CN104834979A - Leader group scheduling method and system - Google Patents
Leader group scheduling method and system Download PDFInfo
- Publication number
- CN104834979A CN104834979A CN201510271972.9A CN201510271972A CN104834979A CN 104834979 A CN104834979 A CN 104834979A CN 201510271972 A CN201510271972 A CN 201510271972A CN 104834979 A CN104834979 A CN 104834979A
- Authority
- CN
- China
- Prior art keywords
- leader
- adaptive value
- population
- group
- individuality
- 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.)
- Granted
Links
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a leader group scheduling method and system; the method comprises the following steps: obtaining group scheduling needs, presetting algorithm parameters and placing the parameters into a buffer memory; obtaining configuration data and placing the data into the buffer memory; initializing the algorithm parameters so as to generate a population; cloning and intersecting individuals; calculating an individual adaptation value, parsing genes in the individual, and removing the gene of low adaptation value from the genes having condition conflicts; selecting an optimal individual from this time iteration, and removing poor individuals so as to form a new population; carrying out disorder processing for the newly formed population; determining whether a termination condition is reached or not, if yes, outputting the optimal individual and continuing, and if no, returning to the cloning step for continue iteration; decoding genes of the optimal individual storing the genes. The leader group scheduling method and system can obtain an ideal group scheduling result with high efficiency.
Description
Technical field
The present invention relates to a kind of leader row group method and system.
Background technology
Along with the quick growth of team travel business, the group's of sending out amount is increasing, and to relate to factor many for leader row clique problem, and service logic is complicated, inefficiency and very easily makeing mistakes, and leader is not optimized utilization.Because leader is not high with team's matching degree, often can occur service quality problem, customer complaint caused thus also gets more and more.It is more and more outstanding that artificial row rolls into a ball the problem caused, and the pattern of artificial row group cannot adapt to the fast development of business at present, and we need the dispatching method of employing science to solve these problems with advanced system.
Traditionally, the combination select permeability of many-one and one-to-many is related in leader row group, namely can there be multiple leader met with a condition in same team, and same leader also can meet the group's of being with condition of multiple team, this has the combination of enormous quantity level with regard to making the possibility of result of group of the row of leader.Existing row's group method will obtain row group effect preferably, then must carry out traversal to each combination to calculate, or be at least carry out in the mode calculated close to traversal, the time of such meeting at substantial and computational resource, and period can calculate the combination of many low values, group of the row of making efficiency reduces greatly.
Existing leader row group method, inefficiency is also very easily made mistakes, and cannot take into account multiple constraint condition and cannot arrange the stark suitable leader for team efficiently.
Summary of the invention
The technical problem to be solved in the present invention is to overcome leader of the prior art row group method inefficiency and very easily makeing mistakes, and multiple constraint condition cannot be taken into account and the defect of the stark suitable leader for team cannot be arranged efficiently, a kind of leader row group method and system are provided.
The present invention solves above-mentioned technical matters by following technical proposals:
The invention provides a kind of leader row group method, its feature is, comprises the following steps:
S
1, obtain row group a demand, preset algorithm parameter also puts into buffer memory, and wherein algorithm parameter comprises population scale, iterations, the maximum operation duration of algorithm, adaptive value condition;
S
2, according to described row group Requirement Acquisition configuration data, and put it into buffer memory, wherein configuration data be meet the group of demand of described row group, group's information, leader, leader's information, country's pin sign duration information and working day information, information of wherein leading a group comprises certificate information and the calendar information of leader;
S
3, initialization algorithm parameter, according to described population scale, generation comprises the population of several body (being also often called as chromosome), each individuality is by multiple unduplicated genomic constitution, wherein gene be row unite fruit minimum unit, to be numbered by a team and leader's numbering forms the correspondence that described minimum unit is team and leader, calculate the adaptive value of each gene according to adaptive value Condition Matching;
S
4, to step S
3individuality in the population generated carries out intersecting after cloning between two and adds in population:
S
5, to step S
4the individuality of middle generation carries out cloning laggard row variation, adds in population by the new individuality produced;
S
6, calculate the individual fitness of each individuality in population, gene in each individuality is resolved to the combination of team and leader, travel through each leader, find out all groups that each leader is corresponding, and the gene (branch hazard here can be such as that two groups have overlap on the stroke date) that in the gene conflicted according to the adaptive value size exclusion existence condition of gene, adaptive value is lower;
S
7, the individuality in population is carried out descending sort according to the size of adaptive value, select the maximum individuality of individual fitness as the optimum individual in current iteration, then according to population scale, eliminate from population and be positioned at the individuality of sequence rear end, to form new population;
S
8, to S
7individuality in the population of middle formation carries out out of order process;
S
9, judge whether to reach end condition, if reach end condition, termination of iterations also exports optimum individual as a result and perform step S
10if do not reach end condition, return S
4;
S
10, each genetic decoding in optimum individual is become corresponding group information and leader's information, and store the information that obtains of decoding.
The design that the present invention have employed genetic algorithm in leader row clique problem is optimized row group mode.From the angle of method flow of the present invention, also never only terminate in the design that make use of genetic algorithm, but the optimization of many aspects has been carried out to flow process and algorithm.One of them is exactly the integrated evaluating method that have employed gene and individual two dimensions.
It should be noted that the part term occurred in present specification, those skilled in the art can understand its implication with reference to the term in common genetic algorithm.Such as, the term " group " occurred in present specification, " team " implication are identical, and can be regarded as the team meeting predefine row group and require, term " gene " represents that row unites the minimum unit of fruit, and namely team is corresponding with what to lead a group with corresponding.Term " individuality " i.e. " chromosome ", is made up of multiple unduplicated genetic unit, and wherein gene number depends on team's number of group of the row of participation.Term " population ", represents by repeatably multiple, the individuality composition that quantity is indefinite in algorithm implementation." gene adaptive value " can be regarded as, and according to the result of the algorithm of the good and bad degree of a series of judgement genes of row group conditional definition, " individual fitness " can be regarded as the result of the algorithm of the individual good and bad degree of a series of judgements according to row group conditional definition.
Individuality in population is being carried out to the stage of crossover and mutation operation, traditional genetic algorithm can directly operate the individuality in current population, and the present invention is (at above-mentioned steps S
4and S
5) then can carrying out intersecting or all cloning the individuality in original seed group before mutation operation, produce the copy of new population, then to the operation that the individuality in copy population intersects or makes a variation, after operation terminates, copy population is incorporated in original seed group, carry out next step operation, finally at the end of an iteration, eliminate the lower individuality of fitness according to population scale again.The benefit done like this is that excellent genes is remained to the next generation by individuality as much as possible that make each have merit, reduce the destruction may caused defect individual gene in the process of intersecting or make a variation, reduce scope or degree that result departs from optimum solution, accelerate the speed of convergence of optimum solution with this, improve efficiency of algorithm.
On the other hand, the present invention is according to some characteristics in leader row group process, devise when being described the fitness of individuality, the abstract of two dimensions and definition are carried out, the i.e. genes of individuals dimension of minimum unit and the dimension of individuality itself, the benefit done like this is convenient to clear and definite and defines the rule of each adaptive value and it is to the coverage of final fitness, and the independent weight passed through each dimension and algorithm design, the impact of the concrete service application rule of statement on net result that can be more accurately three-dimensional, the result that algorithm is produced more meets expection.In following preferred technical scheme, this advantage seems more outstanding.
Preferably, this end condition adopts condition one or condition two, the adaptive value that condition one is the optimum individual reaching iterations or reach in algorithm maximum operation duration or current iteration meets default business demand expectation value, and condition two is for reaching iterations or reaching the maximum operation duration of algorithm.
Easy understand ground, the concrete form of this end condition, can be pre-set according to the actual requirements by those skilled in the art.
Preferably, step S
4in the intersection of carrying out, genes of individuals crossing-over rate is more than or equal to 0.2 and is less than or equal to 0.5.
Preferably, step S
5in the variation carried out, genes of individuals aberration rate is more than or equal to 0.02 and is less than or equal to 0.07.
Preferably, step S
1the population scale of middle setting between 20-200, iterations between 300-1500, the maximum operation duration of algorithm is between 20-100 minute.
More specifically, when population scale is more than or equal to 20 and is less than or equal to 110, iterations optimum range should be more than or equal to 800 and be less than or equal to 1500.When population scale is more than or equal to 110 and is less than or equal to 200, iterations optimum range should be more than or equal to 300 and be less than or equal to 800.Such Selecting parameter is conducive to the good and bad degree taking into account efficiency of algorithm and result.
Preferably, the adaptive value of gene is by formula
calculate, wherein GF is the adaptive value of gene, and GFi is sub-adaptive value, and the numerical range of each sub-adaptive value is between 0-1, and x is the number of sub-adaptive value, and Wi is the weighted value of corresponding sub-adaptive value in the adaptive value of default gene.Obviously, in above-mentioned formula, i is actual is all integers of traversal 1 to x.Wherein, what sub-adaptive value comprised in following 7 kinds is several:
Row group area priorities adaptive value, for characterizing the sequencing priority participating in row group according to predetermined team region, it is identical that the row of the group's correspondence in same team region rolls into a ball area priorities adaptive value;
The leader's group of being with area priorities adaptive value, for characterizing each leader with the priority on a region;
Leader's certificate adaptive value, for certificate of characterizing each leader to a regular adaptedness;
Leader's schedule adaptive value, for schedule of characterizing each leader same day to a regular adaptedness;
Team's grade and leader's ratings match degree adaptive value, the degree of closeness in grade for the group of sign and leader;
Sailing date closes on degree adaptive value in team, for the sailing date of the group of sign and the proximity of current date;
Leader's sex adaptive value, for characterize leader's sex with a regular adaptedness.
The employing of above-mentioned formula, Essential Action played in the present invention adds data smoothing strategy when the adaptive value of COMPREHENSIVE CALCULATING gene dimension, can in the complicated business application scenarios of reality because of business factor occur that suddenly the deviation of or even the order of magnitude comparatively large compared to regime values scope appears in the result of calculation of some or multiple adaptive value time, carry out to a certain degree to data level and smooth, reduce the harmful effect because business scenario fluctuation produces algorithm.
In addition, the caching mechanism of the present invention designed by individual Gene sufficiency, avoids and duplicates computing in an iterative process, accelerate iteration speed, effectively improve the efficiency of whole algorithm.
Present invention also offers a kind of leader row group system, its feature is, comprising:
One pretreatment unit, for the row's of acquisition group's demand and preset algorithm parameter put into buffer memory, wherein algorithm parameter comprises population scale, iterations, the maximum operation duration of algorithm, adaptive value condition;
One configuration data acquiring unit, for rolling into a ball Requirement Acquisition configuration data according to described row, and put it into buffer memory, wherein configuration data be meet the group of demand of described row group, group's information, leader, leader's information, country's pin sign duration information and working day information, information of wherein leading a group comprises certificate information and the calendar information of leader;
One initialization unit, for initialization algorithm parameter, according to described population scale, generation comprises the population of several body (i.e. chromosome), each individuality is by multiple unduplicated genomic constitution, wherein gene be row unite fruit minimum unit, to be numbered by a team and leader's numbering forms the correspondence that described minimum unit is team and leader, calculate the adaptive value of each gene according to adaptive value Condition Matching;
One cross unit, intersects after cloning between two and add in population for the individuality in the population that generates this initialization unit:
One variation unit, the individuality for producing this cross unit clones laggard row variation, is added in population by the new individuality produced, and enables a resolution unit;
This resolution unit, for calculating the individual fitness of each individuality in population, gene in each individuality is resolved to the combination of team and leader, travel through each leader, find out all groups that each leader is corresponding, and the gene that in the gene conflicted according to the adaptive value size exclusion existence condition of gene, adaptive value is lower, then enable a population updating block;
This population recruitment unit, for the individuality in population is carried out descending sort according to the size of adaptive value, select the maximum individuality of individual fitness as the optimum individual in current iteration, then according to population scale, the individuality being positioned at sequence rear end is eliminated, to form new population from population;
One out of order processing unit, carries out out of order process for the individuality in the population that formed this population recruitment unit, then enables a judging unit;
This judging unit judges whether to reach end condition, if reach end condition, termination of iterations also exports optimum individual as a result and enable a decoding unit, if do not reach end condition, again enables this cross unit;
This decoding unit, for each genetic decoding in optimum individual is become corresponding group information and leader's information, and stores the information of decoding and obtaining.
Preferably, this end condition adopts condition one or condition two, the adaptive value that condition one is the optimum individual reaching iterations or reach in algorithm maximum operation duration or current iteration meets default business demand expectation value, and condition two is for reaching iterations or reaching the maximum operation duration of algorithm.
Preferably, in this cross unit intersection of carrying out, genes of individuals crossing-over rate is more than or equal to 0.2 and is less than or equal to 0.5.
Preferably, in the variation carried out of this variation unit, genes of individuals aberration rate is more than or equal to 0.02 and is less than or equal to 0.07.
Preferably, the population scale arranged in this pretreatment unit between 20-200, iterations between 300-1500, the maximum operation duration of algorithm is between 20-100 minute.
More specifically, when population scale is more than or equal to 20 and is less than or equal to 110, iterations optimum range should be more than or equal to 800 and be less than or equal to 1500.When population scale is more than or equal to 110 and is less than or equal to 200, iterations optimum range should be more than or equal to 300 and be less than or equal to 800.Such Selecting parameter is conducive to the good and bad degree taking into account efficiency of algorithm and result.
Preferably, the adaptive value of gene is by formula
calculate, wherein GF is the adaptive value of gene, and GFi is sub-adaptive value, the numerical range of each sub-adaptive value is between 0-1, x is the number of sub-adaptive value, and Wi is the weighted value of corresponding sub-adaptive value in the adaptive value of default gene, and it is several that its neutron adaptive value comprises in following 7 kinds:
Row group area priorities adaptive value, for characterizing the sequencing priority participating in row group according to predetermined team region, it is identical that the row of the group's correspondence in same team region rolls into a ball area priorities adaptive value;
The leader's group of being with area priorities adaptive value, for characterizing each leader with the priority on a region;
Leader's certificate adaptive value, for certificate of characterizing each leader to a regular adaptedness;
Leader's schedule adaptive value, for schedule of characterizing each leader same day to a regular adaptedness;
Team's grade and leader's ratings match degree adaptive value, the degree of closeness in grade for the group of sign and leader;
Sailing date closes on degree adaptive value in team, for the sailing date of the group of sign and the proximity of current date;
Leader's sex adaptive value, for characterize leader's sex with a regular adaptedness.
Those skilled in the art are to be understood that, above-mentioned leader is arranged in the explanation of a system, omit the explanation of to arrange the aspects such as the corresponding technical characteristic of a method, technology contents and the advantage that possesses to above-mentioned leader, but these contents are still and are able to clear understanding by above-mentioned explanation.
On the basis meeting this area general knowledge, above-mentioned each optimum condition, can combination in any, obtains the preferred embodiments of the invention.
Positive progressive effect of the present invention is: the leader in the present invention arranges a method and system, the good results in previous generation can be retained and successive optimization in iteration, effectively can reduce traversal number of times significantly to improve row group efficiency, and more feasible solution can be searched in the short period of time, avoid the interference that a large amount of low value combines, hunting zone also can effectively expand, computation process is covered more comprehensively, the higher complexity that the development that can also adapt to following tourism well brings.
Accompanying drawing explanation
Fig. 1 is the process flow diagram that the leader of the embodiment of the present invention 1 arranges a method.
Fig. 2 is that the leader of the embodiment of the present invention 2 arranges a process flow diagram for method enforcement.
Fig. 3 is the schematic diagram that the leader of the embodiment of the present invention 3 arranges a system.
Embodiment
Mode below by embodiment further illustrates the present invention, but does not therefore limit the present invention among described scope of embodiments.
Embodiment 1
As shown in Figure 1, the leader of the present embodiment arranges a method and comprises the following steps:
S
1, obtain row group a demand, preset algorithm parameter also puts into buffer memory, and wherein algorithm parameter comprises population scale, iterations, the maximum operation duration of algorithm, adaptive value condition;
S
2, according to described row group Requirement Acquisition configuration data, and put it into buffer memory, wherein configuration data be meet the group of demand of described row group, group's information, leader, leader's information, country's pin sign duration information and working day information, information of wherein leading a group comprises certificate information and the calendar information of leader;
S
3, initialization algorithm parameter, according to described population scale, generation comprises the population of several body (being also often called as chromosome), each individuality is by multiple unduplicated genomic constitution, wherein gene be row unite fruit minimum unit, to be numbered by a team and leader's numbering forms the correspondence that described minimum unit is team and leader, calculate the adaptive value of each gene according to adaptive value Condition Matching;
S
4, to step S
3individuality in the population generated carries out intersecting after cloning between two and adds in population:
S
5, to step S
4the individuality of middle generation carries out cloning laggard row variation, adds in population by the new individuality produced;
S
6, calculate the individual fitness of each individuality in population, gene in each individuality is resolved to the combination of team and leader, travel through each leader, find out all groups that each leader is corresponding, and the gene that in the gene conflicted according to the adaptive value size exclusion existence condition of gene, adaptive value is lower;
S
7, the individuality in population is carried out descending sort according to the size of adaptive value, select the maximum individuality of individual fitness as the optimum individual in current iteration, then according to population scale, eliminate from population and be positioned at the individuality of sequence rear end, to form new population;
S
8, to S
7individuality in the population of middle formation carries out out of order process;
S
9, judge whether to reach iterations and whether reach the maximum operation duration of algorithm, if at least one judged result is for being, termination of iterations also exports optimum individual as a result and perform step S
10if two judged results are otherwise return S
4;
S
10, each genetic decoding in optimum individual is become corresponding group information and leader's information, and store the information that obtains of decoding.
Wherein, step S
4in the intersection of carrying out, genes of individuals crossing-over rate is 0.3.Step S
5in the variation carried out, genes of individuals aberration rate is 0.03.Step S
1the population scale of middle setting is 100, iterations is 1000 times, the maximum operation duration of algorithm is 1 hour.
Further, the adaptive value of gene is by formula
calculate, wherein GF is the adaptive value of gene, and GFi is sub-adaptive value, and the numerical range of each sub-adaptive value is between 0-1, and x is the number of sub-adaptive value, and Wi is the weighted value of corresponding sub-adaptive value in the adaptive value of default gene.Obviously, in above-mentioned formula, i is actual is all integers of traversal 1 to x.Wherein, sub-adaptive value comprises following 7 kinds:
1, row group area priorities adaptive value, for characterizing the sequencing priority participating in row group according to predetermined team region, it is identical that the row of the group's correspondence in same team region rolls into a ball area priorities adaptive value;
2, to lead a group the group of being with area priorities adaptive value, for characterizing each leader with the priority on a region;
3, to lead a group certificate adaptive value, for certificate of characterizing each leader to a regular adaptedness;
4, to lead a group schedule adaptive value, for schedule of characterizing each leader same day to a regular adaptedness;
5, team's grade and leader's ratings match degree adaptive value, the degree of closeness in grade for the group of sign and leader;
6, sailing date closes on degree adaptive value in team, for the sailing date of the group of sign and the proximity of current date;
7, to lead a group sex adaptive value, for characterize leader's sex with a regular adaptedness.
The employing of above-mentioned formula, in fact add data smoothing strategy when the adaptive value of COMPREHENSIVE CALCULATING gene dimension, can in the complicated business application scenarios of reality because of business factor occur that suddenly the deviation of or even the order of magnitude comparatively large compared to regime values scope appears in the result of calculation of some or multiple adaptive value time, carry out to a certain degree to data level and smooth, reduce the harmful effect because business scenario fluctuation produces algorithm.
Below lift an object lesson above-mentioned data smoothing strategy is explained.In above-mentioned formula, the value of GFi is all between 0 ~ 1, due to the difference of each conditions and the various of operating environment in the application of reality, the value of adaptive value GFi and the adaptive value of other factors that often there will be some or multiple factor have larger discrepancy on numerical value or the order of magnitude, the result so directly caused is exactly to after the weights W i that the adaptive value application of each factor is corresponding, to the influence power (influence power of the adaptive value of each factor to the final adaptive value calculated is defined as by we: the proportion accounting for net result after the weight that each factor adaptive value application is corresponding) of the result of calculation of final adaptive value, difference is huge.
Although effectively can control the difference of its order of magnitude when defining the adaptive value numerical range of each factor, the bigger difference of numerical value is still difficult to avoid.Such as, suppose the value average out to 0.5 of GF1 under normal conditions, and the value average out to 0.7 of GF2, now according to the weighted value W1=0.6 that this reference design goes out, W2=0.4.So according to general weighting algorithm GF=GF1*W1+GF2*W2=0.5*0.6+0.7*0.4=0.58, so GF1 can be calculated as the influence power of GF: GF1*W1/GF=0.5*0.6/0.58*100%=51.7%; If under the specific condition of certain operating environment the mean value of GF1 reduce to 0.05 and the mean value of GF2 unchanged be still 0.7, so still easily draw GF=0.05*0.6+0.7*0.4=0.31 according to general weighting algorithm; Now GF1 to the influence power of GF is: 0.05*0.6/0.31*100%=9.6%.As can be seen here, the influence power of this situation GF1 to GF decreases an order of magnitude (in other words, reaching more than 5 times).But, we expect that the influence power of adaptive value to net result of each factor of our definition is in different environments that difference is little, because the adaptive value condition of a certain factor is be in peer-level concerning each gene under the condition of a certain operating environment, just differentiated between different operating environmental baseline, therefore should not occur that the influence power of the adaptive value of same factor, with environmental change, great variety occurs, therefore in the present invention, use above-mentioned algorithmic formula to reduce this impact:
Still suppose the value average out to 0.5 of GF1 under normal conditions, and the value average out to 0.7 of GF2, now according to the weighted value W1=0.6 that this reference design goes out, W2=0.4.So easily calculate GF=0.576 according to above-mentioned formula algorithm; Now the influence power of GF1 to GF easily calculates
when the mean value of GF1 under specific condition reduce to 0.05 and the mean value of GF2 unchanged be still 0.7 time, by above-mentioned formulae discovery GF=0.22; Now GF1 to the influence power of GF can calculate into
as can be seen here, when have employed above-mentioned formula, although the mean value of GF1 reduces an order of magnitude, therefore GF1 does not reduce an order of magnitude to the influence power of GF, but remains on peer-level and only numerically slightly reduce.
Embodiment 2
Below for an application example, to illustrate in practical operation, how method of the present invention is implemented according to the demand of reality leader row group or to perform.
The requirement rules that the present embodiment faces is as follows:
Treat that the group of walkthrough adheres to different regions separately, and row group should send leader according to following zone sequence automatically: Taiwan, Europe, America, Canada, Australia, East Africa in other, Japan, Korea S, other.
Each leader has oneself can with the region of group, and there is corresponding priority round values in each region, and the larger priority of numerical value is higher, is 20 to the maximum.
The certificate of leader must meet the condition of regional, and if have an object country of team repeatedly come and go visa, preferentially send.
The schedule of leader must meet wanted condition, namely can not with date with group have and conflict.Lead a group under certain situation had the same day other to arrange but can the group of being with a dating conflict yet, but priority is relatively low.
The brill level (suitable with the team's grade in above-mentioned explanation) of team need meet following rule with the grade of leader:
Bore level descending by team and carry out row group, first arrange 6 and bore, then arrange 5 brills, and priority is the highest;
Optimum Matching degree: A level leader corresponding 6,5 group of boring; B level leader corresponding 5,4 group of boring, C level leader corresponding 4,3 group of boring, D level leader corresponding 3 bore below level.4, during 5,6 group of the row of boring, correspondence first arranges A level leader, is finished rear row B level leader by that analogy (descending); 2, during 3 group of the row of boring, correspondence first arranges D level leader, is finished rear row C level leader by that analogy (ascending order);
Further, successfully sent the team of leader, in certain hour section, sailing date is more forward better, and the group that namely sailing date is nearer preferentially should send leader.When sending leader, the male sex is preferential.
Because the present embodiment relies on the job of timing operation, every day operation once and fix on morning, therefore can run and can not impact other business working time the long period.Further, in the present embodiment, each parameter is defined as follows:
Team number m, determines by meeting group's quantity that row group requires automatically when running.
Leader number n, determines by meeting leader's number that row group requires automatically.
Chromosome length is identical with team number m.
Initial population scale a=60.
Genes of individuals aberration rate Rm=0.02.
Genes of individuals crossing-over rate Rc=0.3.
Algorithm iteration number of times I=1000.
Algorithm runs maximum duration d=1H (hour).
In the present embodiment for each gene adaptive value and weight definition as follows:
1, automatic row group area priorities adaptive value: GF1
The adaptive value in region is defined as 9 grades by priority orders according to demand: Taiwan->1, Europe->0.95, America->0.9, Canada->0.85, Australia->0.8, other Middle East non->0.75, Japan->0.7, Korea S->0.65, other->0.6.
The weights W 1=0.15 corresponding to the significance level setting of service impact according to this condition.
2, lead a group the group's of being with area priorities adaptive value: GF2
(20 are to the maximum according to the numerical value arranged with region of each leader, numerical value is larger, and to represent priority higher) calculate, computing method: GF2=region arranges numerical value/20, if leader L is with in a region do not have team T affiliated area, GF2=0.
The weights W 1=0.2 corresponding to the significance level setting of service impact according to this condition.
3, lead a group certificate adaptive value: GF3
If meet other certificate requirements and leader L have the object country of corresponding team T repeatedly come and go visa time GF3=1, if GF3=0.8 when meeting other certificate requirements but repeatedly come and go visa without object country, if discontented sequitur part requires, GF3=0.
The weights W 1=0.1 corresponding to the significance level setting of service impact according to this condition.
4, lead a group schedule adaptive value: GF4
The L band corresponding group T if lead a group in gene, then in the group of being with and during sending pin label, lead a group and have schedule to conflict then GF4=0, if having other schedules, GF4=0.8 without schedule conflict, if period is without other schedules, and GF4=1.
The weights W 1=0.05 corresponding to the significance level setting of service impact according to this condition.
5, level (i.e. team's grade) and leader's ratings match degree adaptive value: GF5 bore in team
Adaptive value according to priority level initializing corresponding to the brill level (regardless of boring, 2 to bore, 3 to bore, 4 to bore, 5 to bore, 6 to bore) of grade (A, B, C, D) and the team T of the L that leads a group in gene is as shown in table 1 below:
Table 1
| Level is bored by group | Leader's grade | Adaptive value |
| Six bore | A | 0.95 |
| Six bore | B | 0.9 |
| Six bore | C | 0.2 |
| Six bore | D | 0 |
| Five bore | A | 0.9 |
| Five bore | B | 0.85 |
| Five bore | C | 0.25 |
| Five bore | D | 0 |
| Four bore | A | 0.7 |
| Four bore | B | 0.95 |
| Four bore | C | 0.9 |
| Four bore | D | 0 |
| Three bore | A | 0.15 |
| Three bore | B | 0.25 |
| Three bore | C | 0.9 |
| Three bore | D | 0.95 |
| Two bore | A | 0.1 |
| Two bore | B | 0.2 |
| Two bore | C | 0.85 |
| Two bore | D | 0.95 |
| Regardless of brill | A | 0.1 |
| Regardless of brill | B | 0.15 |
| Regardless of brill | C | 0.9 |
| Regardless of brill | D | 0.95 |
The weights W 1=0.3 corresponding to the significance level setting of service impact according to this condition.
6, sailing date closes on degree adaptive value in team: GF6
Whether the sailing date of rolling into a ball T in gene closes on the adaptive value of current time, and it is larger more to close on adaptive value.The minimum sailing date of all team and the interval of maximum sailing date in this row group are divided into 10 intervals, each interval correspondence adaptive value, and time more forward interval adaptive value is larger.Use the interval l belonging to following formulae discovery T:
Group sailing date span d=max (all rolling into a ball sailing date)-min (all rolling into a ball sailing date)
Burst length len: if d%10=0 (namely group's sailing date span is the integral multiple of 10), len=d/10, if d%10 unequal to 0 (namely group's sailing date span is not the integral multiple of 10), len=d/10+1.
Current T and minimum sailing date differ from number of days td=and roll into a ball T Chu and send out Qi – min (all rolling into a ball sailing date)
Grade interval l belonging to current gene: if td/len>0 (namely differed from number of days is more than or equal to an interval span) and td%len=0 (namely differed from number of days is the integral multiple of burst length), l=td/len, otherwise l=td/len+1.
Corresponding team's sailing date proximity adaptive value is as following table 2:
Table 2
| Group's sailing date is interval | Adaptive value |
| 1 | 1 |
| 2 | 0.95 |
| 3 | 0.9 |
| 4 | 0.85 |
| 5 | 0.8 |
| 6 | 0.75 |
| 7 | 0.7 |
| 8 | 0.65 |
| 9 | 0.6 |
| 10 | 0.55 |
The weights W 1=0.15 corresponding to the significance level setting of service impact according to this condition.
7, lead a group sex adaptive value: GF7
The adaptive value of leader's sex, the male sex is the group of row preferentially, the GF7=1 when sex of leading a group is man, the GF7=0.8 when sex of leading a group is female
The weights W 1=0.05 corresponding to the significance level setting of service impact according to this condition.
The all Feasible Basis comprised in individuality are because of total adaptive value: CF1
A leader can be with multiple group, thus to need to get rid of in individuality identical leader in all genes with team in the date have the gene of conflict, then calculate the adaptive value summation of residue gene.
Below with reference to Fig. 2, the implementing procedure leader of the present embodiment being arranged to a method is described.Shown in Figure 2, the detailed process that example of the present invention is implemented is divided into three parts: acquisition algorithm runs necessary data and buffer memory, execution algorithm, analytical algorithm operation result preserving.Concrete implementing procedure is as follows.
1. acquisition algorithm runs necessary data and buffer memory
The 1.1 middle team's requirements participating in row group automatically according to demand, inquire the group and related data buffer memory that satisfy the demands from database.
The 1.2 middle leader's requirements participating in row group automatically according to demand, inquire the leader and related data buffer memory that satisfy the demands from database.
1.3 inquire the relevant certificate information of the leader satisfied the demands and buffer memory from database.
1.4 inquire the calendar information of the leader satisfied the demands and buffer memory from database.
1.5 get country from database sells the configuration information buffer memory of signing duration.
1.6 get workaday configuration information and buffer memory from database, for subsequent calculations working day.
1.7 read the parameter of algorithms and configuration data and put into unique caching.
2. execution algorithm
2.1 initialization algorithm parameters, according to the population scale 60 of configuration, determine generation 60 chromosomes (individuality).According to iterations and the maximum operation duration of algorithm of configuration, end condition is set for reaching iterations or having moved to maximum duration.When producing individual, each gene is made up of team ID and corresponding leader ID, and during the leader L selecting team T corresponding, the adaptive value condition according to design is mated, and the adaptive value that the weight calculation gene of applying gene adaptive value is final, for the adaptive value of subsequent calculations individuality.
Individuality in the initial population produced in 2.2 pairs 2.1 intersects after cloning between two and adds in population:
First combination of two is carried out to the individuality in population, the individuality that grand two the individual generations two of every component Buick are new, ensure not destroy former individuality and excellent genes thereof with this.Selecting length to the crossing-over rate of two individualities 0.3 newly generated is m*0.3, and the random genetic fragment of reference position is carried out cross exchanged and produced two new individualities, and joins in population.
It is newly individual that individuality in the new population produced in 2.3 pairs 2.2 clones rear generation, the genes of individuals aberration rate Rm=0.02 of configuration is used to select number for m*0.02 to the gene in new individuality, leader in the gene of random choose carries out stochastic transformation, the leader of conversion need meet row group and to require and with the right requirement of this group, if the incompetent leader met, be empty, the new individuality produced is added in population.
The individual fitness of each individuality in 2.4 calculating populations:
Gene in each individuality is resolved to the combination of team and leader, travel through each leader, find out in individuality all by this leader with group, and have the gene of conflict according to the date that adaptive value in the gene of the adaptive value size exclusion existence condition conflict of gene is lower.Owing to only having this to the calculating of individual fitness in this example, calculate so last the adaptive value that the total adaptive value remaining gene in individuality is this individuality.
Individuality in population is carried out descending sort according to the size of individual fitness by 2.5, select the maximum individuality of individual fitness and be optimum individual in current iteration, then according to the population scale of configuration, from population, eliminate the individuality after sequence, produce new population.
Individuality in 2.6 pairs of new populations carries out out of order process, so that successive iterations.Intersect between the individuality avoiding adaptive value close, affect diversity.
2.7 carry out end condition judgement: judge whether reach iterations and whether reach operation duration, if meet end condition, termination of iterations also exports optimum individual as solution, if do not reach end condition, continue to perform from 2.2.
3. analytical algorithm operation result preserving
Each genetic decoding in optimum individual to be become corresponding team and leader's data by 3.1, and the data required for information such as record leader schedule.
The 3.2 data write into Databasces that will decode.
The leader of the present embodiment arranges a method and runs complete.
Embodiment 3
Shown in figure 3, the leader of the present embodiment arranges a system, comprises out of order processing unit 8, judging unit 9, decoding unit 0 of pretreatment unit 1, configuration data acquiring unit 2, initialization unit 3, cross unit 4, variation unit 5, resolution unit 6, population updating block 7.
This pretreatment unit is for the row's of acquisition group's demand and preset algorithm parameter put into buffer memory, and wherein algorithm parameter comprises population scale, iterations, the maximum operation duration of algorithm, adaptive value condition.
This configuration data acquiring unit, for rolling into a ball Requirement Acquisition configuration data according to described row, and put it into buffer memory, wherein configuration data be meet the group of demand of described row group, group's information, leader, leader's information, country's pin sign duration information and working day information, information of wherein leading a group comprises certificate information and the calendar information of leader.
This initialization unit, for initialization algorithm parameter, according to described population scale, generate the population comprising several body, each individuality is by multiple unduplicated genomic constitution, wherein gene be row unite fruit minimum unit, to be numbered by a team and leader's numbering forms the correspondence that described minimum unit is team and leader, calculate the adaptive value of each gene according to adaptive value Condition Matching.
This cross unit, intersects after cloning between two and add in population for the individuality in the population that generates this initialization unit.This variation unit, the individuality for producing this cross unit clones laggard row variation, is added in population by the new individuality produced, and enables this resolution unit.
This resolution unit, for calculating the individual fitness of each individuality in population, gene in each individuality is resolved to the combination of team and leader, travel through each leader, find out all groups that each leader is corresponding, and the gene that in the gene conflicted according to the adaptive value size exclusion existence condition of gene, adaptive value is lower, then enable this population recruitment unit.
This population recruitment unit, for the individuality in population is carried out descending sort according to the size of adaptive value, select the maximum individuality of individual fitness as the optimum individual in current iteration, then according to population scale, the individuality being positioned at sequence rear end is eliminated, to form new population from population.
This out of order processing unit, carries out out of order process for the individuality in the population that formed this population recruitment unit, then enables this judging unit.
This judging unit judges whether to reach end condition, if reach end condition, termination of iterations also exports optimum individual as a result and enable a decoding unit, if do not reach end condition, again enables this cross unit.This decoding unit, for each genetic decoding in optimum individual is become corresponding group information and leader's information, and stores the information of decoding and obtaining.
Wherein, this end condition is: the adaptive value of the optimum individual reaching iterations or reach in algorithm maximum operation duration or current iteration meets default business demand expectation value.
Wherein, in this cross unit intersection of carrying out, genes of individuals crossing-over rate is 0.2, genes of individuals aberration rate equals 0.02 in the variation carried out of this variation unit.The population scale arranged in this pretreatment unit is 100, iterations is between 1500, the maximum operation duration of algorithm is 100 minutes.
Further, the adaptive value of gene is by formula
calculate, wherein GF is the adaptive value of gene, and GFi is sub-adaptive value, the numerical range of each sub-adaptive value is between 0-1, x is the number of sub-adaptive value, and Wi is the weighted value of corresponding sub-adaptive value in the adaptive value of default gene, and its neutron adaptive value comprises following 5 kinds:
Row group area priorities adaptive value, for characterizing the sequencing priority participating in row group according to predetermined team region, it is identical that the row of the group's correspondence in same team region rolls into a ball area priorities adaptive value;
The leader's group of being with area priorities adaptive value, for characterizing each leader with the priority on a region;
Leader's schedule adaptive value, for schedule of characterizing each leader same day to a regular adaptedness;
Team's grade and leader's ratings match degree adaptive value, the degree of closeness in grade for the group of sign and leader;
Sailing date closes on degree adaptive value in team, for the sailing date of the group of sign and the proximity of current date.
Although the foregoing describe the specific embodiment of the present invention, it will be understood by those of skill in the art that these only illustrate, protection scope of the present invention is defined by the appended claims.Those skilled in the art, under the prerequisite not deviating from principle of the present invention and essence, can make various changes or modifications to these embodiments, but these change and amendment all falls into protection scope of the present invention.
Claims (10)
1. a leader row group method, is characterized in that, comprise the following steps:
S
1, obtain row group a demand, preset algorithm parameter also puts into buffer memory, and wherein algorithm parameter comprises population scale, iterations, the maximum operation duration of algorithm, adaptive value condition;
S
2, according to described row group Requirement Acquisition configuration data, and put it into buffer memory, wherein configuration data be meet the group of demand of described row group, group's information, leader, leader's information, country's pin sign duration information and working day information, information of wherein leading a group comprises certificate information and the calendar information of leader;
S
3, initialization algorithm parameter, according to described population scale, generate the population comprising several body, each individuality is by multiple unduplicated genomic constitution, wherein gene be row unite fruit minimum unit, to be numbered by a team and leader's numbering forms the correspondence that described minimum unit is team and leader, calculate the adaptive value of each gene according to adaptive value Condition Matching;
S
4, to step S
3individuality in the population generated carries out intersecting after cloning between two and adds in population:
S
5, to step S
4the individuality of middle generation carries out cloning laggard row variation, adds in population by the new individuality produced;
S
6, calculate the individual fitness of each individuality in population, gene in each individuality is resolved to the combination of team and leader, travel through each leader, find out all groups that each leader is corresponding, and the gene that in the gene conflicted according to the adaptive value size exclusion existence condition of gene, adaptive value is lower;
S
7, the individuality in population is carried out descending sort according to the size of adaptive value, select the maximum individuality of individual fitness as the optimum individual in current iteration, then according to population scale, eliminate from population and be positioned at the individuality of sequence rear end, to form new population;
S
8, to S
7individuality in the population of middle formation carries out out of order process;
S
9, judge whether to reach end condition, if reach end condition, termination of iterations also exports optimum individual as a result and perform step S
10if do not reach end condition, return S
4;
S
10, each genetic decoding in optimum individual is become corresponding group information and leader's information, and store the information that obtains of decoding.
2. leader row group method as claimed in claim 1, it is characterized in that, this end condition adopts condition one or condition two, the adaptive value that condition one is the optimum individual reaching iterations or reach in algorithm maximum operation duration or current iteration meets default business demand expectation value, and condition two is for reaching iterations or reaching the maximum operation duration of algorithm.
3. leader row group method as claimed in claim 1, is characterized in that, step S
4in the intersection of carrying out, genes of individuals crossing-over rate is more than or equal to 0.2 and is less than or equal to 0.5, step S
5in the variation carried out, genes of individuals aberration rate is more than or equal to 0.02 and is less than or equal to 0.07.
4. leader row group method as claimed in claim 1, is characterized in that, step S
1the population scale of middle setting between 20-200, iterations between 300-1500, the maximum operation duration of algorithm is between 20-100 minute.
5. as the leader in claim 1-4 as described in any one arranges a method, it is characterized in that, the adaptive value of gene is by formula
calculate, wherein GF is the adaptive value of gene, and GFi is sub-adaptive value, the numerical range of each sub-adaptive value is between 0-1, x is the number of sub-adaptive value, and Wi is the weighted value of corresponding sub-adaptive value in the adaptive value of default gene, and it is several that its neutron adaptive value comprises in following 7 kinds:
Row group area priorities adaptive value, for characterizing the sequencing priority participating in row group according to predetermined team region, it is identical that the row of the group's correspondence in same team region rolls into a ball area priorities adaptive value;
The leader's group of being with area priorities adaptive value, for characterizing each leader with the priority on a region;
Leader's certificate adaptive value, for certificate of characterizing each leader to a regular adaptedness;
Leader's schedule adaptive value, for schedule of characterizing each leader same day to a regular adaptedness;
Team's grade and leader's ratings match degree adaptive value, the degree of closeness in grade for the group of sign and leader;
Sailing date closes on degree adaptive value in team, for the sailing date of the group of sign and the proximity of current date;
Leader's sex adaptive value, for characterize leader's sex with a regular adaptedness.
6. a leader row group system, is characterized in that, comprising:
One pretreatment unit, for the row's of acquisition group's demand and preset algorithm parameter put into buffer memory, wherein algorithm parameter comprises population scale, iterations, the maximum operation duration of algorithm, adaptive value condition;
One configuration data acquiring unit, for rolling into a ball Requirement Acquisition configuration data according to described row, and put it into buffer memory, wherein configuration data be meet the group of demand of described row group, group's information, leader, leader's information, country's pin sign duration information and working day information, information of wherein leading a group comprises certificate information and the calendar information of leader;
One initialization unit, for initialization algorithm parameter, according to described population scale, generate the population comprising several body, each individuality is by multiple unduplicated genomic constitution, wherein gene be row unite fruit minimum unit, to be numbered by a team and leader's numbering forms the correspondence that described minimum unit is team and leader, calculate the adaptive value of each gene according to adaptive value Condition Matching;
One cross unit, intersects after cloning between two and add in population for the individuality in the population that generates this initialization unit:
One variation unit, the individuality for producing this cross unit clones laggard row variation, is added in population by the new individuality produced, and enables a resolution unit;
This resolution unit, for calculating the individual fitness of each individuality in population, gene in each individuality is resolved to the combination of team and leader, travel through each leader, find out all groups that each leader is corresponding, and the gene that in the gene conflicted according to the adaptive value size exclusion existence condition of gene, adaptive value is lower, then enable a population updating block;
This population recruitment unit, for the individuality in population is carried out descending sort according to the size of adaptive value, select the maximum individuality of individual fitness as the optimum individual in current iteration, then according to population scale, the individuality being positioned at sequence rear end is eliminated, to form new population from population;
One out of order processing unit, carries out out of order process for the individuality in the population that formed this population recruitment unit, then enables a judging unit;
This judging unit judges whether to reach end condition, if reach end condition, termination of iterations also exports optimum individual as a result and enable a decoding unit, if do not reach end condition, again enables this cross unit;
This decoding unit, for each genetic decoding in optimum individual is become corresponding group information and leader's information, and stores the information of decoding and obtaining.
7. leader row group as claimed in claim 6 system, it is characterized in that, this end condition adopts condition one or condition two, the adaptive value that condition one is the optimum individual reaching iterations or reach in algorithm maximum operation duration or current iteration meets default business demand expectation value, and condition two is for reaching iterations or reaching the maximum operation duration of algorithm.
8. leader row group as claimed in claim 6 system, it is characterized in that, in the intersection that this cross unit carries out, genes of individuals crossing-over rate is more than or equal to 0.2 and is less than or equal to 0.5, and in the variation that this variation unit carries out, genes of individuals aberration rate is more than or equal to 0.02 and is less than or equal to 0.07.
9. a leader row group as claimed in claim 6 system, is characterized in that, the population scale arranged in this pretreatment unit between 20-200, iterations between 300-1500, the maximum operation duration of algorithm is between 20-100 minute.
10. as the leader in claim 6-9 as described in any one arranges a system, it is characterized in that, the adaptive value of gene is by formula
calculate, wherein GF is the adaptive value of gene, and GFi is sub-adaptive value, the numerical range of each sub-adaptive value is between 0-1, x is the number of sub-adaptive value, and Wi is the weighted value of corresponding sub-adaptive value in the adaptive value of default gene, and it is several that its neutron adaptive value comprises in following 7 kinds:
Row group area priorities adaptive value, for characterizing the sequencing priority participating in row group according to predetermined team region, it is identical that the row of the group's correspondence in same team region rolls into a ball area priorities adaptive value;
The leader's group of being with area priorities adaptive value, for characterizing each leader with the priority on a region;
Leader's certificate adaptive value, for certificate of characterizing each leader to a regular adaptedness;
Leader's schedule adaptive value, for schedule of characterizing each leader same day to a regular adaptedness;
Team's grade and leader's ratings match degree adaptive value, the degree of closeness in grade for the group of sign and leader;
Sailing date closes on degree adaptive value in team, for the sailing date of the group of sign and the proximity of current date;
Leader's sex adaptive value, for characterize leader's sex with a regular adaptedness.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510271972.9A CN104834979B (en) | 2015-05-25 | 2015-05-25 | Leader row group method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510271972.9A CN104834979B (en) | 2015-05-25 | 2015-05-25 | Leader row group method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104834979A true CN104834979A (en) | 2015-08-12 |
| CN104834979B CN104834979B (en) | 2018-07-20 |
Family
ID=53812853
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510271972.9A Active CN104834979B (en) | 2015-05-25 | 2015-05-25 | Leader row group method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104834979B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101256648A (en) * | 2008-04-09 | 2008-09-03 | 永凯软件技术（上海）有限公司 | Genetic operation operator based on indent structure for producing quening system |
| CN102054211A (en) * | 2009-10-28 | 2011-05-11 | 北京交通大学 | Police deployment method and system |
| CN103177292A (en) * | 2011-12-21 | 2013-06-26 | 重庆金美通信有限责任公司 | Metro crew working scheduling algorithm based on hybrid genetic algorithm |
-
2015
- 2015-05-25 CN CN201510271972.9A patent/CN104834979B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101256648A (en) * | 2008-04-09 | 2008-09-03 | 永凯软件技术（上海）有限公司 | Genetic operation operator based on indent structure for producing quening system |
| CN102054211A (en) * | 2009-10-28 | 2011-05-11 | 北京交通大学 | Police deployment method and system |
| CN103177292A (en) * | 2011-12-21 | 2013-06-26 | 重庆金美通信有限责任公司 | Metro crew working scheduling algorithm based on hybrid genetic algorithm |
Non-Patent Citations (3)
| Title |
|---|
| 彭小峰: "基于遗传算法的排课问题研究及其应用", 《中国优秀硕士学位论文全文数据库（信息科技辑）》 * |
| 赵光哲: "大学排课问题中的遗传算法设计", 《延边大学学报(自然科学版)》 * |
| 魏云飞: "遗传算法在人员排班问题上的应用研究", 《中国优秀硕士学位论文全文数据库（信息科技辑）》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104834979B (en) | 2018-07-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104636871B (en) | A kind of control method of the single phase multi-product batch processing based on data | |
| CN104268722A (en) | Dynamic flexible job-shop scheduling method based on multi-objective evolutionary algorithm | |
| CN111079987A (en) | Semiconductor workshop production scheduling method based on genetic algorithm | |
| Zhang et al. | Multi-objective scheduling simulation of flexible job-shop based on multi-population genetic algorithm | |
| CN112882449B (en) | Multi-variety small-batch multi-target flexible job shop energy consumption optimization scheduling method | |
| CN102323952A (en) | Reconfigurable assembly line sequencing method based on improved genetic algorithm | |
| CN111080408A (en) | Order information processing method based on deep reinforcement learning | |
| CN113341889B (en) | Distributed blocking flow workshop scheduling method and system with assembly stage and energy consumption | |
| CN108053152A (en) | The method that improved adaptive GA-IAGA based on polychromatic sets solves dynamic job shop scheduling | |
| CN110471274A (en) | Based on the machine components process line dispatching method for improving unified particle swarm algorithm | |
| CN112884590A (en) | Power grid enterprise financing decision method based on machine learning algorithm | |
| CN104077634A (en) | Active-reactive type dynamic project scheduling method based on multi-objective optimization | |
| CN104199820A (en) | Cloud platform MapReduce workflow scheduling optimizing method | |
| Cao et al. | Simulation optimization for a digital twin using a multi-fidelity framework | |
| CN113705978B (en) | Static and dynamic integrated decision-making method and system for multi-machine task cutter | |
| CN104281917A (en) | Fuzzy job-shop scheduling method based on self-adaption inheritance and clonal selection algorithm | |
| CN110458478A (en) | Job-shop scheduling method based on discrete invasive weed algorithm | |
| CN105427058A (en) | Engine test wire station distribution optimization method based on genetic algorithm | |
| Ilich et al. | Modeling of employment structural transformations | |
| CN113485278B (en) | Flexible job shop scheduling multi-target distribution estimation method for optimizing two production indexes | |
| Wieczorek et al. | (r, Q) inventory management in complex distribution systems of the One Belt One Road initiative | |
| CN104834979A (en) | Leader group scheduling method and system | |
| CN117196462A (en) | Supplier screening method, equipment and medium based on demand order | |
| CN110705844A (en) | Robust optimization method of job shop scheduling scheme based on non-forced idle time | |
| Auzina-Emsina et al. | Competitiveness and economic development scenarios of Latvia |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C41 | Transfer of patent application or patent right or utility model | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20160203 Address after: 200335 Shanghai city Changning District Admiralty Road No. 968 Building No. 16 10 floor Applicant after: SHANGHAI XIECHENG BUSINESS CO., LTD. Address before: 200335 Shanghai City, Changning District Fuquan Road No. 99, Ctrip network technology building Applicant before: Ctrip computer technology (Shanghai) Co., Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |