CN106156981A - Logistics collaboration processing method based on cloud computing - Google Patents

Abstract

The invention provides a kind of logistics collaboration processing method based on cloud computing, the method includes: cloud platform electricity business accepts the sequence information of user, minimizes target selection shipper according to user's cost of serving and carries out transportation service；Solve above-mentioned target, in conjunction with the user's statistical data obtained in advance, it was predicted that for the optimal time of each user's delivery.The present invention proposes a kind of logistics collaboration processing method based on cloud computing, based on the big data analysis of user, while ensureing to meet user's subjectivity demand, logistic resources can be optimized again, farthest improve the satisfaction of user, achieve double benefit economical and environmentally friendly.

Description

Logistics collaboration processing method based on cloud computing

Technical field

The present invention relates to Intelligent logistics, particularly to a kind of logistics collaboration processing method based on cloud computing.

Background technology

Popularizing and developing rapidly of big data, also brings unprecedented while carrying out major opportunity to logistics transportation industrial belt Challenge, personalized service and the environmental protection concept of user are pulled to unprecedented height.Order allocation in enterprise In planning time, the individual demand of user and the ability of each shipper, resource, satisfaction etc., also vehicle transport mistake Oil consumption in journey and discharge are all the decision factors that logistics service is important.Therefore, the potential hobby of analysis mining user, use The Behavior law at family and policy of environment protection factor etc., and go to improve the service ability of loglstics enterprise according to analytical data result, favorably In the development the most stable with enterprise and the lifting of economic and social benefits.Existing logistics transportation scheme considers only emphatically transport road How line is planned shortening time, vehicle could reach peak load the most as far as possible, and to ecological requirements and the transport of user Demand includes that the factor such as expected time, satisfaction considers deficiency, and then cannot distribute logistics task targetedly.

Summary of the invention

For solving the problem existing for above-mentioned prior art, the present invention proposes at a kind of logistics collaboration based on cloud computing Reason method, including:

Cloud platform electricity business accepts the sequence information of user, minimizes target selection shipper according to user's cost of serving and carries out Transportation service；

Solve above-mentioned target, in conjunction with the user's statistical data obtained in advance, it was predicted that for the optimal time of each user's delivery.

Preferably, described user's cost of serving minimizes target and user's cost of serving is expressed as the price of each shipper Constituting with penalty, its object function is:

$\min Z=\underset{i=1}{\overset{n}{\Σ}}\underset{j=1}{\overset{m}{\Σ}}{y}_{ij}{p}_{ij}+c^{\′}{{}_i}_j-\underset{i=1}{\overset{n}{\Σ}}{q}_i{c}_i$

And meet following constraint

$A_{ij}=(1+(p_j-p_{ij})/p_j){(Q_{ij}/Q_j)_{{\mathrm{\γ}}_j}}^{{\mathrm{\θ}}_j}$

A is made when there is k_ijMore than A_kjTime, y_ij=1, wherein i ≠ k,

Otherwise y_ij=0

When t is more than T_jTime c '_ij=-π (t-T_j)p_ij,

Otherwise c '_ij=0

M=a_minT(A∩B)

$q_i=\underset{j=1}{\overset{m}{\Σ}}{y}_{ij}$

$H_i=\underset{j=1}{\overset{m}{\Σ}}{y}_{ij}(p_{ij}-c_{ij})$

$\underset{i=1}{\overset{n}{\Σ}}{q}_i=m$

q_i≤x_i

p_ij≤p_j

γ_j, θ_j∈[0,1]；

Wherein t represents the time of actual service, T_jRepresent the desired optimum delivery of user drawn based on big data prediction Time；π represents time-out penalty coefficient；N represents shipper quantity；A_ijRepresent the user j evaluation of estimate to shipper i；q_iRepresent i-th The delivering amount of individual shipper；c_ijRepresent the shipper i cost of serving to j；p_ijRepresent the shipper i price to user j；Q_ijTable Show the shipper i service quality to j；p_jRepresent the acceptable ceiling price of user j；Q_jRepresent service quality expected from user j； γ_jRepresent the user j attention rate to price change；θ_jRepresent the user j attention rate to service quality；H_iRepresent the profit of shipper Profit；S_iRepresent that shipper can participate in the minimum profit of service；x_iRepresent shipper i maximum service ability；c_iRepresent and product is provided Price；M is sales volume, and view rate and concern time are respectively A and B, choose a large number of users browsing data, represent with amin and browse There is the degree of association between sales volume more than predetermined threshold value in rate and concern time, according to first checking method, excavate generation and meet simultaneously The Candidate Set that support C is minimum, obtains degree of association a_min；Then by formula m=a_minT (A ∩ B) calculates user's request amount m；T(A∩ B) | represent event A and event B concurrent number；Wherein support C is T (A ∩ B) |/T | (A) |, T | (A) | represents thing The sum that part A occurs；

The described optimal time being predicted as each user's delivery, farther includes:

Step 1: prediction user's optimum delivery availability, input time, sequence, obtained segmentation matrix Q_ij；Calculate each Q_ij's Quantity, and utilize degree of association formula to calculate all adjacent Q_ijBetween relatedness, and find out the Q meeting condition_ij；

Step 2: stochastic generation initial solution Ω₍₀₎；The scale determining initial population is R, and crossover probability is P_j, mutation probability For P_bWith termination evolutionary criterion, arranging evolutionary generation enumerator t is 0；

Step 3: calculate individual fitness, using the inverse of target function value minZ as individual fitness value；

Step 4: from Ω_(t)Middle utilization selection opertor selects N/2 to parent, and wherein N > R, according to probability P_jPerform X-type Become individual in the middle of N number of；N number of middle individuality is carried out independently, according to probability P_bPerform variation, form N number of candidate individual；From time Select in individuality and select R individual composition a new generation population Ω according to ideal adaptation degree_(t+1)；

Step 5: as met end condition set in advance, then export Ω_(t+1)Individual as optimal solution, terminate calculating, Otherwise put t and increase 1, and go to step 4 and continue executing with.

The present invention compared to existing technology, has the advantage that

The present invention proposes a kind of logistics collaboration processing method based on cloud computing, based on the big data analysis of user, is protecting While card meets user's subjectivity demand, logistic resources can be optimized again, farthest improve the satisfaction of user, obtain Double benefit economical and environmentally friendly.

Accompanying drawing explanation

Fig. 1 is the flow chart of logistics collaboration processing method based on cloud computing according to embodiments of the present invention.

Detailed description of the invention

Hereafter provide retouching in detail one or more embodiment of the present invention together with the accompanying drawing of the diagram principle of the invention State.Describe the present invention in conjunction with such embodiment, but the invention is not restricted to any embodiment.The scope of the present invention is only by right Claim limits, and the present invention contains many replacements, amendment and equivalent.Illustrate in the following description many details with Thorough understanding of the present invention is just provided.These details are provided for exemplary purposes, and without in these details Some or all details can also realize the present invention according to claims.

An aspect of of the present present invention provides a kind of logistics collaboration processing method based on cloud computing.Fig. 1 is according to the present invention The logistics collaboration process flow figure based on cloud computing of embodiment.

The interest purpose of Behavior law and user in order to analyze user more accurately, knowing of the big data that the present invention combines Know discovery and go digging user historical data, it was predicted that the optimum delivery availability of user, thus reduce entreprise cost and provide the user Preferably service.The present invention considers that view rate, incidence relation between concern time and sales volume are to predict sales volume simultaneously.If sales volume It is respectively A and B for m, view rate and concern time, chooses a large number of users browsing data, represent with a and browse and the concern time is more than There is the degree of association between sales volume in predetermined threshold value, according to first checking method, excavate and produce the candidate meeting support C minimum simultaneously Collection, obtains degree of association a_min.Then by formula m=a_minT (A ∩ B) calculates user's request amount m.T (A ∩ B) | represent event A and event B concurrent number.Wherein support C is T (A ∩ B) |/T | (A) |, T | (A) | represents the sum that event A occurs.

The position coordinates of user is expressed as (x_n, y_n, t_n), (x_n, y_n) represent customer location transverse and longitudinal coordinate, t_nRepresent user The time data of relevant position.First, according to the position data of user, user is carried out segmentation at the time of one day, will after segmentation All positional distances are not more than the position of setting minimum range and cluster, and its position is changed the position consistent with cluster centre Coordinate, then recycles the behavior of Association Rule Analysis user.After data mining, obtain the relatedness ratio of user's time period Relatively, the time period selecting relatedness the highest delivers.As drawn multiple time period, then according to the delivery address choice of user's order The time period nearest from user's ship-to delivers.

The logistics service of the present invention is to be made up of user terminal, cloud platform electricity business, shipper Three-level Supply Chain.Cloud platform electricity Business receives order and the information of user, according to requirement and the information of forecasting of user, selects to best suit the shipper that user requires Service.Shipper quantity is represented with n；A_ijRepresent the user j evaluation of estimate to shipper i；q_iRepresent sending out of i-th shipper Goods amount；c_ijRepresent the shipper i cost of serving to j；p_ijRepresent the shipper i price to user j；Q_ijRepresent that shipper i is to j Service quality；p_jRepresent the acceptable ceiling price of user j；Q_jRepresent service quality expected from user j；γ_jRepresent user j Attention rate to price change；θ_jRepresent the user j attention rate to service quality；H_iRepresent the profit of shipper；S_iRepresent delivery Shang Suoneng participates in the minimum profit of service；x_iRepresent shipper i maximum service ability；c_iRepresent the price that product is provided；

Introducing user's evaluation of estimate to shipper, expression formula is:

$A_{ij}=(1+(p_j-p_{ij})/p_j){(Q_{ij}/Q_j)_{{\mathrm{\γ}}_j}}^{{\mathrm{\θ}}_j}$

To optimize the cloud platform electricity minimum target of business's cost of serving, in the situation that guarantee user satisfaction and service quality are high Under, user's cost of serving can reduce accordingly, and user's cost of serving is made up of price and the penalty of each shipper.Its mesh Scalar functions is:

$\min Z=\underset{i=1}{\overset{n}{\Σ}}\underset{j=1}{\overset{m}{\Σ}}{y}_{ij}{p}_{ij}+c^{\′}{{}_i}_j-\underset{i=1}{\overset{n}{\Σ}}{q}_i{c}_i$

And have

$A_{ij}=(1+(p_j-p_{ij})/p_j){(Q_{ij}/Q_j)_{{\mathrm{\γ}}_j}}^{{\mathrm{\θ}}_j}$

A is made when there is k_ijMore than A_kjTime, y_ij=1, wherein i ≠ k,

Otherwise y_ij=0

When t is more than T_jTime c '_ij=-π (t-T_j)p_ij,

Otherwise c '_ij=0

M=a_minT(A∩B)

$q_i=\underset{j=1}{\overset{m}{\Σ}}{y}_{ij}$

$H_i=\underset{j=1}{\overset{m}{\Σ}}{y}_{ij}(p_{ij}-c_{ij})$

$\underset{i=1}{\overset{n}{\Σ}}{q}_i=m$

q_i≤x_i

p_ij≤p_j

γ_j, θ_j∈[0,1]

T represents the time of actual service, T_jRepresent the desired optimum delivery availability of user drawn based on big data prediction. π represents time-out penalty coefficient.

For the derivation algorithm of above-mentioned model, time slice data mining, cluster and sequential correlation are excavated and combines The optimum delivery availability of prediction user.Algorithm flow is as follows

Step 1: prediction user's optimum delivery availability, input time, sequence, obtained segmentation matrix Q_ij.Calculate each Q_ij's Quantity, and utilize degree of association formula to calculate all adjacent Q_ijBetween relatedness.And find out the Q meeting condition_ij。

Step 2: stochastic generation initial solution Ω₍₀₎.The scale determining initial population is R, and crossover probability is P_j, mutation probability For P_bWith termination evolutionary criterion, arranging evolutionary generation enumerator t is 0.

Step 3: calculate individual fitness, using the inverse of target function value minZ as individual fitness value.

Step 4: from Ω_(t)Middle utilization selection opertor selects N/2 to parent, and wherein N > R, according to probability P_jPerform X-type Become individual in the middle of N number of.N number of middle individuality is carried out independently, according to probability P_bPerform variation, form N number of candidate individual.From time Select in individuality and select R individual composition a new generation population Ω according to ideal adaptation degree_(t+1)。

Step 5: as met end condition set in advance, then export Ω_(t+1)Individual as optimal solution, terminate calculating, Otherwise put t and increase 1, and go to step 4 and continue executing with.

Further, in terms of to vehicle scheduling in, for reaching economic and environment-friendly target, the present invention receives and dispatches having simultaneously The user of goods demand delivers, and is modeled planning vehicle route with the problem of the minimum target of vehicle oil consumption.Then set Meter solves the chromosome algorithm of this problem.

If vehicle fleet is K, have L type of vehicle, n_lFor the quantity of type of vehicle l, meetVehicle The vehicle-mounted capacity of type l is Q_l, oil consumption coefficient is a_l、b_l, n initial user is carried out shipping and receiving, the demand of receiving of user i For p_i, delivery demand amount is d_i, D_ijDistance for user node i to user node j.The optimization aim of model is so that vehicle oil Consumption minimum, before modeling, uses the computational methods of following oil consumption.

F_ij ^l=D_ij(a_l·(p_ij+q_ij)+b_l)

Wherein: F_ijFor the vehicle of l type from node i to the oil consumption of node j；p_ijThe vehicle of node j has been gone to for node i The goods total amount of the user accessed；q_ijThe goods total amount of the vehicle loading of node j is gone to for node i；(p_ij+q_ij) be node i before Total useful load toward the vehicle of node j.

The model of Vehicle Routing Problems is as follows:

Object functionAnd have and retrain as follows:

p_ij+q_ij≤Q_lx_ijk

x_ijk=1 or 0i, j, k ∈ V

∑_r∈Vp_jr-∑_i∈Vp_ij=p_j i,j∈V

∑_r∈Vq_jr-∑_i∈Vq_ij=d_j i,j∈V

$\begin{array}{cc}\underset{i=1}{\overset{n}{\Σ}}{p}_{i0}=\underset{i=1}{\overset{n}{\Σ}}{p}_i& \underset{j=1}{\overset{n}{\Σ}}{q}_{0j}=\underset{j=1}{\overset{n}{\Σ}}{d}_j\end{array}$

$K=\underset{l=1}{\overset{L}{\Σ}}{n}_i$

p_ij≥p_ix_ijk i,j∈V

q_ij≥d_jx_ijk i,j∈V

Wherein V represents the set of all user nodes.

The solution strategies of problem is as follows:

(1) randomly generate the population comprising multiple quantum chromosomes, be the problem of L for number of users, for each amount Daughter chromosome, uses the three-dimensional quantum bit matrix of L × L × 2 to represent, random initializtion quantum bit produces 0 and 1 random number, so The rear probability α that quantum bit is 0 and 1 being assigned to correspondence respectively_ijAnd β_ij。

(2) first produce the random number between [0,1], produce the two-dimensional observation matrix of L × L, and it is every to adjust matrix guarantee Row each column all only has the abscissa at 11,1 place to represent the order of service, the user corresponding to vertical coordinate representative.Forming car During path, first randomly choose a type of vehicle, when this vehicle cannot meet next user's request, then select at random Lower a kind of type of vehicle, when the vehicle number of selected type exceedes existing vehicle number, repeats to choose type of vehicle, until selected Type of vehicle is less than the type vehicle number.

(3) after user's shipping and receiving order and vehicle route distribute, near insertion approach is used to optimize local route Sequence, flow process is as follows:

3.1 take the logistics center 0 starting point as path；

3.2 find first user node i so that logistics center to user node i returns again to the vehicle oil of logistics center Consumption minimum, constitutes local path 0-i-0；

3.3 for the local path formed, and in the user node being not belonging to this path, finds pre-selected users node Collection Q, the point in set Q meets the closest of the point on path, and shipping and receiving weight is the lightest；

3.4 circulations take the user node that Q concentrates, and insert Q respectively and concentrate user node so that the road being newly formed on path The vehicle oil consumption that footpath produces is minimum；That is, the user node k concentrated as distribution Q inserts arc (i, j) vehicle oil during position in path Consuming minimum, (i, j) path of position is as new route just user node k to be inserted arc；

3.5. step 3.1-3.4 is repeated, until all of user is accessed.In the path obtained, for any two joints Point i, j, it may be judged whether meet with (i, j), (i+1, j+1) replace (i, i+1), (j, j+1) total oil consumption afterwards lower, the most then hold The above-mentioned replacement of row.

(4) after each chromosome to population decodes, respectively according to above object function minima minU；Make chromosome Fitness function is U, solves the fitness of homologue, it is simply that will solve the oil consumption of every chromosome, every chromosome is again It is made up of the shipping and receiving of mulitpath, after vehicle distributes path, for each two point on path, by the loading capacity of every 2 Obtain the oil consumption of each point-to-point transmission, thus obtain the oil consumption of every paths, the finally oil consumption of every chromosome.

Comprehensive both examples above, below will use particle cluster algorithm generation with two object function Z and U as general objective For above solution procedure, solve globally optimal solution.

Use the three dimensional particles coded method rounded based on particle position, if user node has n, the first dimension user node Numbering is used for being numbered user；Second dimension particle position X_iIt is used for being ranked up, first Wesy's family node serial number with really Determine the sequencing of user's dispensing；Third dimension particle position (Y_i) (and 0, n) in the range of use the carry side of rounding after stochastic generation Method determines the path of K vehicle；During decoding, user is resequenced by the size being first according to the second dimension particle position, Then third dimension particle position is carried out carry floor operation.

Third dimension particle position (Y_i) be (0, n) in the range of use after stochastic generation carry to round method to determine vehicle Path, if Y during particle iteration updates_iBeyond (0, n) scope, then it is carried out particle correction, to ensure more Particle after Xin remains as feasible solution.The method of correction is: if third dimension particle position (Y after particle renewal_i) beyond (0, n) Scope, then carry out (0, n) scope stochastic generation again, is corrected with this to it.

The flow process that solves of the particle cluster algorithm of Bi-objective optimizing scheduling based on user's cost of serving Z and vehicle oil consumption U is:

1. pair basic parameter is configured, including the inertia weight at the end of Population Size, initial and iteration, study because of Son, constraint of velocity and maximum iteration time, determine the weights of two target updates；

2. using user node data, vehicle data and correlation coefficient as input；When initialization of population according to setting The initialized particle populations of Population Size stochastic generation, initialize each particle the second peacekeeping third dimension particle position and Particle rapidity, initializes individual optimal particle and global optimum's particle；

3. calculating user's cost of serving that in population, each particle is corresponding, the individuality initializing all particles is optimum, and from All of individual optimum finds out the minimum path of vehicle oil consumption to initialize global optimum；

Solve the adaptive value of particle after each renewal the most successively, if the weighted sum of user's cost of serving and oil consumption after Geng Xining It is better than preceding value, then just update current individual optimal value.Then, all individual optimal particle after updating is found out The figure of merit, contrasts with current global optimum, if new optimal value weighted sum is more excellent, is then updated to the current overall situation Excellent.

5. the speed of each particle of renewal:

V_ij=c₁×random₁×(B_ij-X_ij)+c₂×random₂×(G_j-X_ij)

Wherein random₁And random₂The random number being between 0-1, V_ij, X_ijIt is respectively i-th particle at jth dimension sky Speed between and position；B_ijFor the current optimal location of this particle, G_jFor the current optimal location of population；

Update the position of each particle:

X_ij=X_ij+V_ij

Particle position is judged simultaneously, the particle of the constraint that goes beyond the scope is corrected, it is ensured that the particle after renewal It is still that feasible solution；

6. after updating, iterations is increased by 1 every time, be used for carrying out the judgement of algorithm end condition；If reaching maximum Iterations, then algorithm terminates, and inputs current global optimum, it is possible to obtain optimal case, otherwise, proceeds iteration, weight Multiple step 4 and 5.

Additionally, Bi-objective weights give different weights to two optimization aim, it is weighted summation.Assume certain particle User's cost of serving and the oil consumption of the i-th result searched are followed successively by Z_iAnd U_i, user's cost of serving of jth result and oil Consumption is followed successively by Z_jAnd U_j, the current individual optimum of this particle is followed successively by: optimal cost difference PG, optimum oil consumption PT, workload difference Weights be w1, the weights of operation completion date are w2, then particle individuality update Bi-objective more new regulation can be expressed as:

w1*PG+w2*TG>w1**abs(Z_i-Z_j)+w2*max{U_i,U_j}

Wherein, abs () represents the function that takes absolute value.If meeting above formula, then individuality currently optimum is updated to successively:

PG=abs (Z_i-Z_j)

TG=max{U_i,U_j}；

If being unsatisfactory for above formula, then will retain individual current optimum；

Global optimum is found out from the current individual optimum of all particles.

In sum, the present invention proposes a kind of logistics collaboration processing method based on cloud computing, based on the big data of user Analyze, while ensureing to meet user's subjectivity demand, logistic resources can be optimized again, farthest improve expiring of user Meaning degree, achieves double benefit economical and environmentally friendly.

Obviously, it should be appreciated by those skilled in the art, each module of the above-mentioned present invention or each step can be with general Calculating system realize, they can concentrate in single calculating system, or be distributed in multiple calculating system and formed Network on, alternatively, they can realize with the executable program code of calculating system, it is thus possible to by they store Performed by calculating system within the storage system.So, the present invention is not restricted to the combination of any specific hardware and software.

It should be appreciated that the above-mentioned detailed description of the invention of the present invention is used only for exemplary illustration or explains the present invention's Principle, and be not construed as limiting the invention.Therefore, that is done in the case of without departing from the spirit and scope of the present invention is any Amendment, equivalent, improvement etc., should be included within the scope of the present invention.Additionally, claims purport of the present invention Whole within containing the equivalents falling into scope and border or this scope and border change and repair Change example.

Claims (2)

1. a logistics collaboration processing method based on cloud computing, it is characterised in that including:

Cloud platform electricity business accepts the sequence information of user, minimizes target selection shipper according to user's cost of serving and transports Service；

Solve above-mentioned target, in conjunction with the user's statistical data obtained in advance, it was predicted that for the optimal time of each user's delivery.

Method the most according to claim 1, it is characterised in that described user's cost of serving minimizes target and user serviced Cost statement is shown as price and the penalty composition of each shipper, and its object function is:

$\min Z=\underset{i=1}{\overset{n}{\Σ}}\underset{j=1}{\overset{m}{\Σ}}{y}_{ij}{p}_{ij}+{c^{\′}}_{ij}-\underset{i=1}{\overset{n}{\Σ}}{q}_i{c}_i$

And meet following constraint

$A_{\mathrm{ij}}={(1+(p_j-p_{\mathrm{ij}})/p_j)}^{{\mathrm{\γ}}_j}{(Q_{\mathrm{ij}}/Q_j)}^{{\mathrm{\θ}}_j}$

A is made when there is k_ijMore than A_kjTime, y_ij=1, wherein i ≠ k,

Otherwise y_ij=0

When t is more than T_jTime c '_ij=-π (t-T_j)p_ij,

Otherwise c '_ij=0

M=a_minT(A∩B)

$q_i=\underset{j=1}{\overset{m}{\Σ}}{y}_{ij}$

$H_i=\underset{j=1}{\overset{m}{\Σ}}{y}_{ij}(p_{ij}-c_{ij})$

$\underset{i=1}{\overset{n}{\Σ}}{q}_i=m$

q_i≤x_i

p_ij≤p_j

γ_j, θ_j∈[0,1]；

Wherein t represents the time of actual service, T_jRepresent the desired optimum delivery availability of user drawn based on big data prediction；π Represent time-out penalty coefficient；N represents shipper quantity；A_ijRepresent the user j evaluation of estimate to shipper i；q_iRepresent i-th delivery The delivering amount of business；c_ijRepresent the shipper i cost of serving to j；p_ijRepresent the shipper i price to user j；Q_ijRepresent delivery The business i service quality to j；p_jRepresent the acceptable ceiling price of user j；Q_jRepresent service quality expected from user j；γ_jRepresent The user j attention rate to price change；θ_jRepresent the user j attention rate to service quality；H_iRepresent the profit of shipper；S_iTable Show that shipper can participate in the minimum profit of service；x_iRepresent shipper i maximum service ability；c_iRepresent the price that product is provided； M is sales volume, and view rate and concern time are respectively A and B, choose a large number of users browsing data, represent view rate and pass with amin There is the degree of association between sales volume more than predetermined threshold value in the note time, according to first checking method, excavate generation and meet support C simultaneously Minimum Candidate Set, obtains degree of association a_min；Then by formula m=a_minT (A ∩ B) calculates user's request amount m；T (A ∩ B) | represent Event A and event B concurrent number；Wherein support C is T (A ∩ B) |/T | (A) |, T | (A) | represents that event A occurs Sum；

The described optimal time being predicted as each user's delivery, farther includes:

Step 1: prediction user's optimum delivery availability, input time, sequence, obtained segmentation matrix Q_ij；Calculate each Q_ijQuantity, And utilize degree of association formula to calculate all adjacent Q_ijBetween relatedness, and find out the Q meeting condition_ij；

Step 2: stochastic generation initial solution Ω₍₀₎；The scale determining initial population is R, and crossover probability is P_j, mutation probability is P_bWith Terminating evolutionary criterion, arranging evolutionary generation enumerator t is 0；

Step 3: calculate individual fitness, using the inverse of target function value minZ as individual fitness value；

Step 4: from Ω_(t)Middle utilization selection opertor selects N/2 to parent, and wherein N > R, according to probability P_jExecution intersects to form N number of Middle individual；N number of middle individuality is carried out independently, according to probability P_bPerform variation, form N number of candidate individual；From candidate Body selects R individual composition a new generation population Ω according to ideal adaptation degree_(t+1)；

Step 5: as met end condition set in advance, then export Ω_(t+1)Individual as optimal solution, terminate calculating, otherwise Put t and increase 1, and go to step 4 and continue executing with.