1. real-valued reorganization (real valued recombination)
● discrete recombination (discrete recombination);
● middle reorganization (intermediate recombination);
● linear reorganization (linear recombination);
● the linear reorganization of expansion (extended linear recombination).
2. scale-of-two intersects (binary valued crossover)
● single-point intersection (single-point crossover);
● multiple spot intersection (multiple-point crossover);
● evenly intersect (uniform crossover);
● the intersection of shuffling (shuffle crossover);
● dwindle agency's intersection (crossover with reduced surrogate).
3. variation (mutation)
The variation of filial generation experience is actually the variation that the filial generation gene produces by the small probability disturbance after intersecting.Difference according to individual coded representation method can have following algorithm:
1. real-valued variation;
2. scale-of-two variation.
Intersect and mutation operation below in conjunction with the binary-coded once roulette selection of simple case expedition, a single-point.
Shown in Figure 2 is the initial population of the individuality composition of one group of scale-of-two gene code formation, and individual fitness evaluation value is as calculated by the numeric representation in the bracket, and big more this individuality of expression of fitness is good more.
Roulette in the similar gaming of roulette system of selection, as shown in Figure 3, the ideal adaptation degree is converted in proportion chooses probability, and wheel disc is divided into 10 sectors, because Yao Jinhang selects for 10 times, so produce 10 [0,1] random number between is equivalent to rotate wheel disc No. 10 times, pointer position when obtaining No. 10 rotating disks and stopping, pointer stops in a certain sector, and the individuality of this sector representative is promptly selected.
Suppose that producing random number sequence is 0.070 221,0.545 929,0.784 567,0.446 93,0.507 893,0.291 198,0.716 34,0.272 901,0.371 435,0.854 641, this random series and the cumulative probability of calculating acquisition are compared, then sequence number is 1 successively, 8,9,6,7,5,8,4,6,10 is individual selected.Obviously the individual selected probability that fitness is high is big, and may be selected; The individuality that fitness is low then probably is eliminated.In the struggle for existence test first time, sequence number is that 2 individuality (0101111001) and 3 individuality (0000000101) are eliminated, and replaces the higher individuality of fitness 8 and 6, and this process is called as regeneration (reproduction).Important genetic manipulation is to intersect after the regeneration, is called hybridization biologically, can be considered as the place why biology is evolved.We with single-point intersect (one-pointcrossover) be example, select arbitrarily through in the population behind the selection operation two individual as intersecting objects, promptly two fathers are individual produces two son individualities through chromosomes exchange reorganization, as shown in Figure 4.Produce position, a point of crossing at random, father individual 1 and father's individuality 2 exchange at the portion gene sign indicating number on the right side of position, point of crossing, form son individual 1 and son individual 2.Finish other individual interlace operations similarly.
Realizing evolutionary mechanism if only consider interlace operation, is not all right as a rule, and this and organic sphere close relative grow and influence the evolution course is similar.Because the number of individuals of population is limited, through some generation interlace operations, because come from the individual phenomenon that is full of whole population gradually of a better ancestors' son, problem meeting premature convergence (premature convergence), certainly, the individuality of Huo Deing can not be represented the optimum solution of problem at last.For avoiding premature convergence, be necessary to add during evolution and have new genetic individuality.One of solution is to follow the nature biotechnology variation.Sudden change has taken place in the gene code that the variation of biological character is actually this proterties of control, and this is very important for keeping bio-diversity.The genetic manipulation of mimic biology variation for the individual population that binary gene code is formed, is realized the small probability upset of gene code, promptly reaches the purpose of variation.
As shown in Figure 5,, determine the 4th gene upset, be about to 1 and be changed to 0 with small probability for individual 1001110100 generation variations.
Generally speaking, the simple evolutionary process of a generation has just comprised selection and regeneration, intersection and the mutation operation based on fitness.
All top population genetic operations are integrated, and the first generation evolutionary process of initial population is shown in Fig. 1 .6.Initial population is through selection operation, fitness higher No. 8 and No. 6 individualities copy 2 respectively, suffer exit for No. 2 and No. 3 that fitness is lower, next selected 4 pairs of father's individualities to finish interlace operation respectively, carry out the single-point intersection in the position of determining at random and generate 4 pairs of individualities by certain probability.Choose certain individual gene code position by small probability at last, produce variation.Just formed the colony of the first generation like this through said process.Later generations of evolutionary process so circulation is gone down, and each is for finishing all to produce new population.The algebraically that develops depends primarily on the convergence state that the representative problem is separated, and optimized individual is separated as the best fit approximation of problem in the last reign of a dynasty population.
The genetic algorithm evolution modelling as shown in Figure 7, individuality develops into optimum individual in the search volume, its propagation probability on high fitness is by increasing progressively from generation to generation, the individual shade of performance is represented the probability distribution of individual propagation among the figure.
The general flow of genetic algorithm is as shown in Figure 8:
The 1st step produced initial population at random, and individual number is certain, and each individuality is expressed as chromosomal gene code;
The 2nd step was calculated individual fitness, and judged whether to meet the optimization criterion, if meet, and the optimum solution of output optimized individual and representative thereof, and finish to calculate; Otherwise turned to for the 3rd step;
The 3rd step selected regeneration individual according to fitness, the individual selected probability height that fitness is high, and the individuality that fitness is low may be eliminated;
The 4th step generated new individuality according to certain crossover probability and cross method;
The 5th step generated new individuality according to certain variation probability and variation method;
The 6th step turned back to for the 2nd step by intersecting and variation generation population of new generation.
Optimization criterion in the genetic algorithm, generally the difference according to problem has different definite modes.For example, can adopt one of following criterion as Rule of judgment:
1. individual maximum adaptation degree surpasses preset value in the population;
2. individual average fitness surpasses preset value in the population;
3. generation number surpasses preset value.
Based on the foregoing invention thinking, technical scheme of the present invention is, bulk goods ship stowage method of iron and steel product, and it comprises the steps:
(1) data are prepared, and determine goods information, cabin information and the prewired information of ship; Goods closes single information and goods managing detailed catalogue etc. to close simple form formula tissue, to comprise, these information leave in the database with the form of tables of data;
(2) select the goods at a harbour to close single and corresponding cabin;
(3) evolutionary generation t=0 is put in genetic algorithm initialization, and maximum evolutionary generation is T, by coding method is singly encoded in selected pass, and generates M chromosome at random, as initial population POP (0);
(4) calculate the prestowage index, to each chromosome, the application decoder method obtains the goods distribution scheme of a correspondence, according to this scheme, to the goods in each cabin, the application goods is put computing method and is determined the concrete disposing way of goods in cabin, and calculates the volume of compartment utilization factor in this cabin, is calculated the prestowage index of goods distribution scheme by the volume of compartment utilization factor in each cabin;
(5) evolve, to the chromosome among the population POP (t), according to its prestowage index, operator, crossover operator and pRepl*M individuality of mutation operator generation are selected in utilization, and (probability parameter of pRepl for determining formed interim population tmpPOP (t); Population POP (t) and interim population tmpPOP (t) are merged, therefrom eliminate poor individuality, make population scale return to M, obtain population of future generation: t=t+1;
(6) the genetic algorithm stop criterion is judged, if t＜T then forwarded for the 4th step to; Otherwise optimum separating among the output POP (t) is as the prestowage scheme of this harbour goods;
(7) if also have the harbour goods not carry out Calculation of stowage on board, then forwarded for the 2nd step to, otherwise stop.
Wherein, whole Calculation of stowage on board process is directly used the computation process of finding the solution the TSP problem in the genetic algorithm, calls goods and put computing method in the objective function of this computation process, calculates the prestowage index.
It is as follows that goods is put computing method:
A. the chromosome of genetic algorithm is decoded, obtain goods and close singly distribution scheme in each cabin;
B. to each cabin, according to single managing detailed catalogue in the pass of this hold cargo thing and cabin information, the utilization heuristic is determined specifically putting in proper order of goods, calculates the volume of compartment utilization factor in this cabin;
C. calculate the prestowage index of goods distribution scheme according to the volume of compartment utilization factor in each cabin, the prestowage index is returned genetic algorithm, carry out next step iteration.
Wherein, close single information and comprise that pass odd numbers, ship lot number, contract are to the port, to port sequence number, name of product, kind numbering, net weight, gross weight, number of packages etc.
The goods managing detailed catalogue comprise close odd numbers, bill of landing number, name of product, kind numbering, contract number, accurately send out number, thickness, width, length, net weight, gross weight, number of packages.
Bill of landing number has provided the information that goods is deposited the warehouse Before Loading, and pass singly comprises can a plurality of bills of lading, can comprise a plurality of contracts again in the bill of lading, and in the goods of same contract, what thickness, width, length, weight were identical is divided into one group, is called a standard and sends out.
Cabin information comprises the boats and ships code, name of vessel, and cabin number, bilge width 1, bilge width 2, highly, length.
The prewired information of ship is that the ship requires the goods stacked at the particular location of each cabin to the port with allow weight, comprises numbering, the ship lot number, and cabin number, to the port sequence number, permission weight, reference position, end position.
1, coding
Because goods is to close simple form formula tissue, after single the unique numbering in given each pass, a full arrangement of these numberings just can be used as a kind of coding of goods distribution scheme, is called chromosome again.For example have 10 to close singly, numbering is followed successively by 1,2 ..., 10, then a kind of goods distribution scheme can be expressed as 2,1,4,5,7,8,10,3,6,9}.The coding form of therefore separating, promptly chromosome is P={P1, P2 ..., Pn}, n is for closing odd number, and gene Pi is 1 to n positive integer, and is unequal mutually, and single numbering is closed in expression, is example with separating above, P1=2 then, P2=1 ..., P10=9.In the evolutionary process of genetic algorithm, in fact the operation that genetic operator carries out chromosome is exactly to change putting in order of coding, thereby produces different distribution schemes.
2, decoding
The coding of separating, promptly chromosome must just can become the distribution scheme of goods by decoding.The quantity of cabin and weight and the error that each cabin allows loading have been provided in ship's information, when decoding according to coding in gene, promptly close putting in order and each cabin permission weight of single numbering, grouping is calculated and is closed single weight sum, just can obtain the corresponding distribution scheme of goods each cabin from the coding of separating.For example, be provided with 3 cabins, allow weight to be respectively 2000,3000,2000 (unit: ton), error is 100.For that group coding above, { 2,1,3 of 4} close single weight sum in the allowed band ([1900,2100]) in the 1st cabin, and these 3 close list and can be placed on 1 cabin if be numbered; Equally, if 5,7, and 8}, { 10,3,6, the weight of these two groups pass lists of 9} in the allowed band in 2,3 cabins, has at this moment just obtained a feasible goods distribution scheme respectively.Also might certain to separate be infeasible, for example, suppose that { 2,1, the weight of 4} is less than 1900, and { 2,1,4, the weight of 5} illustrates then that greater than 2100 the corresponding distribution scheme of this coding is infeasible, and its prestowage index can be made as a great value.
Obtain the distribution scheme of goods pass list by decoding after, goods to each cabin calls heuristic respectively, just can obtain specifically putting and the volume of compartment utilization factor of each hold cargo thing,, then can calculate the prestowage index of this distribution scheme the volume of compartment utilization factor addition in each cabin.
3, fitness fitness (prestowage index).
It is unit that goods in the cabin is put with row, calculates the maximum number of packages that this row can put according to every row's bilge width, and every row's filling rate=actual number of packages/maximum number of packages of putting.For example coil of strip is generally stacked into a row with isosceles triangle in cabin, can put 10 if calculate ground floor according to storehouse bottom width degree and coil diameter, then second and third layer is 9,8, maximum number of packages=10+9+8=27 spare that this row can put, but in the process of putting, 25 have only been put, then filling rate=25/27=0.926 of this row owing to qualitative restrain.Each row's filling rate sum is the volume of compartment utilization factor in this cabin in one cabin, and each cabin space availability ratio sum then is the volume of compartment utilization factor of this prestowage scheme, is used for estimating the quality of this scheme.We define:
The volume of compartment utilization factor of this prestowage scheme of prestowage index=1-of a prestowage scheme
As chromosomal fitness, the prestowage index is more little with the prestowage index, and then the volume of compartment utilization factor is big more under the condition that satisfies the stacking rule, and this distribution scheme is also just excellent more.
Design by above three parts, we have converted the prestowage problem to a circulation traveling salesman problem (TSP problem): coding of separating just is equivalent to a traverse path between n the city, the prestowage index is equivalent to path, ask a shortest path, i.e. the distribution scheme of prestowage index minimum by the order that changes coding.In GALib, there is direct method to use and find the solution the TSP problem.
4, genetic operator
Each genetic operator has accomplished in many ways in GALib, we test through lot of data at the characteristics of prestowage problem, for each genetic operator has been selected suitable method respectively, and explanation respectively below.
(1) selects operator
The inventive method adopts roulette to select operator and stable state genetic algorithm, can guarantee that more excellent individual inheritance to of future generation, can guarantee higher global convergence again.
It is to more excellent individuality that roulette is selected the basic thought of operator (Roulette-wheel Selector), and the distribution scheme that prestowage index just is less is given bigger selection probability, makes it be genetic to the next generation with bigger probability, calculates cumulative probability again.By the even random number that distributes between producing 0～1, come the definite individuality that will select according to the cumulative probability interval that it falls into.Stable state genetic algorithm (steady-state genetic algorithm) is that operator selection individuality intersects, making a variation obtains an interim colony with selecting from original colony, it and original colony are merged, therefrom return to the scale of original colony behind the individuality of superseded difference, as colony of future generation.The size of interim colony is by determined probability parameter pRepl (0＜pRepl＜1) decision, and the scale of establishing original colony is M, and then the scale of interim colony is pRepl
*M.Have certain overlapping (overlapping) between the adjacent two generation colonies of this algorithm, this is the essential condition that genetic algorithm converges arrives globally optimal solution.
(2) crossover operator
For similar TSP problem with character-coded chromosome P, interleaved mode has part mapping to intersect, order is intersected, circulation intersects and limit reorganization etc.This method adopts the limit recombination form.
(3) mutation operator
The mutation operator that this method adopts is: with the order of gene among the less probability P m chiasmatypy P.
(4) stop criterion
Evolution T stops after generation calculating, and exports best separating as the optimum solution of trying to achieve.
The Calculation of stowage on board process is as follows:
The 1st step: data are prepared: comprise single information, cabin information and the prewired information of ship of closing;
The 2nd step: select the goods at a harbour to close single and corresponding cabin;
The 3rd step: the genetic algorithm initialization: put evolutionary generation t=0, maximum evolutionary generation is T, by coding method is singly encoded in selected pass, and generates M chromosome at random, as initial population POP (0);
The 4th step: calculate the prestowage index.To each chromosome, the application decoder method obtains the goods distribution scheme of a correspondence, according to this scheme, to the goods in each cabin, the application goods is put computing method and is determined the concrete disposing way of goods in cabin, and calculate the volume of compartment utilization factor in this cabin, calculate the prestowage index of goods distribution scheme by the volume of compartment utilization factor in each cabin;
The 5th step: evolve: to the chromosome among the POP (t), according to its prestowage index, utilization selects operator, crossover operator and pRepl*M of mutation operator generation individual, forms interim population tmpPOP (t); POP (t) and tmpPOP (t) are merged, therefrom eliminate poor individuality, make population scale return to M, obtain population of future generation; T=t+1;
The 6th step: the genetic algorithm stop criterion is judged.If t＜T then forwarded for the 4th step to; Otherwise optimum separating among the output POP (t) is as the prestowage scheme of this harbour goods;
The 7th step:, then forwarded for the 2nd step to, otherwise stop if also having the harbour goods not carry out Calculation of stowage on board.
The framework of whole Calculation of stowage on board process can directly use the computation process of finding the solution the TSP problem among the GALib as shown in Figure 9, calls goods and put computing method in the objective function of this computation process, calculates the prestowage index.
It is as follows that goods is put computing method:
In the genetic algorithm iterative process, after obtaining closing singly distribution scheme in each cabin, and put quality requirements according to the specifying information of goods, call the computing method of putting of goods in each cabin, to determine specifically putting in proper order of goods, calculate volume of compartment utilization factor and prestowage index.Detailed process is as follows:
The 1st step: the chromosome to genetic algorithm is decoded, and obtains goods and closes singly distribution scheme in each cabin;
The 2nd step: to each cabin, according to single managing detailed catalogue in the pass of this hold cargo thing and cabin information, the utilization heuristic is determined specifically putting in proper order of goods, calculates the volume of compartment utilization factor in this cabin;
The 3rd step: calculate the prestowage index of goods distribution scheme according to the volume of compartment utilization factor in each cabin, the prestowage index is returned genetic algorithm, carry out next step iteration.
Be example still, establish and be numbered that { 2,1,3 passes of 4} singly are placed in No. 1 cabin, and then heuristic closes single goods managing detailed catalogue and cabin information according to these 3, and that determines goods specifically puts order, calculates the volume of compartment utilization factor in this cabin with above prestowage scheme.After the goods in each cabin all calculated, obtain the prestowage index of this goods distribution scheme.In the objective function of genetic algorithm, call heuristic, again the prestowage index is returned genetic algorithm, carry out next step iteration.
Beneficial effect of the present invention
According to the requirement of putting of ship's stowage requirement and different goods, calculate rapidly goods in each cabin distribution and specifically put order, realized the automatic stowage of goods, ensured the freight transportation quality; Determine the delivery order of goods to have improved shipment efficient according to the result of Calculation of stowage on board.
Description of drawings
Fig. 1 is the process synoptic diagram of genetic algorithm;
Fig. 2 is the initial population distribution schematic diagram of genetic algorithm;
Fig. 3 is that the roulette of genetic algorithm is selected synoptic diagram;
Fig. 4 is the single-point intersection synoptic diagram of genetic algorithm;
Fig. 5 is the variation synoptic diagram of genetic algorithm;
Fig. 6 is the evolutionary process synoptic diagram of genetic algorithm;
Fig. 7 is the evolution modelling synoptic diagram of genetic algorithm;
Fig. 8 is the general flow figure of genetic algorithm;
Fig. 9 is the process flow diagram of prestowage process of the present invention;
Figure 10 is the prestowage process flow diagram of one embodiment of the invention coil of strip;
Figure 11 is the prestowage process flow diagram of another embodiment of the present invention steel plate.
Embodiment
The present invention will be further described below by embodiment and accompanying drawing.
Referring to Fig. 9, the process flow diagram of prestowage process of the present invention; Table 1 has provided goods information, and totally 33 close list, and volume and two kinds of plate are arranged, and volume comprises a plurality of kinds again.Provided each in the table and closed single number of packages and weight, the concrete weight of every goods and size have not just been listed one by one because length is limited.Table 4 is ship's information, and the ship provides 3 zones for this shipments, respectively 2,3, and 4 cabins, starting position and end position refer to along the start-stop position of each zone of the longitudinal axis in cabin.
Singly encode to closing, promptly close single numbering to each, shown in number column in the table 1, the full arrangement of any one of numbering is exactly a chromosome.Chromosome is decoded according to the decoding rule, close single distribution scheme with regard to obtaining corresponding goods, for example the goods distribution scheme in the table 5 is exactly by chromosome { 1,4,6,7,8,13,17,24,25,26,29,32,3,9,12,16,19,20,31,2,5,10,11,14,15,18,21,22,23,27,28,30, the 33} decoding obtains.According to the goods distribution scheme, the goods in each cabin is used the computing method of putting of different goods, obtain the filling rate of specifically putting order and every row of goods, the putting in proper order and filling rate of 1 row's goods as shown in table 2.Calculate the volume of compartment utilization factor in each cabin and the prestowage index of goods distribution scheme again.Genetic operator generates one group of new goods distribution scheme according to the prestowage index, and so continuous iteration is up to reaching maximum iteration time, and the prestowage scheme of output prestowage index minimum is shown in table 2 and table 3.Table 5 has provided and has closed single distribution situation and each regional Weight Loaded, taken capacity length; Table 3 provides wherein the concrete arrangement method of row's goods, and this row's goods is positioned at the 3rd row in zone 1, is 24 meters along transverse axis cabin width, and the actual width of putting of goods is 23.1 meters, can put 19+18+17=54 spare, actually puts 51.As can be seen, disposing way is to satisfy to stow rule, as level requirement, top heavy, up big and down small etc.
Table 2 ship information
| Regional number | Cabin number | The starting position | End position | Allow |
| 1 | 2 | .00 | 15.00 | 3800.000 |
| 2 | 3 | 9.00 | 18.00 | 2500.000 |
| 3 | 4 | .00 | 18.00 | 4000.000 |
Table 3 result of calculation-pass is single to distribute
| Regional number | Actual load-carrying | Physical length | The sequence number of the goods of putting |
| 1 | 3759 | 12.44 | 1，4，6，7，8，13，17，24，25，26，29，32 |
| 2 | 2449 | 9.2 | 3，9，12，16，19，20，31 |
| 3 | 3937 | 14.66 | 2，5，10，11，14，15，18，21，22，23，27，28，30，33 |
Because various goods have different characteristics, therefore need design corresponding heuristic for every kind of goods.Explanation respectively below.
Embodiment--volume class assembly method
Coil of strip finished product outward appearance is a right cylinder, is rolled into by sheet steel, and we claim that the cylindrical cross-sectional diameter of coil of strip is a coil diameter, and cylindrical height is the width of coil of strip.
According to the unlike material of steel plate, coil of strip is divided into different kinds, as hot rolled coil, pickling volume, cold rolling coil, roll hard volume, prepainted coil, electrician's volume, zinc-plated volume and stainless-steel roll etc.The width of coil of strip is by user's contract for future delivery decision, and the width of different contract coil of strips may be different.Also by the contract decision, coil of strip is heavy more for the weight of coil of strip, and coil diameter is big more.
When with boats and ships transportation coil of strip, at first the problem that will solve is exactly how coil of strip is put in cabin.The transversal section of cabin is trapezoidal, and the bilge is a rectangle or trapezoidal, and two cabin bilge of general forward andor aft are trapezoidal, and middle cabin is a rectangle.
When loading coil of strip in cabin, coil of strip is fixed with material colligations such as tie, flitches in a row with Chinese character pin-shaped stacking perpendicular to the shipping agency direction, and 1 to 4 layer of every row by the decision of coil of strip kind, keeps certain intervals between row and row.For guaranteeing shipping mass, the requirement certain to being mounted with of coil of strip:
(1) heap high request can be stacked 3 floor heights as hot rolled coil, pickling volume, cold rolling coil, rolls hard volume and can stack 2 floor heights, and hot zinc volume, electric zinc volume etc. can only be stacked 1 floor height;
(2) stack order: when different types of volume is deposited in same row, to stacking order requirement is arranged, can be placed on the hot rolled coil as the pickling volume, cold rolling coil can be placed on pickling and roll up, and prepainted coil can be placed on the cold rolling coil, then do not allow conversely, or the like;
(3) weight requires: the volume that weight is big is placed on lower floor, and the volume that weight is little is placed on the upper strata;
(4) width requirement: the volume that width is big is placed on lower floor, and the volume that width is little is placed on the upper strata.
The ship is before leaning on ship, can be according to voyage and whole ship prestowage balance, provide the cabin number of Hong kong cargo loading and regional extent and each cabin and can distribute tonnage, the area information that this moment, port side will provide according to the ship and the carriage requirement of goods, work out the loading order of goods, i.e. the cargo distribution scheme.Because outlet boats and ships car loading is generally very big, and various in style, specification, weight differ, and it is very difficult to work out a rational prestowage scheme by hand, need design effective algorithm and solve this problem by computing machine.
The prestowage method is referring to Figure 10.
The 1st step: determine the level number and the maximum floor height of the volume of each kind, level number big volume can be placed on number little volume of level, otherwise does not then allow, and is as shown in the table:
Table 4: the level of coil of strip number and maximum floor height
| Kind number | The name of an article | Level number | Maximum floor height |
| 0101 | Hot rolled |
1 | 3 |
| 0102 | The pickling |
2 | 3 |
| 0103 | |
3 | 2 |
| 0104 | Roll |
3 | 2 |
| 0105 | |
4 | 1 |
| 0106 | |
4 | 1 |
| 0107 | |
4 | 1 |
| 0108 | Electrician's |
4 | 1 |
| 0109 | Zinc-plated |
4 | 1 |
| 0110 | Stainless- |
4 | 1 |
The 2nd step: the span of weight, width is set, calculates the weight specification and the width specifications of each volume.So-called span is meant the unit of quantity that prestowage person sets, and is made as 500 kilograms as the weight span, and the width span is made as 100 millimeters.
Weight specification=[w/w span], width specifications=[width/width span] ([] expression rounds).
For example the weight of certain volume is 12665 kilograms, width is 1310 millimeters, and then its weight specification is 25, and width specifications is 13, can ignore the nuance between different volume weight and the width like this, help when guaranteeing loading mass, improving space availability ratio.
The 3rd step: the volume that is placed on same zone is classified and gathered according to level number, kind number, weight specification, width specifications, and these 4 identical volumes of attribute are considered as same group.
The 4th step: the assembling of finishing row's volume is calculated:
The 4.1st step: calculate the width of this row cabin bilge, estimate the number of packages that this row bottom is rolled up according to the average coil diameter of unassembled volume, and according to number definite this row's of the minimum level in the unassembled volume maximum floor height;
The 4.2nd step: the order according to 4,3,2,1 is filled this row from top to bottom with the volume of four levels number respectively.Concrete grammar is: selecting earlier level number is 4 group, the group of same level number selects to fill this row from small to large according to weight specification, width specifications again, up to the maximum floor height that reaches number permission of this level, select the group of next level number again, fill according to identical method, up to the maximum floor height that reaches this row.
Why select this fill method from top to bottom, for following consideration:
● consider earlier to allow the little level number big volume of floor height, be easier to find weight specification, width specifications to be placed on lower floor like this, so that improve space availability ratio than its big volume;
● consider earlier to be easier to impaired level number big volume, cabin in they just preferentially have been deposited in like this reduces impaired possibility;
● first cabin in the twisting in of dress, the fork truck haul distance is bigger, the more close hatch of volume of back dress, the fork truck haul distance is less, therefore adorns lightweight volume earlier, and the big volume of dress weight helps improving fork truck efficient earlier.
The 5th step: if also have knocked-down volume, then returned for the 4th step, otherwise finish and calculating total space utilization factor.
Embodiment--plate class assembly method
Steel plate finished product outward appearance is a cube, piles up colligation by sheet steel and forms, and is divided into different kinds according to the unlike material of steel plate, as hot rolled plate, acid-cleaning plate, cold-reduced sheet, roll hardboard, Coil Coating Products, electrician's plate, tin plate and corrosion resistant plate etc.The width of steel plate and length are considered the convenience of transportation by user's contract for future delivery decision, and the height of each bale packing of same contract is generally consistent.When with boats and ships transportation steel plate, at first the problem that will solve is exactly how steel plate is put in cabin.The transversal section of cabin is trapezoidal, and the bilge is a rectangle or trapezoidal, and two cabin bilge of general forward andor aft are trapezoidal, and middle cabin is a rectangle.
During load steel plates, steel plate is fixed with material colligations such as tie, flitches stack in a row (and length direction is perpendicular to shipping agency direction) perpendicular to shipping agency direction tooled joint in cabin, and 4 to 6 layers of every rows are determined by the steel plate kind, keep certain intervals between row and row.Be to guarantee shipping mass, to the certain requirement of being mounted with of steel plate:
The floor height requirement can be stacked 6 floor heights as hot rolled plate, acid-cleaning plate, cold-reduced sheet, rolls hardboard, hot zine plate, electric zine plate etc. and can stack 4 floor heights;
Weight requires: the plate that weight is big is placed on lower floor, and the plate that weight is little is placed on the upper strata;
Width requirement: the plate that width is big is placed on lower floor, and the plate that width is little is placed on the upper strata.
Because the steel plate that ship is criticized is generally of less types, and a lot of with the number of packages of kind, and the board size of different cultivars differs greatly, and can't pile up mutually, so the plate of each kind generally is independent stacking.
The prestowage method is referring to Figure 11.
The 1st step: determine the level number and the maximum floor height of the plate of each kind, as shown in the table:
Table 5: the level of steel plate number and maximum floor height
| Kind number | The name of an article | Level number | Maximum floor height |
| 0101 | Hot rolled |
1 | 6 |
| 0102 | Acid- |
1 | 6 |
| 0103 | Cold-reduced |
1 | 6 |
| 0104 | |
1 | 4 |
| 0105 | |
1 | 4 |
| 0106 | The |
1 | 4 |
| 0107 | |
1 | 4 |
| 0108 | Electrician's |
1 | 4 |
| 0109 | |
1 | 4 |
| 0110 | Corrosion |
1 | 4 |
The 2nd step: the span of weight, width is set, calculates the weight specification and the width specifications of each plate.So-called span is meant the unit of quantity that prestowage person sets, and is made as 100 kilograms as the weight span, and the width span is made as 100 millimeters.
Weight specification=[w/w span], width specifications=[width/width span] ([] expression rounds).
For example the weight of certain plate is 2051 kilograms, and width is 944 millimeters, and then its weight specification is 20, and width specifications is 9, can ignore the nuance between different plate weight and the width like this, helps improving when guaranteeing loading mass space availability ratio.
The 3rd step: the plate that is placed on same zone is classified and gathered according to kind number, weight specification, width specifications, and these 3 identical plates of attribute are considered as same group.
The 4th step: the assembling of finishing row's plate is calculated:
The 4.1st step: calculate the width of this row cabin bilge, estimate the number of packages of this row bottom plate according to the average length of being unkitted matching board, and according to the maximum floor height that is unkitted minimum level number definite this row in the matching board;
The 4.2nd step: fill this row from top to bottom with the plate of this kind according to weight specification, width specifications order from small to large.
The 5th step: if also have knocked-down plate, then returned for the 4th step, otherwise finish and calculating total space utilization factor.
Claims (6)
1. bulk goods ship stowage method of iron and steel product, it comprises the steps:
(1) data are prepared, and determine goods information, cabin information and the prewired information of ship; Goods closes single information and goods managing detailed catalogue to close simple form formula tissue, to comprise, these information leave in the database with the form of tables of data;
(2) select the goods at a harbour to close single and corresponding cabin;
(3) genetic algorithm initialization: put evolutionary generation t=0, maximum evolutionary generation is T, by coding method is singly encoded in selected pass, and generates M chromosome at random, as initial population POP (0);
(4) calculate the prestowage index: to each chromosome, the application decoder method obtains the goods distribution scheme of a correspondence, according to this scheme, to the goods in each cabin, the application goods is put computing method and is determined the concrete disposing way of goods in cabin, and calculate the volume of compartment utilization factor in this cabin, calculate the prestowage index of goods distribution scheme by the volume of compartment utilization factor in each cabin;
(5) evolve: to the chromosome among the population POP (t), according to its prestowage index, utilization selects operator, crossover operator and mutation operator to generate pRepl*M individuality, and wherein interim population tmpPOP (t) is formed in 0＜pRepl＜1; Population POP (t) and interim population tmpPOP (t) are merged, therefrom eliminate poor individuality, make population scale return to M, obtain population of future generation: t=t+1;
(6) the genetic algorithm stop criterion is judged: if t＜T then forwarded for the 4th step to; Otherwise optimum separating among the output POP (t) is as the prestowage scheme of this harbour goods;
(7) if also have the harbour goods not carry out Calculation of stowage on board, then forwarded for the 2nd step to, otherwise stop; Wherein, whole Calculation of stowage on board process is directly used the computation process of finding the solution the TSP problem in the genetic algorithm, calls goods and put computing method in the objective function of this computation process, calculates the prestowage index, and that described goods is put computing method is as follows:
A. the chromosome of genetic algorithm is decoded, obtain goods and close singly distribution scheme in each cabin;
B. to each cabin, according to single managing detailed catalogue in the pass of this hold cargo thing and cabin information, the utilization heuristic is determined specifically putting in proper order of goods, calculates the volume of compartment utilization factor in this cabin;
C. calculate the prestowage index of goods distribution scheme according to the volume of compartment utilization factor in each cabin, the prestowage index is returned genetic algorithm, carry out next step iteration.
2. bulk goods ship stowage method of iron and steel product as claimed in claim 1 is characterized in that, closes single information and comprises that pass odd numbers, ship lot number, contract are to the port, to port sequence number, name of product, kind numbering, net weight, gross weight, number of packages etc.
3. bulk goods ship stowage method of iron and steel product as claimed in claim 1, it is characterized in that, the goods managing detailed catalogue comprise close odd numbers, bill of landing number, name of product, kind numbering, contract number, accurately send out number, thickness, width, length, net weight, gross weight, number of packages.
4. bulk goods ship stowage method of iron and steel product as claimed in claim 1, it is characterized in that, bill of landing number has provided the information that goods is deposited the warehouse Before Loading, one closes list and can comprise a plurality of bills of lading, can comprise a plurality of contracts again in the bill of lading, in the goods of same contract, what thickness, width, length, weight were identical is divided into one group, is called a standard and sends out.
5. bulk goods ship stowage method of iron and steel product as claimed in claim 1 is characterized in that cabin information comprises the boats and ships code, name of vessel, and cabin number, bilge width 1, bilge width 2, highly, length.
6. bulk goods ship stowage method of iron and steel product as claimed in claim 1, it is characterized in that, the prewired information of ship is that ship's goods of requiring to stack at the particular location of each cabin is to port and permission weight, comprise numbering, ship lot number, cabin number, to the port sequence number, allow weight, reference position, end position.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100283764A CN100495434C (en) | 2005-07-29 | 2005-07-29 | Bulk goods ship stowage method for iron and steel product |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100283764A CN100495434C (en) | 2005-07-29 | 2005-07-29 | Bulk goods ship stowage method for iron and steel product |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1903655A CN1903655A (en) | 2007-01-31 |
| CN100495434C true CN100495434C (en) | 2009-06-03 |
Family
ID=37673054
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2005100283764A Active CN100495434C (en) | 2005-07-29 | 2005-07-29 | Bulk goods ship stowage method for iron and steel product |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100495434C (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI497441B (en) * | 2012-12-21 | 2015-08-21 | Evaluation method of three - dimensional cargo loading and its computer program products | |
| CN103500390B (en) * | 2013-09-18 | 2016-07-06 | 东北大学 | A kind of automatic stowage method improving coiled sheet water transport stability |
| CN107330214B (en) * | 2017-07-11 | 2021-06-01 | 哈尔滨工业大学（威海） | Spatial layout optimization method based on discretization and heuristic evolutionary algorithm |
| CN110027727B (en) * | 2019-04-10 | 2022-05-13 | 上海交通大学 | Aircraft structural strength test optimized installation method based on genetic algorithm |
| CN111859806B (en) * | 2020-07-22 | 2023-10-27 | 大连陆海科技股份有限公司 | Intelligent ship-oriented liquid tank loading method and system |
| CN113780949B (en) * | 2021-09-14 | 2023-09-05 | 福建三钢闽光股份有限公司 | Truck steel plate and steel coil mixed loading method based on order information |
| CN116822755B (en) * | 2023-08-31 | 2024-01-02 | 山东捷瑞数字科技股份有限公司 | Roll-on-roll-off ship vehicle management method, system, device and medium based on digital twin |
-
2005
- 2005-07-29 CN CNB2005100283764A patent/CN100495434C/en active Active
Non-Patent Citations (5)
| Title |
|---|
| . . |
| 优化普零货物拼箱配装的遗传算法. 卜雷，尹传忠，蒲云.交通运输工程学报，第4卷第4期. 2004 |
| 优化普零货物拼箱配装的遗传算法. 卜雷，尹传忠，蒲云.交通运输工程学报，第4卷第4期. 2004 * |
| 遗传算法在货位精确选址中的应用研究. 张英锋，柏l林.系统仿真学报，第16卷第2期. 2004 |
| 遗传算法在货位精确选址中的应用研究. 张英锋，柏l林.系统仿真学报，第16卷第2期. 2004 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1903655A (en) | 2007-01-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100495434C (en) | Bulk goods ship stowage method for iron and steel product | |
| CN107480922B (en) | Method for establishing goods position distribution scheduling model under two-end type same-rail double-vehicle running mode | |
| CN110084545B (en) | Integrated scheduling method of multi-lane automatic stereoscopic warehouse based on mixed integer programming model | |
| CN105836356B (en) | A kind of intensive warehousing system hybrid optimization dispatching method | |
| CN109597304B (en) | Intelligent partitioned storage method for mold library based on artificial bee colony algorithm | |
| CN111178606A (en) | Automatic warehouse storage position allocation optimization method based on NSGA-II | |
| CN101271543A (en) | Production scheduling system and method using genetic algorithm based on elite solution pool | |
| CN103544586A (en) | Cargo allocation method for improving quay crane operation efficiency and vessel stability of containers | |
| CN105976030A (en) | Multi-agent-based platform scheduling intelligent sorting model structure | |
| CN104063778A (en) | Method for allocating cargo positions for cargoes in three-dimensional warehouse | |
| CN108550007A (en) | A kind of slotting optimization method and system of pharmacy corporation automatic stereowarehouse | |
| CN107967586A (en) | A kind of power grid goods and materials storage optimization method | |
| CN103870893A (en) | Optimization method for solving encasement problem under multiple weight restrictions based on three-dimensional space | |
| CN101789093A (en) | Distribution method of container quay berths and shore bridges | |
| CN101587571A (en) | Container terminal yard optimization system | |
| CN105858043B (en) | The warehousing system Optimization Scheduling that a kind of lift is combined with shuttle | |
| CN103246941A (en) | Scheduling method for export container wharf pile-up space | |
| CN114417696A (en) | Automatic stereoscopic warehouse goods space allocation optimization method based on genetic algorithm | |
| CN106773686A (en) | With piler scheduling path model method for building up under the double car operational modes of rail | |
| CN111507644B (en) | Multi-point unloading constrained three-dimensional multi-container loading method | |
| Ayachi et al. | A Genetic algorithm to solve the container storage space allocation problem | |
| CN113570025A (en) | E-commerce storage center goods space distribution method based on discrete particle swarm algorithm | |
| CN110294251A (en) | Box stereo garage warehouse compartment design method in a kind of case of multi-varieties and small-batch article | |
| CN110378644A (en) | A kind of harbour dangerous material container space distribution model and algorithm | |
| Siswanto et al. | Multi-heuristics based genetic algorithm for solving maritime inventory routing problem |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |