CN108537338A - Disaster assistance emergency resources dispatching method based on multi-Agent Genetic Algorithm - Google Patents

Disaster assistance emergency resources dispatching method based on multi-Agent Genetic Algorithm Download PDF

Info

Publication number
CN108537338A
CN108537338A CN201810321384.5A CN201810321384A CN108537338A CN 108537338 A CN108537338 A CN 108537338A CN 201810321384 A CN201810321384 A CN 201810321384A CN 108537338 A CN108537338 A CN 108537338A
Authority
CN
China
Prior art keywords
intelligent body
emergency
energy
agent
intelligent
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
Application number
CN201810321384.5A
Other languages
Chinese (zh)
Other versions
CN108537338B (en
Inventor
刘静
秦永伟
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xidian University
Original Assignee
Xidian University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Xidian University filed Critical Xidian University
Priority to CN201810321384.5A priority Critical patent/CN108537338B/en
Publication of CN108537338A publication Critical patent/CN108537338A/en
Application granted granted Critical
Publication of CN108537338B publication Critical patent/CN108537338B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/12—Computing arrangements based on biological models using genetic models
    • G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G06Q10/06312—Adjustment or analysis of established resource schedule, e.g. resource or task levelling, or dynamic rescheduling

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • Human Resources & Organizations (AREA)
  • Biophysics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Theoretical Computer Science (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Biology (AREA)
  • Strategic Management (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Economics (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Biomedical Technology (AREA)
  • Development Economics (AREA)
  • Molecular Biology (AREA)
  • Educational Administration (AREA)
  • Genetics & Genomics (AREA)
  • Game Theory and Decision Science (AREA)
  • Physiology (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a kind of disaster assistance emergency resources dispatching method based on multi-Agent Genetic Algorithm mainly solves the problems, such as that prior art emergency resources scheduling freight charges are high and rescue delay time length.Its scheme is：1) emergency resources dispatch network is built, and as an intelligent body, Agent Grid is built into 25 intelligent bodies, the energy of each intelligent body in computational intelligence volume mesh；2) maximum iteration T is set, neighborhood competition, neighborhood intersection, mutation operation is carried out successively to Agent Grid, obtains local optimum intelligent body；3) self study operation is carried out to local optimum intelligent body；4) judge whether cyclic algebra reaches maximum iteration, if so, the optimal scheduling scheme of output emergency resources network, otherwise, by T plus 1, return to step 3).The freight generated present invention reduces emergency scheduling and total delay time improve the promptness and high efficiency of emergency resources scheduling, can be used for the safe handling to natural calamity.

Description

Disaster assistance emergency resources dispatching method based on multi-Agent Genetic Algorithm
Technical field
The invention belongs to field of computer technology, more particularly to a kind of disaster assistance emergency resources dispatching method can be used for Safe handling to natural calamity.
Background technology
In recent years, various frequent natural calamities occur, and production and life to the people bring serious influence, also give state Family brings huge property loss.With the promotion of scientific and technological level, the ability of the human knowledge world and reforming world is continuous to be obtained To reinforcement, it is significantly improved to the pre-alerting ability of the natural calamity of certain scale.But some accidents are coped with, as Shake, flood etc., there is also certain defects.It is fragile to the comparison of planning strategies for of emergency resources when handling these accidents, standard Change degree is not high, and scheduling structure is unreasonable.Existing emergency resources Operation Measures are it is difficult to ensure that the promptness and high efficiency rescued. China is the influence for coping with various natural calamities and corresponding accident, has been formulated《National Emergent Public Events are totally answered Anxious prediction scheme》, to ensure the smooth development of emergency management and rescue, it is a degree of full to ensure that the basic living of the disaster area masses can obtain Foot.But facing to complicated situation and many uncertain factors, how to ensure that relief goods are reasonably dispatched, to mitigate people The life and property loss of the people masses ensures the promptness and high efficiency of rescue, is still the important topic that we study.
Paper " the Emergency resources scheduling based that Zhang Liming et al. are delivered at it on adaptively mutate genetic algorithm”(《Computer in Human Behavior》Article is numbered： 1493-1498 (2011)) in disclose a kind of emergency resources dispatching method based on Adaptive Mutation Genetic Algorithms Based.This method is logical It crosses and creates n × m matrix at random to represent a kind of emergency materials scheduling scheme, one is selected from population using wheel disc bet method A better solution completes individual crossover operation between population, using y-bend space cut tree come real using matrix crossover operator Existing TSP question operation retains the maximum individual of fitness value in population, stops until meeting end condition by successive ignition Only.The shortcoming of this method is that emergency resources are dispatched scale and increased, and speed of searching optimization is slow, and convergence rate is slow, is easily trapped into part It is optimal.
A kind of patent " scheduling of resource optimization based on genetic algorithm of the emergent Science and Technology Ltd. of the glad latitude in Hubei in its application Method " (application number：201410154950.X application publication number：103927584 A of CN) in disclose and a kind of calculated based on heredity The scheduling of resource optimization method of method.This approach includes the following steps：1. generating initial population S at random：To the various types of moneys of N in S Source individual is encoded, and all resource individuals form a population；2. according to preset fitness function, calculate each in population The fitness value of individual；3. carrying out the selection of individual, intersection, mutation operation；4. judging whether current individual meets termination item Part, if so, decoding obtains optimal resource scheduling scheme, if it is not, the progeny population return to step 2 then generated based on individual into The processing of row next iteration.The shortcoming of this method is that calculating speed is slow, local optimum is easily trapped into, to emergency resources It dispatches not prompt enough efficient.
Invention content
It is an object of the invention to the deficiency for above-mentioned prior art, provide a kind of based on multi-Agent Genetic Algorithm Disaster assistance emergency resources dispatching method prevents from being absorbed in local optimum to improve algorithm speed, makes the scheduling of emergency resources more In time efficiently.
Technical scheme of the present invention includes as follows to achieve the above object：
(1) emergency resources dispatch network is built：
Three Emergency Goods Supply point labels of setting are respectively S1, S2, S3；
Three emergency materials demand point labels of setting are respectively D1, D2, D3；
Three Emergency Goods Supply point supply emergency resources capability flags of setting are respectively SQ1, SQ2, SQ3；
The emergency resources demand label of three emergency materials demand points of setting is respectively DQ1, DQ2, DQ3；
The object time label that configuration scheduling unit emergency materials reach three demand points is respectively t1, t2, t3；
It sets from Emergency Goods Supply point SiTo emergency materials demand point DjThe emergency materials amount of allocation and transportation is labeled as cij；
It sets the time value from i-th of supply centre traffic unit emergency materials to j-th of demand point and is labeled as tij, wherein i It is the integer from 1 to 3 with j；
Three Emergency Goods Supplys o'clock provide required emergency resources, each emergency materials to three emergency materials demand points The goods and materials amount that supply centre is provided to emergency materials demand point is supplied no more than the deliverability of itself, all Emergency Goods Supply points The goods and materials amount answered should meet the demand of each emergency materials demand point, and thread emergency resources to emergency materials demand point are produced The raw required time, thus form emergency resources dispatch network；
(2) using an emergency resources dispatch network as an intelligent body, it is 5 × 5 to be built into size with 25 intelligent bodies Agent Grid, i.e., in Agent Grid, often row has 5 intelligent bodies, often shows 5 intelligent bodies, and according to fitness function, meter Calculate the ENERGY E of each intelligent body in the Agent Grid；
(3) it is based on multi-Agent Genetic Algorithm, the Agent Grid for being 5 × 5 to size carries out neighborhood competition, neighborhood successively Intersect, mutation operation, obtains the local optimum intelligent body in 5 × 5 grids；
(4) self study operation is carried out to local optimum intelligent body：
(4a) builds the Agent Grid that a size is 3 × 3, i.e. intelligent body with local optimum intelligent body in 5 × 5 grids Often row is there are three intelligent body in grid, and there are three intelligent bodies for each column；
(4b) calculates the energy e of each intelligent body in the Agent Grid that size is 3 × 3 according to fitness function；
The Agent Grid that (4c) is 3 × 3 to size carries out neighborhood competition, mutation operation successively, obtains in 3 × 3 grids Local optimum intelligent body；
The ENERGY E of 5 × 5 local optimum intelligent bodies is compared by (4d) with the energy e of 3 × 3 local optimum intelligent bodies, if When E is less than e, then updates 5 × 5 local optimum intelligent bodies with 3 × 3 local optimum intelligent bodies and otherwise do not update；
(4e) is using updated 5 × 5 local optimum intelligent body as optimal emergency resources dispatch network；
It is t=20 that the iterations that self study operates, which are arranged, in (4f), judge current self study operation cyclic algebra whether Reach the iterations of setting, if so, thening follow the steps (5), otherwise, after the cyclic algebra that self study operates is added 1, returns to step Suddenly (4c)；
(5) maximum iteration of setting multi-Agent Genetic Algorithm is T=1000；Judge that current multiple agent heredity is calculated Whether the cyclic algebra of method reaches maximum iteration, if so, the optimal scheduling scheme of output emergency resources network, otherwise, After the cyclic algebra of multi-Agent Genetic Algorithm is added 1, return to step (3).
The present invention has the following advantages that compared with prior art：
First, since the present invention uses Agent Grid as initialization population, reduce the iterations of searching process, It has been quickly found out optimal emergency resources scheduling scheme, has overcome that conventional method convergence rate in the prior art is slow, iterations More disadvantage.So that the present invention accelerates the convergence rate of optimization emergency resources scheduling, shortens and carry out emergency resources scheduling The spent time.
Second, since the present invention executes neighborhood contention operation, neighborhood crossover operation, mutation operation, contracting to Agent Grid The small search space of emergency resources dispatch network optimizing, accelerates the searching process of emergency resources dispatch network, overcomes existing Have the shortcomings that the conventional method search space in technology is big, computationally intensive.So that the present invention reduces search space, greatly reduce The calculation amount of searching process.
Third is adapted to large-scale emergency resources tune since the present invention performs self study operation to Agent Grid Degree network optimization problem is easily trapped into local optimum when overcoming conventional method in the prior art to solving the problems, such as extensive, The too slow disadvantage of speed.When so that the present invention solving large-scale emergency resources dispatch network optimization problem, can quickly it search out Optimal solution prevents from being absorbed in local optimum.
Description of the drawings
The implementation flow chart of Fig. 1 present invention；
Fig. 2 emergency resources dispatch network schematic diagrames.
Specific implementation mode
The present invention will be further described below in conjunction with the accompanying drawings：
Step 1, emergency resources dispatch network is built.
(1a) setting network marks and parameter：
Three Emergency Goods Supply point labels of setting are respectively S1, S2, S3；
Three emergency materials demand point labels of setting are respectively D1, D2, D3；
Three Emergency Goods Supply point supply emergency resources capability flags of setting are respectively SQ1, SQ2, SQ3；
The emergency resources demand label of three emergency materials demand points of setting is respectively DQ1, DQ2, DQ3；
The object time label that configuration scheduling unit emergency materials reach three emergency materials demand points is respectively t1, t2, t3；
It sets from Emergency Goods Supply point SiTo emergency materials demand point DjThe emergency materials amount of allocation and transportation is labeled as cij；
It sets the time value from i-th of supply centre traffic unit emergency materials to j-th of demand point and is labeled as tij, wherein i It is the integer from 1 to 3 with j.
(1b) sets three Emergency Goods Supply point S1, S2, S3To three emergency materials demand point D1, D2, D3Needed for providing Emergency resources, each Emergency Goods Supply point SiTo emergency materials demand point DjThe goods and materials amount of offer is no more than the supply of itself Ability SQi, all Emergency Goods Supply point S1, S2, S3The goods and materials amount of supply should meet each emergency materials demand point D1, D2, D3 Demand DQj, from Emergency Goods Supply point SiTo emergency materials demand point DjTime needed for thread emergency materials is tij, from three Emergency Goods Supply point S1, S2, S3To three emergency materials demand point D1, D2, D3The mesh of thread emergency materials It is respectively t to mark time value1, t2, t3, emergency resources dispatch network is consequently formed, as shown in Figure 2.
Step 2, in computational intelligence volume mesh each intelligent body ENERGY E.
The emergency resources dispatch network that (2a) builds step 1 is built into greatly as an intelligent body with 25 intelligent bodies Small is 5 × 5 Agent Grids, i.e., in Agent Grid, often row has 5 intelligent bodies, often shows 5 intelligent bodies；
(2b) calculates the ENERGY E of each intelligent body in 5 × 5 Agent Grids：
(2b1) calculates the target function value f of each intelligent body in 5 × 5 Agent Grids by following objective function Equation (a)：
Wherein, the intelligent body in the Agent Grid that a is 5 × 5；S is search space, it represents all emergency resources networks Schedule sequences；
(2b2) obtains the ENERGY E of each intelligent body in 5 × 5 Agent Grid by following fitness function formula：
Wherein, Energy (a) is fitness function, i.e., the ENERGY E of each intelligent body is equal in 5 × 5 Agent Grid The fitness of its own.
Step 3, be based on multi-Agent Genetic Algorithm, to size be 5 × 5 Agent Grid carry out successively neighborhood competition, Neighborhood intersection, mutation operation, obtain the local optimum intelligent body in 5 × 5 grids.
(3a) carries out neighborhood contention operation：
(3a1) optional intelligent body finds out energy in four neighborhoods from four neighborhoods up and down of selected intelligent body Measure maximum intelligent body；
The energy of energy maximum intelligent body in neighborhood is compared by (3a2) with the energy of selected intelligent body, if in neighborhood When the energy of energy maximum intelligent body is more than the energy of selected intelligent body, then energy maximum intelligent body in neighborhood is used to substitute selected intelligence Energy body, obtains updated intelligent body, otherwise, without substituting, retains in selected intelligent body survival to Agent Grid；
(3b) carries out neighborhood crossover operation：
In (3b1) first Agent Grid after completing neighborhood competition, an optional intelligent body, from the upper of selected intelligent body In the neighborhood of lower left and right four, the maximum intelligent body of energy in four neighborhoods is found out；
(3b2) randomly generates two different crosspoint point1 and point2, and the size in the two crosspoints is less than The length of agent encoding；
(3b3) hands over the gene between two crosspoints of the maximum intelligent body of energy in selected intelligent body and its neighborhood It changes；
(3c) carries out mutation operation, i.e., in the Agent Grid after completing neighborhood intersection, an optional intelligent body, by one A random perturbation for meeting Gaussian Profile is added on the intelligent body, changes the genic value on the intelligent body with this, after obtaining variation New intelligent body；
(3d) calculates energy possessed by each intelligent body in 5 × 5 Agent Grids, by comparing the energy of each intelligent body Amount determines there is the intelligent body of highest energy, local optimum intelligent body in as 5 × 5 Agent Grids.
Step 4, self study operation is carried out to local optimum intelligent body.
(4a) builds the Agent Grid that a size is 3 × 3, the i.e. intelligence with local optimum intelligent body in 5 × 5 grids Often row is there are three intelligent body in volume mesh, and there are three intelligent bodies for each column；
(4b) calculates the energy e of each intelligent body in the Agent Grid that size is 3 × 3：
(4b1) calculates the object function of each intelligent body in 3 × 3 Agent Grid by following objective function Equation Value f (a)：
Wherein, the intelligent body in the Agent Grid that a is 3 × 3；S is search space, it represents all emergency resources networks Schedule sequences；
(4b2) calculates the energy e of each intelligent body in 3 × 3 Agent Grid by following fitness function formula.
Wherein, Energy (a) is fitness function, and the energy e of each intelligent body is equal to it in 3 × 3 Agent Grid The fitness of itself.
The Agent Grid that (4c) is 3 × 3 to size carries out neighborhood competition, mutation operation successively, obtains in 3 × 3 grids Local optimum intelligent body；
(4c1) carries out neighborhood contention operation：
(4c11) optional intelligent body is found out from four neighborhoods up and down of selected intelligent body in four neighborhoods The maximum intelligent body of energy；
The energy of energy maximum intelligent body in neighborhood is compared by (4c12) with the energy of selected intelligent body, if in neighborhood When the energy of energy maximum intelligent body is more than the energy of selected intelligent body, then energy maximum intelligent body in neighborhood is used to substitute selected intelligence Energy body, obtains updated intelligent body, otherwise, without substituting, retains in selected intelligent body survival to Agent Grid；
(4c2) carries out mutation operation, i.e., in the Agent Grid after completing neighborhood intersection, an optional intelligent body will One random perturbation for meeting Gaussian Profile is added on the intelligent body, is changed the genic value on the intelligent body with this, is made a variation New intelligent body afterwards；
(4c3) calculates energy possessed by each intelligent body in 3 × 3 Agent Grids, by comparing each intelligent body Energy determines there is the intelligent body of highest energy, local optimum intelligent body in as 3 × 3 Agent Grids；
The ENERGY E of 5 × 5 local optimum intelligent bodies is compared by (4d) with the energy e of 3 × 3 local optimum intelligent bodies, if When E is less than e, then updates 5 × 5 local optimum intelligent bodies with 3 × 3 local optimum intelligent bodies and otherwise do not update；
(4e) is using updated 5 × 5 local optimum intelligent body as optimal emergency resources dispatch network；
It is t=20 that the iterations that self study operates, which are arranged, in (4f), judge current self study operation cyclic algebra whether Reach the iterations of setting, if so, thening follow the steps (5), otherwise, after the cyclic algebra that self study operates is added 1, returns to step Suddenly (4c).
Step 5, the maximum iteration that multi-Agent Genetic Algorithm is arranged is T=1000；Judge that current multiple agent is lost Whether the cyclic algebra of propagation algorithm reaches maximum iteration, if so, the optimal scheduling scheme of output emergency resources network, no Then, after the cyclic algebra of multi-Agent Genetic Algorithm being added 1, return to step 3.
The effect of invention can be further described by following simulation result.
1. simulated conditions：
The processor that the emulation experiment of the present invention selects is Intel (R) Core (TM) i5-2450M CPU@ 2.50GHz2.50GHz inside saves as 4G, hard disk 500G, and operating system is Microsoft windows7, and programmed environment is Visual studio 2010。
Data set such as table 1, table 2, the table 3 that policy of the present invention uses, wherein：
Under table 1 indicates no-delay, from Emergency Goods Supply point SiUnit emergency materials are allocated and transported to emergency materials demand point Dj's Spend cij；
Table 2 is the test data set about supply centre and demand point；
Table 3 is shown from Emergency Goods Supply point SiUnit emergency materials are allocated and transported to emergency materials demand point DjWhen required Between tij。
Thread emergency materials reach the object time t of three emergency materials demand points1, t2, t33h is respectively set to, 4h,5h。
Under table 1 is no-delay, the cost of unit emergency materials is allocated and transported
Test data set of the table 2 about supply centre and demand point
The 3 traffic unit emergency materials required time of table
2. experiment content and interpretation of result：
5 groups of data in table 3 are optimized respectively with the present invention, are calculated after every group of data simulation 25 times optimal Under total freight and optimal emergency resources scheduling scheme under emergency resources scheduling scheme, optimal emergency resources scheduling scheme Total delay time, and be compared with existing close mother algorithm MA and Genetic Algorithms, the results are shown in Table 4.
4 simulation result list of table
From table 4, it can be seen that with calculated optimal emergency resources scheduling scheme after the present invention every group of emulation data 25 times Under total freight it is smaller, illustrate the present invention when optimizing emergency resources scheduling, efficient, convergence rate, can than very fast With quickly find optimal solution.
From table 4, it can be seen that with calculated optimal emergency resources scheduling scheme after the present invention every group of emulation data 25 times Under total delay time it is smaller, illustrate the present invention optimize emergency resources scheduling when, speed is fast, optimization performance it is good, more increase Effect.
From table 4 it can also be seen that the disaster assistance emergency materials proposed by the present invention based on multi-Agent Genetic Algorithm Dispatching method is effective, to be substantially better than traditional close mother algorithm MA and Genetic Algorithms, illustrates the method that the present invention designs It can be derived that preferable emergency resources scheduling scheme, and more quickly and efficiently.

Claims (5)

1. a kind of disaster assistance emergency resources dispatching method based on multi-Agent Genetic Algorithm, which is characterized in that including：
(1) emergency resources dispatch network is built：
Three Emergency Goods Supply point labels of setting are respectively S1, S2, S3；
Three emergency materials demand point labels of setting are respectively D1, D2, D3；
Three Emergency Goods Supply point supply emergency resources capability flags of setting are respectively SQ1, SQ2, SQ3；
The emergency resources demand label of three emergency materials demand points of setting is respectively DQ1, DQ2, DQ3；
The object time label that configuration scheduling unit emergency materials reach three demand points is respectively t1, t2, t3；
It sets from Emergency Goods Supply point SiTo emergency materials demand point DjThe emergency materials amount of allocation and transportation is labeled as cij；
It sets the time value from i-th of supply centre traffic unit emergency materials to j-th of demand point and is labeled as tij, wherein i and j are equal It is the integer from 1 to 3；
Three Emergency Goods Supplys o'clock provide required emergency resources, each Emergency Goods Supply to three emergency materials demand points The goods and materials amount that is there is provided to emergency materials demand point of point no more than the deliverability of itself, what all Emergency Goods Supply points were supplied Goods and materials amount should meet the demand of each emergency materials demand point, and thread emergency resources to emergency materials demand point generate institute The time needed, thus form emergency resources dispatch network；
(2) using an emergency resources dispatch network as an intelligent body, it is 5 × 5 intelligence to be built into size with 25 intelligent bodies Volume mesh, i.e., in Agent Grid, often row has 5 intelligent bodies, often shows 5 intelligent bodies, and according to fitness function, calculating should The ENERGY E of each intelligent body in Agent Grid；
(3) it is based on multi-Agent Genetic Algorithm, the Agent Grid for being 5 × 5 to size carries out neighborhood competition, Neighborhood Intersection successively Fork, mutation operation, obtain the local optimum intelligent body in 5 × 5 grids；
(4) self study operation is carried out to local optimum intelligent body：
(4a) builds the Agent Grid that a size is 3 × 3, i.e. Agent Grid with local optimum intelligent body in 5 × 5 grids In often row is there are three intelligent body, there are three intelligent bodies for each column；
(4b) calculates the energy e of each intelligent body in the Agent Grid that size is 3 × 3 according to fitness function；
The Agent Grid that (4c) is 3 × 3 to size carries out neighborhood competition, mutation operation successively, obtains part in 3 × 3 grids Optimal intelligent body；
The ENERGY E of 5 × 5 local optimum intelligent bodies is compared by (4d) with the energy e of 3 × 3 local optimum intelligent bodies, if E is small When e, then updates 5 × 5 local optimum intelligent bodies with 3 × 3 local optimum intelligent bodies and otherwise do not update；
(4e) is using updated 5 × 5 local optimum intelligent body as optimal emergency resources dispatch network；
It is t=20 that the iterations that self study operates, which are arranged, in (4f), judges whether the cyclic algebra of current self study operation reaches The iterations of setting, if so, (5) are thened follow the steps, otherwise, after the cyclic algebra that self study operates is added 1, return to step (4c)；
(5) maximum iteration of setting multi-Agent Genetic Algorithm is T=1000；Judge current multi-Agent Genetic Algorithm Whether cyclic algebra reaches maximum iteration, if so, the optimal scheduling scheme of output emergency resources network otherwise will be more After the cyclic algebra of Agent Genetic Algorithm adds 1, return to step (3).
2. according to the method described in right 1, which is characterized in that the neighborhood contention operation in step (3) is realized as follows：
First, an optional intelligent body finds out in four neighborhoods energy most from four neighborhoods up and down of selected intelligent body Big intelligent body；
Then, the energy of energy maximum intelligent body in neighborhood is compared with the energy of selected intelligent body, if energy in neighborhood When the energy of maximum intelligent body is more than the energy of selected intelligent body, then energy maximum intelligent body in neighborhood is used to substitute selected intelligence Body obtains updated intelligent body, otherwise, then without substituting, retains in selected intelligent body survival to Agent Grid.
3. according to the method described in right 1, which is characterized in that neighborhood crossover operation in step (3) is first from completion neighborhood competition In Agent Grid afterwards, an optional intelligent body finds out four neighborhoods from four neighborhoods up and down of selected intelligent body The middle maximum intelligent body of energy；Randomly generate two different crosspoint point1 and point2 again, the two crosspoints it is big The small length for being less than agent encoding；Then by two crosspoints of the maximum intelligent body of energy in selected intelligent body and its neighborhood Between gene swap.
4. according to the method described in right 1, which is characterized in that the mutation operation of step (3) is the intelligence after completing neighborhood intersection In energy volume mesh, an optional intelligent body, the random perturbation for one being met Gaussian Profile is added on the intelligent body, is changed with this Genic value on the intelligent body, the new intelligent body after being made a variation.
5. according to the method described in right 1, which is characterized in that calculate each intelligence in 5 × 5 Agent Grid in step (2) The ENERGY E of energy body, carries out as follows：
First, by following objective function Equation, the target function value f of each intelligent body in 5 × 5 Agent Grid is calculated (a)：
Wherein, the intelligent body in the Agent Grid that a is 5 × 5, S is search space, represents all emergency resources network scheduling sequences Row；
Then, by following fitness function formula, the ENERGY E of each intelligent body is calculated；
Wherein, Energy (a) is fitness function, i.e., the ENERGY E of each intelligent body is equal to the fitness of its own.
CN201810321384.5A 2018-04-11 2018-04-11 Disaster rescue emergency resource scheduling method based on multi-agent genetic algorithm Active CN108537338B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810321384.5A CN108537338B (en) 2018-04-11 2018-04-11 Disaster rescue emergency resource scheduling method based on multi-agent genetic algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810321384.5A CN108537338B (en) 2018-04-11 2018-04-11 Disaster rescue emergency resource scheduling method based on multi-agent genetic algorithm

Publications (2)

Publication Number Publication Date
CN108537338A true CN108537338A (en) 2018-09-14
CN108537338B CN108537338B (en) 2020-07-28

Family

ID=63479622

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810321384.5A Active CN108537338B (en) 2018-04-11 2018-04-11 Disaster rescue emergency resource scheduling method based on multi-agent genetic algorithm

Country Status (1)

Country Link
CN (1) CN108537338B (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109711046A (en) * 2018-12-26 2019-05-03 电子科技大学 The efficient integrated approach of Large Scale Sparse array based on multi-Agent Genetic Algorithm
CN109886584A (en) * 2019-02-27 2019-06-14 安徽大学 A kind of heat supply network dispatching method and device based on intelligent body
CN110020778A (en) * 2019-02-21 2019-07-16 国网山东省电力公司临沂供电公司 The power emergency distribution of materials and configuration system and method
CN113111476A (en) * 2021-04-29 2021-07-13 华北电力大学 Human-vehicle-object emergency resource optimal scheduling method for improving toughness of power grid
CN117745042A (en) * 2024-02-20 2024-03-22 广东省科技基础条件平台中心 Multi-disaster-point emergency material scheduling method, device, medium and equipment

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7246075B1 (en) * 2000-06-23 2007-07-17 North Carolina A&T State University System for scheduling multiple time dependent events
US20080215512A1 (en) * 2006-09-12 2008-09-04 New York University System, method, and computer-accessible medium for providing a multi-objective evolutionary optimization of agent-based models
CN105976030A (en) * 2016-03-15 2016-09-28 武汉宝钢华中贸易有限公司 Multi-agent-based platform scheduling intelligent sorting model structure
CN106910018A (en) * 2017-02-23 2017-06-30 吉林大学 One kind rescue resource regulating method and system
CN107292450A (en) * 2017-07-18 2017-10-24 西安电子科技大学 Disaster assistance ambulance paths planning method based on multi-Agent Genetic Algorithm
CN107730056A (en) * 2017-11-22 2018-02-23 合肥工业大学 A kind of emergency materials modularization dispatching method based on improvement NSGA II algorithms

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7246075B1 (en) * 2000-06-23 2007-07-17 North Carolina A&T State University System for scheduling multiple time dependent events
US20080215512A1 (en) * 2006-09-12 2008-09-04 New York University System, method, and computer-accessible medium for providing a multi-objective evolutionary optimization of agent-based models
CN105976030A (en) * 2016-03-15 2016-09-28 武汉宝钢华中贸易有限公司 Multi-agent-based platform scheduling intelligent sorting model structure
CN106910018A (en) * 2017-02-23 2017-06-30 吉林大学 One kind rescue resource regulating method and system
CN107292450A (en) * 2017-07-18 2017-10-24 西安电子科技大学 Disaster assistance ambulance paths planning method based on multi-Agent Genetic Algorithm
CN107730056A (en) * 2017-11-22 2018-02-23 合肥工业大学 A kind of emergency materials modularization dispatching method based on improvement NSGA II algorithms

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
YAWEN ZHOU ETAL.: "A multi-agent genetic algorithm for multi-period emergency resource scheduling problems in uncertain traffic network", 《2017 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION》 *
蔡冰琦: "资源受限项目调度问题编码分析及进化优化", 《中国优秀硕士学位论文全文数据库 信息科技辑》 *

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109711046A (en) * 2018-12-26 2019-05-03 电子科技大学 The efficient integrated approach of Large Scale Sparse array based on multi-Agent Genetic Algorithm
CN109711046B (en) * 2018-12-26 2019-09-20 电子科技大学 Large Scale Sparse array synthetic method based on multi-Agent Genetic Algorithm
CN110020778A (en) * 2019-02-21 2019-07-16 国网山东省电力公司临沂供电公司 The power emergency distribution of materials and configuration system and method
CN109886584A (en) * 2019-02-27 2019-06-14 安徽大学 A kind of heat supply network dispatching method and device based on intelligent body
CN109886584B (en) * 2019-02-27 2022-09-02 安徽大学 Heat supply network scheduling method and device based on intelligent agent
CN113111476A (en) * 2021-04-29 2021-07-13 华北电力大学 Human-vehicle-object emergency resource optimal scheduling method for improving toughness of power grid
CN117745042A (en) * 2024-02-20 2024-03-22 广东省科技基础条件平台中心 Multi-disaster-point emergency material scheduling method, device, medium and equipment
CN117745042B (en) * 2024-02-20 2024-05-24 广东省科技基础条件平台中心 Multi-disaster-point emergency material scheduling method, device, medium and equipment

Also Published As

Publication number Publication date
CN108537338B (en) 2020-07-28

Similar Documents

Publication Publication Date Title
CN108537338A (en) Disaster assistance emergency resources dispatching method based on multi-Agent Genetic Algorithm
Pholdee et al. Hybridisation of real-code population-based incremental learning and differential evolution for multiobjective design of trusses
CN109840154B (en) Task dependency-based computing migration method in mobile cloud environment
CN109299781A (en) Distributed deep learning system based on momentum and beta pruning
Esmin et al. Hybrid evolutionary algorithm based on PSO and GA mutation
CN104516785B (en) A kind of cloud computing resources dispatch system and method
CN111325356A (en) Neural network search distributed training system and training method based on evolutionary computation
CN106610654A (en) Improved genetic algorithm for flexible workshop scheduling
CN106250650A (en) The resource allocation and optimization method of model in high flux emulation
CN111626545B (en) Emergency rescue material high-dimensional multi-target distribution method and system
CN106610867B (en) On-chip network task scheduling method and device
CN110135584A (en) Extensive Symbolic Regression method and system based on self-adaptive parallel genetic algorithm
WO2014173272A1 (en) Layout method and layout system for monitoring units of digital man-machine interface
Ye et al. A new approach for resource scheduling with deep reinforcement learning
CN106611378A (en) Priority encoding-based hybrid genetic algorithm for solving job-shop scheduling problem
CN104143116B (en) System soft protection combinatorial optimization method based on particle swarm optimization
CN104536831B (en) A kind of multinuclear SoC software image methods based on multiple-objection optimization
CN111414961A (en) Task parallel-based fine-grained distributed deep forest training method
CN105205536B (en) 1553B bus message transmission optimization methods based on genetic algorithm
CN106096722A (en) Transport path combined optimization method based on stage structures Predator-Prey model
CN109582457A (en) Network-on-chip heterogeneous multi-core system task schedule and mapping
CN113848904B (en) Method for optimizing task allocation of multiple mobile robots based on punishment energy consumption
CN111027760B (en) Power load prediction method based on least square vector machine
CN115841230A (en) Emergency resource overall planning and allocating method based on improved NSGA-II algorithm
CN107396375B (en) Quantum brain storm-based optimal target coverage method for heterogeneous sensor network

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant