CN103489042A - Method for optimizing disaster emergency decision system path - Google Patents

Abstract

The invention relates to the technical field of artificial intelligence algorithms, and provides a method for optimizing a disaster emergency decision system path. The method comprises the steps that S11, an objective function and a fitness function are defined; S12, chromosome coding is carried out; S13, operators are selected according to the fitness function; S14, the operators are crossed; S15, the operators are mutated according to the fitness function and pheromone updating guiding mutation rules; S16, a plurality of sets of optimal solutions are generated, and the best solution is output through the fitness function. The quality and efficiency of the solutions are effectively improved through the pheromone guiding mutation rules of an ant colony algorithm. A large amount of useless redundant iteration of a genetic algorithm is avoided through the positive feedback mechanism and the parallelism of the ant colony algorithm, and the solution speed of the genetic algorithm is further improved. The decision model with a single saving point and multiple disaster points based on continuous supply losses is more suitable for emergency situations, and practical significance is achieved.

Description

The method of a kind of calamity emergency decision system path optimization

[technical field]

The present invention relates to the intelligent algorithm technical field, particularly relate to the method for a kind of calamity emergency decision system path optimization.

[background technology]

Over the past thousands of years, the huge threat to human life's composition of estate invariably of disaster, human accident, therefore how to improve urgent distribution of materials efficiency extremely important to reduce injures and deaths and economic loss in research.Along with development in science and technology is rapid, in recent years, people are more and more and more and more deep to the research of calamity emergency system path optimization problem, uncertain factor also appears at emergent decision model limit of consideration, but most research is rescued disaster-stricken some decision model of a list for probabilistic having more, and probabilisticly singly go out to rescue a little how disaster-stricken point and probabilistic having more and rescue some many resource decisions models, the research and comparison of goods and materials continuous consumption that even has time restriction is few.The former is relatively simple, but too idealized, simplification is difficult to fulfill; Both relative complex, still more realistic afterwards.

At present, the algorithm of calamity emergency decision system path optimization is mainly contained to genetic algorithm, ant group algorithm, particle cluster algorithm, simulated annealing, control algolithm, immune algorithm etc., various algorithms respectively have quality, thereby the research that improves various algorithms emerges in an endless stream.Wherein, genetic algorithm, genetic-ant colony algorithm, ant group genetic algorithm of simulated annealing etc. are arranged, by improved algorithm, optimum solution is kept on improving.

Genetic algorithm (Genetic Algorithm) is that the evolution laws (survival of the fittest, survival of the fittest genetic mechanism) that a class is used for reference organic sphere develops and next randomization searching method.At first initialization population, then calculate each individual fitness, select operator according to fitness, to selecting operator to carry out calculated crosswise, obtain crossover operator, carry out mutation operator according to the fitness function value, the fitness of operator relatively, recursive iteration, finally obtain the optimization solution of some groups, and then therefrom select optimum solution.

Ant group algorithm (Ant Colony Optimization, ACO), claim again ant algorithm, is a kind of probability type algorithm that is used for finding in the drawings path optimizing.Each ant starts search of food not telling in advance under where prerequisite of their food.When one find food after, its can discharge a kind of volatile secretion information to environment and usually realize (this material disappearance of As time goes on can volatilizing gradually, the size of pheromone concentration characterizes the distance in path), attract other ant to come, so increasing ant can be found food.Some ant can look for another way, if the road of separately opening up is shorter, so, gradually, more ant attracted on this compare Duan road.Finally, through operation after a while, may occur that a path the shortest repeated by most of ants.

At present, genetic algorithm or ant group algorithm are widely applied in the middle of the optimization of calamity emergency system path.By five operations greatly of genetic algorithm, can more effectively solve singly to go out to rescue a little to enter to rescue a decision model more and rescue a little multiobject Optimized Operation scheme problem with having more to rescue a little to enter more, as solve minimum crash time, least consume, minimumly the desired value such as go out to rescue a little.Genetic algorithm has stronger quick random ability of searching optimum, but, owing to lacking feedback information, often in iterative process, has the redundancy iteration of a large amount of inactions, the accurate efficiency that impact solves.

Ant group algorithm mainly solves shortest path, can solve target, go out to rescue point, the individual layer that enters to rescue point, resource etc. or the various combination problem of multilayer.The accumulation of ant group algorithm by pheromones and renewal converge on active path, though there is the distributed parallel ability of searching optimum, pheromones scarcity originally solves speed very slow.

Above-mentioned decision model has been optimized in the fusion of genetic-ant colony algorithm or genetic algorithm and ant group algorithm.These improved algorithms are respectively got the chief, and then have improved optimization efficiency.Improved algorithm becomes the more strong instrument of shot array problem that solves, but due to the firm rise that improves algorithm, improved route and method are still in the stage of fumbling, and it is optimized space and still can further promote, also can on improved basis, be improved, more be strengthened the effect of optimizing.

Given this, overcoming the existing defect of above-mentioned prior art is the art problem demanding prompt solution.

[summary of the invention]

The technical problem to be solved in the present invention is to provide quality and the efficiency that a kind of raising solves, and meets the method for the calamity emergency decision system path optimization of emergent present situation.

The present invention adopts following technical scheme:

The method of a kind of calamity emergency decision system path optimization comprises the following steps:

Step S11: objective definition function and fitness function, wherein objective function f _r=min S, objective function minimizes in the situation without interruption based on goods and materials, solving crash time S, fitness function

Step S12: set up and contact between the chromosome bit string structure of the actual expression of target problem and genetic algorithm, determine the Code And Decode computing, carry out chromosome coding;

Step S13: according to fitness function F _rselect operator, calculate each individual fitness, select probability and accumulated probability, by repeatedly selecting to select the individuality of intersection;

Step S14: crossover operator, take the discrete recombination of real-valued restructuring, calculate the fitness value after restructuring, the fitness value before and after relatively intersecting, the operator after determining to select to intersect still retains original operator;

Step S15: according to fitness function F _rupgrade and instruct the regular mutation operator of variation with pheromones;

Step S16: generate some groups of optimization solutions, and by fitness function F _routput is preferably separated.

Further, described method also comprises:

The initialization population, comprise initialization current go out to rescue a little provide the quantity initial value of goods and materials, current residue deposit quantity initial value, initial population scale, population scale and the termination evolutionary generation that goes out to rescue current goods and materials a little for a certain disaster-stricken point.

Further, after step S15, also comprise:

Step S15 ＇: more current population's fitness and population's fitness before, if without obviously improving and while arriving iteration parameter, stopping carrying out, otherwise, return to execution step S12.

Further, in described step S12, coded system is a kind of in real coding, binary coding, out of order coding, adaptive coding or tree-encoding sequential coding.

Further, the specific implementation step of initialization population comprises:

Steps A 1: judge whether to exist the actual goods and materials of disaster-stricken some i to distribute not meet it required, if, therefrom choose at random disaster-stricken some i, execution step A2, otherwise, execution step A4;

Steps A 2: in the disaster-stricken some i chosen in determining step A1, whether exist actual goods and materials quantity allotted not meet that it is required, if, therefrom choose at random goods and materials j, execution step A3, otherwise, return to execution step A1;

Steps A 3: judge whether the current residue deposit quantity that goes out to rescue goods and materials j a little is greater than 0, if, generation 0～min(preStoij, Qij-preRij) the random number random between is a little the quantity of disaster-stricken some i dispensing goods and materials j as originally going out to rescue, preRij=preRij+random, preStoij=preStoij-random, return to execution step A2, otherwise, construct unsuccessfully, exit;

Steps A 4:Rij=preRij, be successfully constructed, and generates initial population;

Wherein, preRij goes out to rescue for current the quantity that goods and materials j a little is provided for disaster-stricken some i, and its initial value is 0, preStoj is the current residue deposit quantity that goes out to rescue goods and materials j a little, initial value is Stoj, and Qij is the disaster-stricken some goods and materials j that i is required, and Rij is the quantity that disaster-stricken some i needs j kind emergency resources.

Further, in described step S13, while calculating the selection probability of each individuality, take pro rata fitness assignment, establishing Population Size is L, and individuality is R, selects probability

Combine to select operator according to roulette wheel selection and the direct reservation method of optimum solution, before colony intersects, the optimized individual of some is genetic directly to the next generation, residue is individual to be selected according to roulette wheel selection according to its fitness, fitness is larger, and selected probability is larger;

While repeatedly being selected, produce the random number random of [0,1], random number is uniformly distributed at every turn, by random number, determines final selected individuality.

Further, in described step S14, the specific implementation step of crossover operator comprises:

Step S141: two individualities of random selection from population, produce the random number random of [0,1], judge whether to need to intersect, if random<Pc, Pc is crossover probability, performs step S142, otherwise, retain original operator to of future generation;

Step S142: produce the random number random of [1, p], exchange the distribution condition of goods and materials R in two individualities.

Further, described step S15 specifically comprises:

According to the variation probability P, m carries out mutation operation, the random change point at random of selecting; According to ant group pheromones τ _ij(t) whether determine the exchange mutation point, the pheromones of guaranteeing path before and after the variation position after exchange is high before than exchange.

Compared with prior art, beneficial effect of the present invention is:

1, the pheromones by ant group algorithm instructs the variation rule, has effectively improved the quality and the efficiency that solve;

2, utilize positive feedback mechanism and the concurrency of ant group algorithm, avoid a large amount of unhelpful redundancy iteration of genetic algorithm, further improved genetic algorithm for solving speed;

3, based on the continuous loss of goods and materials singly go out to rescue the some decision model of how disaster-stricken point, more meet emergent present situation, there is practical significance.

[accompanying drawing explanation]

Fig. 1 is the method flow diagram of the calamity emergency decision system path optimization of the embodiment of the present invention.

[embodiment]

In order to make purpose of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein, only in order to explain the present invention, is not intended to limit the present invention.

In addition, below in each embodiment of described the present invention involved technical characterictic as long as form each other conflict, just can mutually not combine.

As shown in Figure 1, be the method flow diagram of the calamity emergency decision system path optimization of the embodiment of the present invention.

the description of problem:

(1) mathematical modeling of calamity emergency system:

If A ₁, A ₂a _mfor m disaster-stricken point, A is for going out to rescue a little, and R is the emergency materials demand.A _iresources requirement be R _j, i=1,2,3 ..., m, j=1,2,3 ... n, and meet

during emergent movable the beginning, consumption of materials speed is V, from A to A _ithe time needed is t _i(>0).If t ₁≤ t ₂≤ ...≤t _m, the known demand of emergency materials separately, on the continual basis of the consumption of materials, make emergency start time the earliest.Supposing that T expresses to rescue participates in emergent deadline, S(>=t ₁) be emergency start time, if option b emergency start time S is feasible continuously, if

have

$\underset{k\&Element;\left\{i\right|t_i<t,i=\mathrm{1,2}...,m\}}{\mathrm{\&Sigma;}}{R}_{\mathrm{kj}}\&GreaterEqual;(t-S)v_j$

In above formula, the left side means the goods and materials amount that t has arrived constantly; The right means the t consumption of goods and materials constantly.

If meet above-mentioned formula, in option b, zero-time S is feasible continuously.

If the set that is continuously feasible scheme all about zero-time S is R _s, the Solve problems of optimal case is:

Scheme 1 is that the j resource is feasible continuously about zero-time s, if right

have

$\underset{k\&Element;\left\{i\right|t_i<t,i=\mathrm{1,2}...,m\}}{\mathrm{\&Sigma;}}{R}_{\mathrm{kj}}\&GreaterEqual;(t-S)v_j$

If all about zero-time s be the j resource continuously the set of feasible all schemes be R ^j _s, thereby many resource problems can be expressed as:

If above-mentioned formula optimal objective value is S ^*, that is:

Adopt matrix form, arbitrary option b can be expressed as:

$\left[\begin{array}{cccc}{R}_{11}& R_{12}& ...& R_{1n}\\ R_{21}& R_{22}& ...& R_{2n}\\ R_{m1}& R_{m2}& ...& R_{\mathrm{mn}}\end{array}\right]$

0≤R wherein _ij≤ R ' _ij,

i=1,2 ..., m, j=1,2 ..., n.

R ' _ijmean A _ij kind resources requirement, R _ija while meaning option b _ineed j kind emergency resources quantity.Visible, disaster-stricken some A of i line display _ithe emergent vector needed, if this vector equals 0, express and rescue the resource emergency requirement that an A does not arrange this disaster-stricken point.The j list illustrates the kind of rescuing a little disaster-stricken some emergency resources.

If t ₁≤ t ₂≤ ...≤t _mthe strict establishment, the crash time of option b early start is:

$\begin{array}{c}{S}^{*}\underset{k\&Element;\{\mathrm{1,2},...p_j\}}{\max }[t_k-{\mathrm{\&Sigma;}}_{i\&Element;\left\{i\right|t_i<t_k,i=\mathrm{1,2}...,p_j\left\}\frac{R_{\mathrm{ij}}}{v_j}\right]}\\ =(t_k-\underset{i=0}{\overset{k-1}{\mathrm{\&Sigma;}}}\frac{R_{\mathrm{ij}}}{v_j})_{k\&Element;\{\mathrm{1,2},...,p_j\}}^{\max }\end{array}$

Corresponding prioritization scheme B _j ^*

$\begin{array}{c}{B_j}^{*}=\\ {(R_{1j},R_{2j},...,R_{\mathrm{pj}-1j},R_j-\underset{k=0}{\overset{\mathrm{pj}-1}{\mathrm{\&Sigma;}}}{R_k}_j,0,...,0)}^T\end{array}$

P _jmeet following formula:

$\underset{k=0}{\overset{\mathrm{pj}-1}{\mathrm{\&Sigma;}}}{R}_{\mathrm{ij}}<R_j<\underset{k=0}{\overset{\mathrm{pj}}{\mathrm{\&Sigma;}}}{R}_{\mathrm{kj}}$

(2) introduce the central ant group algorithm of genetic algorithm:

Pheromones based on improved ant group algorithm-MMAS is upgraded and is instructed the variation rule:

${\mathrm{\&tau;}}_{\mathrm{ij}}(t+1)=\mathrm{\&rho;}\&CenterDot;{\mathrm{\&tau;}}_{\mathrm{ij}}\left(t\right)+\mathrm{\&Delta;}{\mathrm{\&tau;}}^{\mathrm{bes}{t}_{\mathrm{ij}}}$

Wherein, ρ means the lasting factor of pheromones, in order to prevent pheromones, infinitely accumulates, and the span of ρ is initial information element τ _ij(t)=c is maximal value τ _max;

f(s ^best) expression iteration optimum solution (s ^ib) or globally optimal solution (s ^gb) value.Mainly use the iteration optimum solution at this.

For fear of precocious, stagnation behavior, by the impact of restricted information element track, can avoid at an easy rate in the algorithm operational process between each pheromones track difference excessive., MMAS has applied respectively τ to minimum value and the maximal value of pheromones track respectively _minand τ _maxthereby, make all pheromones track τ _ij(t), τ is arranged _min<τ _ij(t)<τ _max.If τ is arranged _ij(t)>τ _max, τ is set _ij(t)=τ _max; If τ _ij(t)>τ _min, τ is set _ij(t)>τ _min.

Based on this, the embodiment of the present invention provides the method for a kind of calamity emergency decision system path optimization.At first initialization population, specifically comprise initialization current go out to rescue a little provide the quantity initial value of goods and materials, current residue deposit quantity initial value, initial population scale, population scale and the termination evolutionary generation that goes out to rescue current goods and materials a little for a certain disaster-stricken point.If preRij goes out to rescue for current the quantity that emergency materials j a little is provided for disaster-stricken some i, its initial value is 0, establishes preStoj and goes out to rescue a little residue deposit quantity of current goods and materials j for this reason, and initial value is Stoj, Qij is the disaster-stricken some goods and materials j that i is required, and Rij is the quantity that disaster-stricken some i needs j kind emergency resources.Initial population scale L is set., population scale L, and stop evolutionary generation G.The specific implementation step of initialization population comprises:

Steps A 1: judge whether to exist the actual goods and materials of disaster-stricken some i to distribute not meet it required, if, therefrom choose at random disaster-stricken some i, execution step A2, otherwise, execution step A4;

Steps A 2: in the disaster-stricken some i chosen in determining step A1, whether exist actual goods and materials quantity allotted not meet that it is required, if, therefrom choose at random goods and materials j, execution step A3, otherwise, return to execution step A1;

Steps A 3: judge whether the current residue deposit quantity that goes out to rescue goods and materials j a little is greater than 0, if, generation 0～min(preStoij, Qij-preRij) the random number random between is a little the quantity of disaster-stricken some i dispensing goods and materials j as originally going out to rescue, preRij=preRij+random, preStoij=preStoij-random, return to execution step A2, otherwise, construct unsuccessfully, exit;

Steps A 4:Rij=preRij, be successfully constructed, and generates initial population.

the method that the present embodiment provides comprises the steps:

Step S11: objective definition function and fitness function.

According to front surface analysis, the target of this calamity emergency decision system is based in goods and materials situation without interruption, and the crash time, S minimized.So objective function is f _r=min S.

The present invention requires the minimum value of objective function, the fitness function of therefore choosing

Step S12: generate at random one group of real coding.

While using the genetic algorithm for solving problem, must between the chromosome bit string structure of the actual expression of target problem and genetic algorithm, set up and contact, determine the Code And Decode computing.Coded system can adopt binary coding, real coding, out of order coding, adaptive coding, tree-encoding sequential coding etc.Adopt real coding in the present embodiment, as option b:

$\begin{array}{c}B=\left\{(A_i,R_{\mathrm{ij}})\right\}\\ B_1^{*}=\{(A_1,9),(A_2,23),(A_3,56),(A_4,32),(A_5,15),(A_6,5)\}\\ B_2^{*}=\{(A_1,3),(A_2,18),(A_3,29),(A_4,41),(A_5,66)\}\end{array}$

The such scheme vector representation is:

$B_1^{*}={\left(\mathrm{9,23,56,32,15,5}\right)}^T$

$B_2^{*}={\left(\mathrm{3,18,29,41,66,0}\right)}^T$

Step S13: according to fitness function F _rselect operator.

Calculate each individual fitness, select probability and accumulated probability, by repeatedly selecting to select the individuality of intersection.

While calculating the selection probability of each individuality, take pro rata fitness assignment, claim again the Monte Carlo method of selecting, establishing Population Size is L, and individuality is R, selects probability

Combine to select operator according to roulette wheel selection and the direct reservation method of optimum solution, before colony intersects, the optimized individual of some is genetic directly to the next generation, residue is individual to be selected according to roulette wheel selection according to its fitness, fitness is larger, and selected probability is larger;

The specific implementation step is as follows:

(1) calculate each individual fitness, select probability and accumulated probability;

(2) in order to select the individuality of intersection, need to repeatedly be selected.Each random number random that produces [0,1], random number need to be uniformly distributed, and by random number, determines final selected individuality.

Step S14: crossover operator acts on population.

Intersect and claim again genetic recombination, two kinds of real-valued restructuring and scale-of-two restructuring are arranged.The present embodiment is taked the discrete recombination of real-valued restructuring, calculates the fitness value after recombinating, the fitness value before and after relatively intersecting, and decision selects the operator after intersection still to retain original operator.Like this, just can there be blindness in crossover operator, and a cognition of intersecting is more outstanding.Its specific implementation step is:

Step S141: two individualities of random selection from population, produce the random number random of [0,1], judge whether to need to intersect, if random<Pc, Pc is crossover probability, performs step S142, otherwise, retain original operator to of future generation;

Step S142: produce the random number random of [1, p], exchange the distribution condition of goods and materials R in two individualities, for example:

（R ₁₁，R ₁₂，…，R ₂₁，R ₂₂，…，R _mn）（R‘ ₁₁，R’ ₁₂，…，R‘ ₂₁，R’ ₂₂，…，R‘ _mn）

New individuality after intersection is:

（R‘ ₁₁，R ₁₂，…，R ₂₁，R’ ₂₂，…，R‘ _mn）（R ₁₁，R’ ₁₂，…，R‘ ₂₁，R ₂₂，…，R _mn）

Step S15: according to fitness function F _rupgrade and instruct the regular mutation operator of variation with pheromones.Specifically comprise:

According to the variation probability P, m carries out mutation operation, the random change point at random of selecting; According to ant group pheromones τ _ij(t) whether determine the exchange mutation point, the pheromones of guaranteeing path before and after the variation position after exchange is high before than exchange.

Step S16: generate some groups of optimization solutions, and by fitness function F _routput is preferably separated.

In the emulation experiment of matlab, algorithm design maximum iterations, also be provided with an iteration parameter, the end condition of algorithm is set, is specially and performs step S15 ＇ after step S15, carry out recursive iteration: more current population's fitness and population's fitness before, if during without obvious raising and arrival iteration parameter, stop carrying out, otherwise, execution step S12 returned to.

The present embodiment has following beneficial effect:

1, improve traditional genetic algorithm by ant group algorithm, optimize calamity emergency decision system path;

2, the pheromones by ant group algorithm instructs the variation rule, guarantees that the routing information element before genetic algorithm variation Hou path, position is than variation is high, has effectively improved the quality and the efficiency that solve;

3, utilize positive feedback mechanism and the concurrency of ant group algorithm, avoid a large amount of unhelpful redundancy iteration of genetic algorithm, make the genetic algorithm mutation operator have more Objective, further improved genetic algorithm for solving speed;

4, by improved ant group genetic algorithm, work out the calamity emergency system decision-making model of more realistic meaning, such as the present embodiment based on the continuous loss of goods and materials singly go out to rescue a little how disaster-stricken many resource decisions of some model, more meet emergent present situation, there is practical significance.

The embodiment of the present invention is by being used the improvement of ant group algorithm, genetic algorithm, can be applicable to the fields such as function optimization, Combinatorial Optimization, control automatically, machine learning, image processing, artificial life, genetic coding and robotics, can solve the optimization of calamity emergency decision system, traveling salesman problem (Traveling Salesman Problem, TSP), evacuating personnel, knapsack problem, figure partition problem and bin packing when Chinese postman problem (Chinese Postman Problem, CPP), accident.

One of ordinary skill in the art will appreciate that all or part of step in the whole bag of tricks of embodiment is to come the hardware that instruction is relevant to complete by program, this program can be stored in a computer-readable recording medium, storage medium can comprise: ROM (read-only memory) (ROM, Read Only Memory), random access memory (RAM, Random Access Memory), disk or CD etc.

The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, all any modifications of doing within the spirit and principles in the present invention, be equal to and replace and improvement etc., within all should being included in protection scope of the present invention.

Claims (8)

1. the method for a calamity emergency decision system path optimization, is characterized in that, comprises the following steps:

Step S11: objective definition function and fitness function, wherein objective function f _r=min S, objective function minimizes in the situation without interruption based on goods and materials, solving crash time S, fitness function

Step S12: set up and contact between the chromosome bit string structure of the actual expression of target problem and genetic algorithm, determine the Code And Decode computing, carry out chromosome coding;

Step S13: according to fitness function F _rselect operator, calculate each individual fitness, select probability and accumulated probability, by repeatedly selecting to select the individuality of intersection;

Step S14: crossover operator, take the discrete recombination of real-valued restructuring, calculate the fitness value after restructuring, the fitness value before and after relatively intersecting, the operator after determining to select to intersect still retains original operator;

Step S15: according to fitness function F _rupgrade and instruct the regular mutation operator of variation with pheromones;

Step S16: generate some groups of optimization solutions, and by fitness function F _routput is preferably separated.

2. the method for calamity emergency decision system according to claim 1 path optimization, is characterized in that, described method also comprises:

The initialization population, comprise initialization current go out to rescue a little provide the quantity initial value of goods and materials, current residue deposit quantity initial value, initial population scale, population scale and the termination evolutionary generation that goes out to rescue current goods and materials a little for a certain disaster-stricken point.

3. the method for calamity emergency decision system according to claim 1 path optimization, is characterized in that, after step S15, also comprises:

Step S15 ＇: more current population's fitness and population's fitness before, if without obviously improving and while arriving iteration parameter, stopping carrying out, otherwise, return to execution step S12.

4. the method for calamity emergency decision system according to claim 1 path optimization, is characterized in that, in described step S12, coded system is a kind of in real coding, binary coding, out of order coding, adaptive coding or tree-encoding sequential coding.

5. the method for calamity emergency decision system according to claim 2 path optimization, is characterized in that, the specific implementation step of initialization population comprises:

Steps A 1: judge whether to exist the actual goods and materials of disaster-stricken some i to distribute not meet it required, if, therefrom choose at random disaster-stricken some i, execution step A2, otherwise, execution step A4;

Steps A 2: in the disaster-stricken some i chosen in determining step A1, whether exist actual goods and materials quantity allotted not meet that it is required, if, therefrom choose at random goods and materials j, execution step A3, otherwise, return to execution step A1;

Steps A 3: judge whether the current residue deposit quantity that goes out to rescue goods and materials j a little is greater than 0, if, generation 0～min(preStoij, Qij-preRij) the random number random between is a little the quantity of disaster-stricken some i dispensing goods and materials j as originally going out to rescue, preRij=preRij+random, preStoij=preStoij-random, return to execution step A2, otherwise, construct unsuccessfully, exit;

Steps A 4:Rij=preRij, be successfully constructed, and generates initial population;

Wherein, preRij goes out to rescue for current the quantity that goods and materials j a little is provided for disaster-stricken some i, and its initial value is 0, preStoj is the current residue deposit quantity that goes out to rescue goods and materials j a little, initial value is Stoj, and Qij is the disaster-stricken some goods and materials j that i is required, and Rij is the quantity that disaster-stricken some i needs j kind emergency resources.

6. the method for calamity emergency decision system according to claim 1 path optimization, is characterized in that, in described step S13, while calculating the selection probability of each individuality, take pro rata fitness assignment, establishing Population Size is L, and individuality is R, selects probability

Combine to select operator according to roulette wheel selection and the direct reservation method of optimum solution, before colony intersects, the optimized individual of some is genetic directly to the next generation, residue is individual to be selected according to roulette wheel selection according to its fitness, fitness is larger, and selected probability is larger;

While repeatedly being selected, produce the random number random of [0,1], random number is uniformly distributed at every turn, by random number, determines final selected individuality.

7. the method for calamity emergency decision system according to claim 1 path optimization, is characterized in that, in described step S14, the specific implementation step of crossover operator comprises:

Step S141: two individualities of random selection from population, produce the random number random of [0,1], judge whether to need to intersect, if random<Pc, Pc is crossover probability, performs step S142, otherwise, retain original operator to of future generation;

Step S142: produce the random number random of [1, p], exchange the distribution condition of goods and materials R in two individualities.

8. the method for calamity emergency decision system according to claim 1 path optimization, is characterized in that, described step S15 specifically comprises:

According to the variation probability P, m carries out mutation operation, the random change point at random of selecting; According to ant group pheromones τ _ij(t) whether determine the exchange mutation point, the pheromones of guaranteeing path before and after the variation position after exchange is high before than exchange.