CN110619424A - Transportation and distribution optimizing system - Google Patents

Abstract

The invention relates to the field of transportation, in particular to a transportation distribution optimization system, which comprises a transportation base site selection module, a transportation loading module and a transportation planning module; the transportation base site selection module is used for selecting a transportation base to meet transportation requirements; the transportation loading module is used for loading the goods into the standard container as much as possible to meet the transportation requirement; the transportation planning module is used for scheduling transportation tools to the maximum extent; the transportation base site selection module provided by the invention fully considers the geographical characteristics of different areas, and widens the application range; the transportation loading module provided by the invention provides a mixed simulated annealing algorithm based on loading block sequences for multi-dimensional multi-backpack NP problems, and approaches to an accurate solution; various modules are organically combined to achieve the comprehensive purpose that 1+1 is larger than 2.

Description

Transportation and distribution optimizing system

Technical Field

The invention relates to the field of transportation, in particular to a transportation distribution optimization system.

Background

In the prior art, the transportation and distribution systems can be divided into: the system comprises a transportation base site selection module, a transportation loading module and a transportation planning module.

In the base site selection module, clustering analysis is often performed on a plurality of base positions, a distance matrix is directly constructed, and then a range is selected. However, the selection of the range is too subjective, the classification is not clear, the site selection is lack of persuasiveness, and the method is not easy to be applied to more practical situations.

In the transportation loading module, the technology of carrying out loading operation by 'constructing a load balance function and forming a cargo unit' is more efficient at present. The origin of this technique is as follows:

LEI Dingyou,HONG Shuhua,ZHANG Yinggui.Model and algorithm for container mixed balanced loading of light and heavy cargo.ComputerEngineering and Applications.

in the technology, the distance of the total gravity center of the cargo deviating from the geometric center of the container is used for representing load balance constraint, a load balance function is used for describing the relation between the total gravity center position of the cargo and the maximum allowable weight of the cargo, and a mixed balance loading model of light and heavy cargos of the container is established; and classifying the goods to be loaded into light and heavy goods units, constructing a central framework by using the heavy goods units to fix the total gravity center position, and selecting the light goods units by using an evaluation function to obtain a complete loading scheme. However, the technology has the following defects: firstly, the loading effect cannot be visually output and only can be used for loading reference in actual application; secondly, the utilization rate of the container load capacity is not considered; third, connectivity to other modules of the invention is poor.

Disclosure of Invention

The invention provides a transportation and distribution optimization system, aiming at solving the problems that in the prior art, the selection of an address range in a transportation and distribution system is too subjective, the output is not intuitive enough, the utilization rate of the container load capacity and the poor connection of module quality inspection are not considered.

In order to solve the technical problems, the technical scheme of the invention is as follows:

a transportation distribution optimization system comprises a transportation base site selection module, a transportation loading module and a transportation planning module;

the transportation base site selection module is used for selecting a transportation base to meet transportation requirements, obtaining the position of the transportation base by adopting clustering analysis, solving the linear programming problem of a transportation line by using an immune algorithm, and transmitting the obtained data to the transportation programming module;

the transportation loading module is used for loading the goods into the standard container as much as possible to meet the transportation requirement; obtaining an optimal solution for loading goods by adopting a combination of a simulated annealing algorithm and a multi-dimensional multi-backpack problem, and transmitting data of the optimal solution to a transportation planning module;

the transportation planning module is used for scheduling transportation tools to the maximum extent, acquiring a transportation plan and a transportation route by using a physical model based on a greedy algorithm according to data obtained from the transportation base site selection module and the transportation loading module, and outputting a final obtained result.

Preferably, the concrete steps of the transportation base site selection module for cluster analysis are as follows:

gr and Gs are defined as two classes of cities, the shortest distance between them being

D₁＝min{d_ij|i∈G_r,j∈G_s} (1)

Wherein d is_ijThe distance between the ith city in the Gr class and the jth city in the Gs class;

define the longest distance between class Gr and class Gs as

D₂＝max{d_ij|i∈G_r,j∈G_s} (2)

Define the mean squared distance between class Gr and class Gs as

Wherein nr and ns are the number of samples contained in Gr and Gs, respectively;

defining the maximum similarity coefficient between class Gr and class Gs as

R₁＝max{r_ij|i∈G_r,j∈G_s} (4)

The minimum similarity coefficient is R₂＝min{r_ij|i∈G_r,j∈G_s}，

Define the Euclidean distance formula between two points as

Wherein (x)_i,y_i,z_j) Is the coordinate of city i, (x)_j,y_j,z_j) Is the coordinate of city j;

wherein rij is a similarity coefficient between the ith city in the class Gr and the jth city in the class Gs;

the method comprises the following specific steps:

step S1: h city classes are constructed, and each city class only contains one object; in an initial state, one object only comprises one city;

step S2: the distance between every two objects is calculated according to a two-point Euclidean distance formula (5) and is marked as d_ijI, j ═ 1,2, …, h), and a distance symmetric array is derived:

step S3: at a distance matrix D⁽⁰⁾The minimum value is found by the middle inquiry and is recorded asSubject i₁And j₁Merging into a new city class, and marking the new city class as the h +1 th city class; at the same time, the original i-th₁，j₁The classes are eliminated, and the total number of the city classes is reduced by 1 due to the merging process, namely the total number of the current city classes is changed into h-1;

step S4: calculating the distances between the new city class and all other city classes according to a two-point Euclidean distance formula (5), wherein the distances between all other city classes are unchanged, and obtaining a first-order-reduced new distance matrix:

representing the new distance matrix after one combination, and reducing the new distance matrix to (h-1) x (h-1) order;

repeating the steps S3 and S4 until the number of the remaining classes is 1 or the distance is higher than the class critical value;

step S5: drawing a cluster map;

step S6: the number and class of classes are determined.

Preferably, the concrete steps of the transportation base site selection module using the immune algorithm to solve the linear planning problem of the transportation route are as follows:

in order to calculate the decision class and the number of classes after the cluster analysis, a threshold is defined, and the threshold is defined as:

wherein S represents the area of the region, L represents the length of a diagonal line of the region, and K is calculated according to the perimeter of the earth and the property of the vehicle;

the following linear programming model is established:

is limited to

i,j∈N.

N ═ {1,2, 3.., N } is the set of candidate cities, F is the sum of the distances between all cities;

because the number of candidate cities is large, the linear programming problem is calculated by adopting an immune algorithm to obtain an optimal solution, and the method comprises the following steps:

step 101: defining an optimization problem of linear planning of a transportation route;

step 102: generating an initial optimal solution of an optimization problem based on city combination;

step 103: evaluating the affinity between the optimization problem and the initial city, and measuring the quality of the optimal solution through the affinity;

step 104: according to the affinity, selecting n optimal cities, wherein the optimal cities meet the objective functionThe optimal solution of (2);

step 105: the n optimal cities are cloned to generate a temporary clone set C, the C comprises the feasible solution optimization problem, the affinity of the optimization problem and the initial city is improved, and the clone set C is used for optimizing the feasible solution;

step 106: generating a new city through immune selection based on n optimal cities and the clone set C to obtain a new feasible solution;

step 107: generating a memory set M through the feasible solutions, wherein M is a set of all the feasible solutions, and the optimal solution is contained in the set of the feasible solutions;

step 108: generating new cities by immune selection and based on the set M, the new cities being generated by modifying some feasible city group in M or one city;

step 109: and checking and analyzing whether the new city meets the requirement of an actual problem, wherein the actual problem requirement is a limiting condition except an objective function for the site of the selected site, namely a condition except the shortest Euclidean distance.

Preferably, the simulated annealing algorithm adopted by the transportation loading module comprises the following specific steps:

suppose there are N objects and a backpack with a volume of V and a capacity of U, ith^thThe maximum number of articles is n [ i ]]Each object having a volume a [ i ]]Weight of each object b [ i ]]The value of each item is w [ i ]]To maximize the value of the items placed in the bag, the transformation function of the problem is:

f(i，u，v)＝max{f(i-1，v，u)，f(i-1，v-l*a[i]，u-l*b[i])+l*w[i]}

l is an integer between 1 and n [ i ];

obtaining the solution by applying a hybrid simulated annealing algorithm;

order: i ═ 1,2, ·, N } a set of all shipments; v_iCargo volume i (i)∈I)；

CV is the volume of the container;

current loading sequence adaptation value (loading rate):

solving the maximum value of the loading rate as:

O＝max{g(B_list)}

the concrete steps of solving are as follows:

step S201: inputting the set I, setting the cooling speed eta, 0<η<1. Maximum number of iterations n and cooling target t_goal；

Step S202: solving a loading block B;

step S203: according to B, randomly generating a sequence B_list(ii) a Setting the initial temperature t>t_goalSimultaneously, let k equal to 1;

step S204: solving the adaptive value G-G (B)_list)；

Step S205: random Generation of New Loading sequence B'_list(ii) a The recipe value G ' ═ G (B ' was calculated using the recipe subroutine '_list)；

Step S206: if G'>G, then order B_list＝B′_listOtherwise, let k be k + 1;

step S207: if k > n, executing step S208, otherwise, executing step S205;

step S208: cooling; setting t as t × η;

step S209: if t<t_goalUsing a loading scheme subroutine and based on B_listGet the loading scheme and update O to max { g (B)_list) }; otherwise, step S204 is executed.

Preferably, the specific step of solving the loading block B in step S202 is as follows:

step S2021: the length, width and height of the input goods.

Step S2022: root of herbaceous plantArranging goods according to the bottom area of the goods to obtain a loading sequence I_list(ii) a Initializing iI ═ 1 and iB ═ 1;

step S2023: placing iI^th) Cargo entering container iB while ordering I_list＝I_list{ iI }, wherein iI^th(iI∈I_list)；

Step S2024: selecting goods jI with the largest bottom area and the goods meeting constraint conditions;

step S2025: if jI exists, put on the existing cargo in iB and make I_list＝I_list{ jI }; if jI does not exist, go to step S2026;

step S2026: if I_listFor empty set, stop inputting, if I_listNot empty set, iB ═ iB + 1; and executes step S2023;

preferably, the specific steps for solving the loading scheme are as follows:

step S301: input loading sequence B_listAnd the identity of the cargo and container and let iB equal 1;

step S302: selection B_listiB of (1)^thA cargo; obtaining a current loading point chart EP; calculating the current iB based on the table EP^thA load point value for the cargo;

step S303: if all the load point values are less than 0, executing step S4; otherwise, the loading point in the table EP having the maximum loading point value and its correspondence with the goods are selected as the iB^thThe loading position and the loading direction of the goods;

step S304: obtaining a new loading scheme by using the current loading block;

step S305: let iB ═ iB + 1;

step S306: if iB>length(B_list) Outputting the result; otherwise, step S302 is performed.

Compared with the prior art, the technical scheme of the invention has the beneficial effects that:

the transportation base site selection module provided by the invention defines the selection of the threshold value to fully consider the geographical characteristics of different areas, thereby widening the application range; the transportation loading module provided by the invention provides a mixed simulated annealing algorithm based on loading block sequences for multi-dimensional multi-backpack NP problems, and approaches to an accurate solution; various modules are organically combined to achieve the comprehensive purpose that 1+1 is larger than 2.

Drawings

FIG. 1 is a system block diagram of the present invention.

FIG. 2 is a block diagram of a transport distribution optimization system provided by the present invention;

FIG. 3 is a diagram of the site selection results of a medical base, using the disaster response system of Podocis as an example;

FIG. 4 is (a) and (b) a top view and a side view of a dry cargo container of the International Standards Organization (ISO) standard in example 2;

FIG. 5 is a top view of the dry cargo container two-loading of the International Standards Organization (ISO) standard in example 2;

Detailed Description

The drawings are for illustrative purposes only and are not to be construed as limiting the patent;

for the purpose of better illustrating the embodiments, certain features of the drawings may be omitted, enlarged or reduced, and do not represent the size of an actual product;

it will be understood by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted.

The technical solution of the present invention is further described below with reference to the accompanying drawings and examples.

Example 1

As shown in fig. 1, a transportation distribution optimization system includes a transportation base site selection module, a transportation loading module and a transportation planning module;

the transportation base site selection module is used for selecting a transportation base to meet transportation requirements, obtaining the position of the transportation base by adopting clustering analysis, solving the linear programming problem of a transportation line by using an immune algorithm, and transmitting the obtained data to the transportation programming module;

the transportation loading module is used for loading the goods into the standard container as much as possible to meet the transportation requirement; obtaining an optimal solution for loading goods by adopting a combination of a simulated annealing algorithm and a multi-dimensional multi-backpack problem, and transmitting data of the optimal solution to a transportation planning module;

the transportation planning module is used for scheduling transportation tools to the maximum extent, acquiring a transportation plan and a transportation route by using a physical model based on a greedy algorithm according to data obtained from the transportation base site selection module and the transportation loading module, and outputting a final obtained result.

As a preferred embodiment, the concrete steps of the transportation base addressing module for cluster analysis are as follows:

gr and Gs are defined as two classes of cities, the shortest distance between them being

D₁＝min{d_ij|i∈G_r,j∈G_s} (1)

Wherein d is_ijThe distance between the ith city in the Gr class and the jth city in the Gs class;

define the longest distance between class Gr and class Gs as

D₂＝max{d_ij|i∈G_r,j∈G_s} (2)

Define the mean squared distance between class Gr and class Gs as

Wherein nr and ns are the number of samples contained in Gr and Gs, respectively;

defining the maximum similarity coefficient between class Gr and class Gs as

R₁＝max{r_ij|i∈G_r,j∈G_s} (4)

The minimum similarity coefficient is R₂＝min{r_ij|i∈G_r,j∈G_s}，

Define the Euclidean distance formula between two points as

Wherein (x)_i,y_i,z_i) Is the coordinate of city i, (x)_j,y_j,z_j) Is the coordinate of city j;

wherein rij is a similarity coefficient between the ith city in the class Gr and the jth city in the class Gs;

the method comprises the following specific steps:

step S1: h city classes are constructed, and each city class only contains one object; in an initial state, one object only comprises one city;

step S2: the distance between every two objects is calculated according to a two-point Euclidean distance formula (5) and is marked as d_ijI, j ═ 1,2, …, h), and a distance symmetric array is derived:

step S3: at a distance matrix D⁽⁰⁾The minimum value is found by the middle inquiry and is recorded asSubject i₁And j₁Merging into a new city class, and marking the new city class as the h +1 th city class; at the same time, the original i-th₁，j₁The classes are eliminated, and the total number of the city classes is reduced by 1 due to the merging process, namely the total number of the current city classes is changed into h-1;

step S4: calculating the distances between the new city class and all other city classes according to a two-point Euclidean distance formula (5), wherein the distances between all other city classes are unchanged, and obtaining a first-order-reduced new distance matrix:

representing the new distance matrix after one combination, and reducing the new distance matrix to (h-1) x (h-1) order;

repeating the steps S3 and S4 until the number of the remaining classes is 1 or the distance is higher than the class critical value;

step S5: drawing a cluster map;

step S6: the number and class of classes are determined.

As a preferred embodiment, the transportation base site selection module uses an immune algorithm to solve the linear programming problem of the transportation line

The specific steps of the subject are as follows:

in order to calculate the decision class and the number of classes after the cluster analysis, a threshold is defined, and the threshold is defined as:

wherein S represents the area of the region, L represents the length of a diagonal line of the region, and K is calculated according to the perimeter of the earth and the property of the vehicle;

the following linear programming model is established:

is limited to

i,j∈N.

N ═ {1,2, 3.., N } is the set of candidate cities, F is the sum of the distances between all cities;

because the number of candidate cities is large, the linear programming problem is calculated by adopting an immune algorithm to obtain an optimal solution, and the method comprises the following steps:

step 101: defining an optimization problem of linear planning of a transportation route;

step 102: generating an initial optimal solution of an optimization problem based on city combination;

step 103: evaluating the affinity between the optimization problem and the initial city, and measuring the quality of the optimal solution through the affinity;

step 104: according to the affinity, selecting n optimal cities, wherein the optimal cities meet the objective functionThe optimal solution of (2);

step 105: the n optimal cities are cloned to generate a temporary clone set C, the C comprises the feasible solution optimization problem, the affinity of the optimization problem and the initial city is improved, and the clone set C is used for optimizing the feasible solution;

step 106: generating a new city through immune selection based on n optimal cities and the clone set C to obtain a new feasible solution;

step 107: generating a memory set M through the feasible solutions, wherein M is a set of all the feasible solutions, and the optimal solution is contained in the set of the feasible solutions;

step 108: generating new cities by immune selection and based on the set M, the new cities being generated by modifying some feasible city group in M or one city;

step 109: and checking and analyzing whether the new city meets the requirement of an actual problem, wherein the actual problem requirement is a limiting condition except an objective function for the site of the selected site, namely a condition except the shortest Euclidean distance.

As a preferred embodiment, the steps of the adopted simulated annealing algorithm of the transportation loading module are as follows:

suppose there are N objects and a backpack with a volume of V and a capacity of U, ith^thThe maximum number of articles is n [ i ]]Each object having a volume a [ i ]]Weight of each object b [ i ]]The value of each item is w [ i ]]To maximize the value of the items placed in the bag, the transformation function of the problem is:

f(i，u，v)＝max{f(i-1，v，u)，f(i-1，v-l*a[i]，u-l*b[i])+l*w[i]}

l is an integer between 1 and n [ i ];

obtaining the solution by applying a hybrid simulated annealing algorithm;

order: i ═ 1,2, ·, N } a set of all shipments; v_iCargo volume I (I ∈ I);

CV is the volume of the container;

current loading sequence adaptation value (loading rate):

solving the maximum value of the loading rate as:

O＝max{g(B_list)}

the concrete steps of solving are as follows:

step S201: inputting the set I, setting the cooling speed eta, 0<η<1. Maximum number of iterations n and cooling target t_goal；

Step S202: solving a loading block B;

step S203: according to B, randomly generating a sequence B_list(ii) a Setting the initial temperature t>t_goalSimultaneously, let k equal to 1;

step S204: solving the adaptive value G-G (B)_list)；

Step S205: random Generation of New Loading sequence B'_list(ii) a The recipe value G ' ═ G (B ' was calculated using the recipe subroutine '_list)；

Step S206: if G'>G, then order B_list＝B′_listOtherwise, let k be k + 1;

step S207: if k > n, executing step S208, otherwise, executing step S205;

step S208: cooling; setting t as t × η;

step S209: if t<t_goalUsing a loading scheme subroutine and based on B_listGet the loading scheme and update O to max { g (B)_list) }; otherwise, step S204 is executed.

As a preferred embodiment, the specific steps of solving the loading block B in step S202 are as follows:

step S2021: the length, width and height of the input goods.

Step S2022: arranging goods according to the bottom area of the goods to obtain a loading sequence I_list(ii) a Initializing iI ═ 1 and iB ═ 1;

step S2023: placing iI^th) Cargo entering container iB while ordering I_list＝I_list{ iI }, wherein iI^th(iI∈I_list)；

Step S2024: selecting goods jI with the largest bottom area and the goods meeting constraint conditions;

step S2025: if jI exists, put on the existing cargo in iB and make I_list＝I_list{ jI }; if jI does not exist, go to step S2026;

step S2026: if I_listFor empty set, stop inputting, if I_listNot empty set, iB ═ iB + 1;

and executes step S2023;

as a preferred embodiment, the specific steps for solving the loading scheme are as follows:

step S301: input loading sequence B_listAnd the identity of the cargo and container and let iB equal 1;

step S302: selection B_listiB of (1)^thA cargo; obtaining a current loading point chart EP; calculating the current iB based on the table EP^thA load point value for the cargo;

step S303: if all the load point values are less than 0, executing step S4; otherwise, the loading point in the table EP having the maximum loading point value and its correspondence with the goods are selected as the iB^thThe loading position and the loading direction of the goods;

step S304: obtaining a new loading scheme by using the current loading block;

step S305: let iB ═ iB + 1;

step S306: if iB>length(B_list) Outputting the result; otherwise, step S302 is performed.

Example 2

As shown in fig. 1-5, in the present embodiment, after the hurricane disaster of puerto rico, a transportable disaster response system can be designed to improve the response capability thereof by using the invention. The use of a rotorcraft provides prepackaged medical supplies that are shipped to designated medical bases for the purpose of providing adequate and timely response during or after a natural disaster.

Firstly, collecting geographic information and material demand information of the region;

and determining the site selection of the traditional Chinese medicine base in the disaster response system of puerto rich by using a transport base site selection module, taking the main traffic nodes on the map of puerto rich as candidate positions, wherein the distance between the two points is the linear geographic position distance. And (3) solving an optimal solution by using cluster analysis and an immune algorithm: base 1 and base 2. FIG. two is a diagram of the result of address selection;

determining the transportation routes of dry cargo containers of different unmanned aerial vehicles and International Standard Organization (ISO) standards by using a transportation planning module and considering the optimal solution of each step;

and (3) using a transportation loading module, respectively using the unmanned aerial vehicle as a backpack, using the medicine as goods, using the container as a backpack and using the unmanned aerial vehicle as goods, and successively performing hybrid simulated annealing algorithm operation. The loading schemes of dry cargo containers of different kinds of drones and International Standards Organization (ISO) standards are determined. The third, fourth and fifth are loading diagrams.

The method provided by the invention can generate a set of scheme for selecting the site of the material transportation and allocation center in a certain area, and can be used for allocating disaster relief materials, daily express delivery materials and other related activities; the three modules of the invention are independent from each other in operation, but are closely combined in actual operation, and are suitable for real social environment.

As shown in fig. 3, the site selection module of the transport base is used for determining the site selection of two medical bases (base 1 and base 2) in the disaster response system of puerto rich; the first table and the second table are respectively the loading scheme and the transportation plan of dry cargo containers of different unmanned aerial vehicles and International Standard Organization (ISO) standards;

TABLE 1

TABLE 2

Determining loading schemes of dry cargo containers of different unmanned aerial vehicles and International Standards Organization (ISO) standards by using a transport loading module;

using a transportation planning module, transportation routes for dry cargo containers of different kinds of drones and International Standards Organization (ISO) standards are determined.

The same or similar reference numerals correspond to the same or similar parts;

the terms describing positional relationships in the drawings are for illustrative purposes only and are not to be construed as limiting the patent;

it should be understood that the above-described embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. And are neither required nor exhaustive of all embodiments. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the claims of the present invention.

Claims (6)

1. A transportation distribution optimizing system is characterized by comprising a transportation base site selection module, a transportation loading module and a transportation planning module;

the transportation base site selection module is used for selecting a transportation base to meet transportation requirements, obtaining the position of the transportation base by adopting clustering analysis, solving the linear programming problem of a transportation line by using an immune algorithm, and transmitting the obtained data to the transportation programming module;

the transportation loading module is used for loading the goods into the standard container as much as possible to meet the transportation requirement; obtaining an optimal solution for loading goods by adopting a combination of a simulated annealing algorithm and a multi-dimensional multi-backpack problem, and transmitting data of the optimal solution to a transportation planning module;

the transportation planning module is used for scheduling transportation tools to the maximum extent, acquiring a transportation plan and a transportation route by using a physical model based on a greedy algorithm according to data obtained from the transportation base site selection module and the transportation loading module, and outputting a final obtained result.

2. The transportation distribution optimization system of claim 1, wherein the transportation site selection module performs cluster analysis by the following steps:

gr and Gs are defined as two classes of cities, the shortest distance between them being

D₁＝min{d_ij|i∈G_r,j∈G_s} (1)

Wherein d is_ijThe distance between the ith city in the Gr class and the jth city in the Gs class;

define the longest distance between class Gr and class Gs as

D₂＝max{d_ij|i∈G_r,j∈G_s} (2)

Define the mean squared distance between class Gr and class Gs as

Wherein nr and ns are the number of samples contained in Gr and Gs, respectively;

defining the maximum similarity coefficient between class Gr and class Gs as

R₁＝max{r_ij|i∈G_r,j∈G_s} (4)

The minimum similarity coefficient is R₂＝min{r_ij|i∈G_r,j∈G_s}，

Define the Euclidean distance formula between two points as

Wherein (x)_i,y_i,z_i) Is the coordinate of city i, (x)_j,y_j,z_j) Is the coordinate of city j;

wherein rij is a similarity coefficient between the ith city in the class Gr and the jth city in the class Gs;

the method comprises the following specific steps:

step S1: h city classes are constructed, each city class only comprises one object, and in an initial state, one object only comprises one city;

step S2: the distance between every two objects is calculated according to a two-point Euclidean distance formula (5) and is marked as d_ijI, j ═ 1,2, …, h), and a distance symmetric array is derived:

step S3: at a distance matrix D⁽⁰⁾The minimum value is found by the middle inquiry and is recorded asSubject i₁And j₁Merging into a new city class, and marking the new city class as the h +1 th city class; at the same time, the original i-th₁，j₁The classes are eliminated, and the total number of the city classes is reduced by 1 due to the merging process, namely the total number of the current city classes is changed into h-1;

step S4: calculating the distances between the new city class and all other city classes according to a two-point Euclidean distance formula (5), wherein the distances between all other city classes are unchanged, and obtaining a first-order-reduced new distance matrix:

representing the new distance matrix after the combination is performed once, reducing the new distance matrix to (h-1) x (h-1) order, and repeating the steps S3 and S4 until the number of the remaining classes is 1 or the distance is higher than the class critical value;

step S5: drawing a cluster map;

step S6: the number and class of classes are determined.

3. The transportation distribution optimization system of claim 2, wherein the transportation site location module uses an immune algorithm to solve the linear planning problem of the transportation route by the specific steps of:

in order to calculate the decision class and the number of classes after the cluster analysis, a threshold is defined, and the threshold is defined as:

wherein S represents the area of the region, L represents the length of a diagonal line of the region, and K is calculated according to the perimeter of the earth and the property of the vehicle;

the following linear programming model is established:

is limited to

i,j∈N.

N ═ {1,2, 3.., N } is the set of candidate cities, F is the sum of the distances between all cities;

because the number of candidate cities is large, the linear programming problem is calculated by adopting an immune algorithm to obtain an optimal solution, and the method comprises the following steps:

step 101: defining an optimization problem of linear planning of a transportation route;

step 102: an initial optimal solution to the optimization problem is generated based on the urban combinations,

step 103: evaluating the affinity between the optimization problem and the initial city, and measuring the quality of the optimal solution through the affinity;

step 104: according to the affinity, selecting n optimal cities, wherein the optimal cities meet the objective functionThe optimal solution of (2);

step 105: the n optimal cities are cloned to generate a temporary clone set C, the C comprises the feasible solution optimization problem, the affinity of the optimization problem and the initial city is improved, and the clone set C is used for optimizing the feasible solution;

step 106: generating a new city through immune selection based on n optimal cities and the clone set C to obtain a new feasible solution;

step 107: generating a memory set M through the feasible solutions, wherein M is a set of all the feasible solutions, and the optimal solution is contained in the set of the feasible solutions;

step 108: generating new cities by immune selection and based on the set M, the new cities being generated by modifying some feasible city group in M or one city;

step 109: and checking and analyzing whether the new city meets the requirement of an actual problem, wherein the actual problem requirement is a limiting condition except an objective function for the site of the selected site, namely a condition except the shortest Euclidean distance.

4. The system of claim 2, wherein the simulated annealing algorithm used by the transport loading module comprises the following steps:

suppose there are N objects and a backpack with a volume of V and a capacity of U, ith^thThe maximum number of articles is n [ i ]]Each object having a volume a [ i ]]Weight of each object b [ i ]]The value of each item is w [ i ]]To maximize the value of the items placed in the bag, the transformation function of the problem is:

f(i，u，v)＝max{f(i-1，v，u)，f(i-1，v-l*a[i]，u-l*b[i])+l*ω[i]}

l is an integer between 1 and n [ i ];

obtaining the solution by applying a hybrid simulated annealing algorithm;

order: i ═ 1,2, ·, N } a set of all shipments; v_iCargo volume I (I ∈ I);

CV is the volume of the container;

current loading sequence adaptation value (loading rate):

solving the maximum value of the loading rate as:

O＝max{g(B_list)}

the concrete steps of solving are as follows:

step S201: inputting the set I, setting the cooling speed eta, 0<η<1. Maximum number of iterations n and cooling target t_goal；

Step S202: solving a loading block B;

step S203: according to B, randomly generating a sequence B_list(ii) a Setting the initial temperature t>t_goalSimultaneously, let k equal to 1;

step S204: solving the adaptive value G-G (B)_list)；

Step S205: random Generation of New Loading sequence B'_list(ii) a The recipe value G ' ═ G (B ' was calculated using the recipe subroutine '_list)；

Step S206: if G'>G, then order B_list＝B′_listOtherwise, let k be k + 1;

step S207: if k > n, executing step S208, otherwise, executing step S205;

step S208: cooling; setting t as t × η;

step S209: if t<t_goalUsing a loading scheme subroutine and based on B_listGet the loading scheme and update O to max { g (B)_list) }; otherwise, step S204 is executed.

5. The transportation distribution optimization system of claim 4, wherein the step S202 of solving the loading block B comprises the following steps:

step S2021: the length, width and height of the input goods.

Step S2022: arranging goods according to the bottom area of the goods to obtain a loading sequence I_list(ii) a Initializing iI ═ 1 and iB ═ 1;

step S2023: placing iI^th) Cargo entering container iB while ordering I_list＝I_list{ iI }, wherein iI^th(iI∈I_list)；

Step S2024: selecting goods jI with the largest bottom area and the goods meeting constraint conditions;

step S2025: if jI exists, put on the existing cargo in iB and make I_list＝I_list{ jI }; if jI does not exist, go to step S2026;

step S2026: if I_listFor empty set, stop inputting, if I_listNot empty set, iB ═ iB + 1; and performs step S2023.

6. The transportation distribution optimization system of claim 4, wherein the specific steps for solving the loading scheme are as follows:

step S301: input loading sequence B_listAnd the identity of the cargo and container and let iB equal 1;

step S302: selection B_listiB of (1)^thA cargo; obtaining a current loading point chart EP; calculating the current iB based on the table EP^thA load point value for the cargo;

step S303: if all the load point values are less than 0, executing step S4; otherwise, the loading point in the table EP having the maximum loading point value and its correspondence with the goods are selected as the iB^thThe loading position and the loading direction of the goods;

step S304: obtaining a new loading scheme by using the current loading block;

step S305: let iB ═ iB + 1;

step S306: if iB>length(B_list) Outputting the result; otherwise, step S302 is performed.