Hindawi

Computational Intelligence and Neuroscience

Volume 2022, Article ID 8468438, 9 pages
https://doi.org/10.1155/2022/8468438

Research Article
Logistics Distribution Route Optimization Based on
Genetic Algorithm

Liu Xin, Peng Xu, and Gu Manyi

School of Humanities and Management, Southwest Medical University, Luzhou 646000, China

Correspondence should be addressed to Gu Manyi; [email protected]

Received 22 April 2022; Revised 6 June 2022; Accepted 13 June 2022; Published 6 July 2022

Academic Editor: Kapil Sharma

Copyright © 2022 Liu Xin et al. This is an open access article distributed under the Creative Commons Attribution License, which
permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Aiming at the problem of logistic division based on genetic algorithm, it is planned to study the improvement of logistic
distribution methods. We first meet the requirements of the genetic algorithm of logistic development, use the division method to
divide the delivery area of the gene, and formulate a functional delivery plan, which generally includes weight measurement,
measurement time, customer value measurement, instrument measurement time, and the whole process index. We set weight
goals and find the best way to improve genetic algorithm delivery. The experimental comparison results show that the optimal
method takes less than 2 minutes to find the optimal method, while the normal process takes 4 minutes to find the optimal
method, and the longest can reach 5 minutes. The comparison shows that the traditional algorithm takes longer to find the correct
way than the algorithm developed this time. Finally, the simple logistic distribution optimization method model and the soft time-
limited logistic distribution processing optimization model are calculated and simulated by the genetic testing algorithm and
genetic algorithm development. The effectiveness of the improved genetic algorithm in local research and the effectiveness of the
logistic transportation allocation solution are determined.

1. Introduction driving position to deliver the goods to the customer in a

short period of time. The functional elements of an express
Due to the increase in the timely delivery of small and large logistic system refer to the basic capabilities of the express
batches of goods, the transportation cost has increased year logistic system. These basic capabilities are effectively
by year, and the transportation of many merchants has combined and connected, becoming the general function of
exceeded the sale price. Choosing the best mode of trans- express logistics, and the overall purpose of the express
portation has become an important measure to control logistic system can be reasonably and effectively realized.
logistic costs. The VRP (vehicle routing problem) is a major The functional elements of the express logistic system are
problem in transportation today and is slowly gaining at- generally considered to have transportation, storage, pack-
tention. In this study, an improved genetic algorithm is used aging, loading and unloading, handling, circulation and
to solve the logistic distribution routing problem. The focus processing, distribution, express logistic information, etc. If
of the distribution operation of the logistic center is how to viewed from the actual work links of express logistic ac-
effectively use the vehicle and determine its most economical tivities to examine, the express logistics is composed of the
driving route map, so that the goods can be delivered to the above seven specific works. In other words, express logistics
customers in the shortest time. Distribution and trans- can achieve the above seven functions. Distribution and
portation are the most important functional elements in shipping are the most important roles in distribution.
logistic and distribution activities, and the rationalization of Rationalization of distribution depends on the ration-
distribution largely depends on the rationalization of dis- alization of distribution and transportation measures. Due
tribution and transportation. The goal of in-store delivery is to the large number of urban customers and the difficulty of
how to properly use the vehicle and determine its best distribution, how to build the best road is not only the
 8483, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/8468438, Wiley Online Library on [07/06/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2 Computational Intelligence and Neuroscience

characteristics of distribution and transportation, but also an Considering the income and total cost, this study proposes a
arduous task. Whether the distribution is reasonable or not new concept of income distribution [6]. Daugherty et al.
has a significant impact on export speed, cost, and benefit. have developed one of the multi-purpose stochastic pro-
Applied research determining product distribution, espe- gramming models for transportation emergency construc-
cially through consumer distribution, is an important and tion using a combination of vertical and horizontal
difficult task in product distribution. It is an important way distribution systems [7]. *e traffic crisis (VRP) was first
to improve the performance of freight forwarders to un- mentioned by Sun et al. Since then, the VRP has become an
derstand the research of logistic distribution vehicle routing important research hotspot and research work in the field of
optimization. combinatorial optimization [8]. Yang et al. proposed a
segmentation method for the VRP problem. To this end,
2. Literature Review they directly determine the set of best possible solutions
(possible solutions) and develop a simple mathematical
*e logistic industry plays an important role in international model of VRPAs, as shown in Figure 1 [9].
trade and plays a leading role, and more importantly, it has a Based on the current research, a study is planned for the
greater impact on all aspects of the global economic influ- delivery of the optimization process. First, in order to meet
ence. By business, transportation, information technology the requirements of the development of genetic algorithm
industry, and the development of talent employment in- logistics, the logistic distribution path of genetic algorithm is
dustry, we prompted the warehouse storage, logistics, goods optimized to solve the problem of finding the optimal path.
processing and distribution, and cargo loading and Second, we adopted the clustering algorithm of genetic al-
unloading industry development; formed a relatively com- gorithm logistic distribution area and establish logistic
plete new industrial chain, optimized the allocation of re- distribution path optimization target function, mainly in-
sources; to a large extent, pulled the domestic demand; also cluding weight index, aging index, customer importance
provided an opportunity for youth entrepreneurship; and index, time window index, and total path index, and set the
brought social and economic development and prosperity. distribution target weight, to find the optimal target path, in
Shipping has gradually become an important challenge for order to realize the genetic algorithm logistic distribution
enterprises [1]. *e goal of export products is to provide path optimization. For today’s enterprises, how to reduce
customers with the best service at the lowest price, so as to operating costs, increase profits, and improve transportation
improve the overall efficiency of the industry, increase the has become a major issue in logistic management, which
overall profit, and improve market competitiveness. directly affects and determines the importance of compe-
*erefore, studies on export improvement are not only tition in the logistic industry. On the one hand, the logistic
concerned with the cost-effectiveness and service efficiency industry is facing good development prospects, and on the
of the logistic industry, but can also provide data for income other hand, it is also facing a bottleneck restricting its de-
distribution. Manufacturers maximize efficiency and en- velopment. Especially with the acceleration of economic
hance the overall image of the enterprise. globalization and information technology, a series of out-
*e optimization problem has always been the focus of dated modes such as traditional and simple transportation
scientists’ research, and the research on the distribution and storage services have limited and delayed the generation
types of joint ventures is in the first stage. Asokan et al. and development of efficient and professional logistic ser-
obtained centralized joint ventures and equipped them with vices. In industrial areas, many logistic centers choose
multiobjective genetic optimization to solve the split order product distribution according to customer needs, resulting
problem in joint ventures [2]. Wang and Ying used fuzzy in less resource usage and high cost. Sharing can be a new
programming models and algorithms to consider the un- way of distribution. Codistribution is to exchange the
certainty of vehicle travel and customer service time, de- previous allocation according to different suppliers and
veloped a collaborative fuzzy programming model with the different products into “heavy equipment and allocation”
problem of minimizing vehicle delivery costs, and adopted that is no different from suppliers and inventories. *at is,
an adaptive discrete particle swarm optimization algorithm transporting goods on the same route, using the same truck
to solve the problem [3]. Based on the concept of virtual to transport multiple customers, reduces operating costs,
business and capital integration, Li et al. provided a model and improves resource utilization and the competitiveness
for the integration of third-party logistic companies in a of the logistic industry.
virtual environment and simulated the model with Auto-
Mod simulation software [4]. Ran et al. have developed an 3. Logistic Distribution Path Optimization
integrated circuit system that takes advantage of the point- Based on Genetic Algorithm
to-point power distribution model and the hub electrical
model to provide research-specific heuristics. Using this 3.1. Significance of Route Distribution Optimization. From
method is very effective in emergency situations [5]. Mar- the application point of view, logistic distribution optimi-
tinson et al. argue that joint ventures can reduce product zation is not only an important part of logistic distribution
distribution costs, increase revenue, and improve business optimization but also an indispensable topic in e-commerce
efficiency through capital sharing. Whether the joint venture games. Optimizing freight transport can improve the effi-
is healthy or not depends on whether the joint venture’s cost ciency of logistic transactions and promote logistic research.
sharing and distribution of benefits are reasonable. *e research theory and process of optimizing freight routes
 8483, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/8468438, Wiley Online Library on [07/06/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Computational Intelligence and Neuroscience 3

Place the order

Confirm

Warehouse
Confirm receipt Stores

Being pushed

Short notice
Return notice

Being pushed

Short notice

Vehicle Management personnel

Figure 1: Logistic distribution path of genetic algorithm.

are the basis for developing transportation, creating a was developed in 1975 by Professor Holland, a global re-
modern reference command system, developing profes- search tool based on natural selection and genetic research.
sional transportation technology, and developing e-com- It stimulates the evolution of genes from genetic selection,
merce. In the construction of logistic distribution crossover, and mutation and is used to determine the most
information network platform in most rural areas, due to efficient chromosomes using an energy function. Genetic
various conditions, there are still many deficiencies in the algorithms modeled on Darwin’s theory of natural evolution
design of the platform. One is that most of the products sold and genetic change are robust, globally optimized, and
in rural areas are still sold and circulated in small shops. designed to solve complex multivariate optimization and
*ere are many sectors of wholesale trade, but most involve combinatorial problems, and various forms of utility bills.
traditional industries. Farmers generally rely on traditional *e relative values are shown in Table 1 [11].
methods to obtain information on the circulation and
distribution of agricultural products, with little access to
news or the internet, and the rest from local markets. 3.3. Genetic Algorithm Based on Multilogistic Center
Traditional practices of raw data, logistics, and distribution Collaborative Distribution Path Optimization
are depriving most rural areas of the environment in which
information networking platforms are built. Second, while 3.3.1. Algorithm Description
some agricultural discussions focus on product availability,
in practice, most sites do not have robust export plans, (1) Choose Code and Create a Population Default. *e
reasonable standards, or affected export directions. For program uses the natural copper encoding method. m2, m4,
example, in the data distribution process, the lack of an and m6 are distributors that randomly generate a large
objective data management interface can lead to data dis- number of n users. 1, 2, 3 . . . n are all users. In this way, m1,
tortion and data expiration. *erefore, the construction of m2, m3, 4, 5, 6, n −1, and nm can be encoded to express the
the existing logistic distribution information platform in classification solution. *ey are m1 and 12m, which means
most rural areas is not perfect [10]. that the distribution center distributes the goods to each
household 1 and 2. Shipping to each distribution center must
meet design constraints. Every operator can prove it by
3.2. Basic Concepts of Genetic Algorithm. *e genetic algo- distributing it to its users. When selecting a distribution area
rithm (GA) is an intelligent algorithm. *is is an important for coding, we combine the previous distribution area with
way for domestic and foreign scholars to study VRP. His idea the next distribution area, convert m1 into I2, calculate the
 8483, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/8468438, Wiley Online Library on [07/06/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
4 Computational Intelligence and Neuroscience

Table 1: Correspondence.
Biological genetic concept Concepts expressed in optimization problems
Individual Feasible solution
Chromosome Solution coding
Gene Characteristics of each component in connection
Individual fitness Objective function of solution
Population *e set of feasible solutions

distribution area including m13 and m33, etc., and ensure that represent the number of n cities) (nmX )2 � vJ . Equation
there is a user relationship between the two. According to d(v) represents the distance from city vi to vi+1 , as shown in
this method, two adjacent parts get the first possible solu- equation (2).
tion, the first group [12].
T � 􏽘 d vi + vi+1 􏼁 + d vm 􏼁. (2)
(2) Fitness Function. In order to evaluate the delivery pro-
cess, the coding method is determined according to the 3.5.2. Genetic Operators for VRP. *e VRP genetic operators
specifics of the vehicle scheduling problem, i.e., each cus- usually include additional operators, crossover operators,
tomer sends, each customer only delivers to one dealer, and exchange operators, and reverse operators [15].
the goods are sent to everyone. Machine constraints will not
be satisfactory after multiple transfers. *erefore, the defi- (1) Personnel Selection. Common selection processes of VRP
nition of the exercise function should reflect both the fea- schemes include roulette wheel selection mechanism, op-
sibility of the solution and the cost of the parallel process. timal placement process, reliability model selection mech-
For self y, the reciprocal of the relative distribution is M, anism, category model selection mechanism, association
0M � 0, which means that this person is possible; if his goal is selection mechanism, and isolation model. *is form uses a
W, then the effective function of y is F, as described in roulette wheel to choose ideas. *e main idea of the roulette
equation (1). selection mechanism is that the outcome of each candidate is
p
􏽐r�1 􏼐wy + my 􏼑 proportional to the value of the energy function. If the
y . (1) population is n and the energy of an i is Fi, then the
w + mp probability of selecting I as the next generation heir is de-
termined by equation (3).
Here, pω represents the severity of the failure (the weight
m
can bring a good number for the maximum value of the
target job). Combined with the actual policy, pω is 1000 [13]. pi � F i 􏽘 Fi . (3)
i�2

(2) Crossover Operator. *e cross operator is an important

3.4. Basic 2eory of Logistic Distribution Path Problem.
operator in the genetic algorithm, which gives a brief in-
Ramser and Danzig originally proposed a vehicle routing
troduction to the more mature cross operator in the genetic
problem (VRP) that solves a specific constraint (e.g., start
algorithm. Common crossover operators have single-point
time, end time, and transit time) and achieves a goal (e.g.,
crossover, double or multipoint crossover, uniform crossing,
shortest total number) within the maximum freight range
arithmetic crossing, etc., single-frequency division, digital
(the shortest time, cost, vehicles, etc.)., which are essential
frequency division, and the like. *is recipe uses a partial
for customers with various special transportation needs.
cross recipe. Intersections with half a graph are also called
Since the VRP problem is a complex problem in NP with
partial intersections. All frequency division is divided into
similar production and environmental problems, the study
two stages: (1) two-point frequency division is required to
of the VRP problem provides theoretical support for
operate on a single line of code; (2) according to the rela-
studying the problem of improving efficiency and effec-
tionship diagram of different subvalues at the intersection,
tiveness [14].
the sowing amount of different seeds outside the intersection
area is adjusted [16].
3.5. Application of Genetic Algorithm in VRP
(3) Change in Business Operator. *e genetic algorithm
3.5.1. Modeling and Description of VRP Problems. *e VRP describes the crossover operation. From a genetic algorithm
(transportation issues) is one of the key issues in the modern perspective, solutions can only be developed through se-
logistic industry. *is is also a difficult problem for NP and lection mechanisms and crossover strategies. Mutation
gradually becomes popular n. We combine cities and their simply fixes and adds some genes involved in the selection
points to find a system where each city is only passed once, and hybridization process. Only change can understand the
and all goods are loaded on time. *e explanation for the world of GA.
VRP problem is simple. In short, to find the shortest path or
reduction value through n sending points, we find the plan (4) Operator Returns. Operators have shown some growth in
for a subset of integers (x � m, where the elements of nX genetic work due to the shortcomings of early integration
 8483, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/8468438, Wiley Online Library on [07/06/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Computational Intelligence and Neuroscience 5

that often arise when using genetic processes to solve the

VRP problems. Experiments show that the calculated op- Ask questions
timal solutions can be seen in a wide range of applications.
*erefore, well-thought-out algorithms can prevent early
interactions and improve the global performance of genetic
work [17]. Chromosome fitness was
calculated

3.6. Steps and Flowchart of Basic Genetic Algorithm. *e key

steps are as follows: ① we determine the number of the first
population, define the coding process, create the first
population, and list all members of the first population as its Meet the conditions
citizens; ② the salary of each citizen will be calculated by
rank, and it will be determined whether the operation is
satisfied cost. If a good solution is encountered, if not, we go
to step ③; ③ according to the cost and choice of the op- Choose
erator, we select the person who is most suitable for edu-
cating future generations; and ④ we create a new person in
the community according to the cross-exchange operator
and return to the step in the new population ②. *e op-
eration diagram is shown in Figure 2 [18]. Variation

4. Experiment and Research Figure 2: Flowchart of basic genetic algorithm.

4.1. Genetic Algorithm. *e genetic algorithm is initiated by

an arbitrarily initiated public, adopts a random selection In Figure 4, the received order is the main part of the
mechanism, and is based on the principle of survival of the shipment. If the shipper and the freight forwarder receive the
fittest in nature, so that very good people can have a lot of shipping plan required by the customer, it affects the pro-
time to enter the future; newborns with strong adaptability cess; then, the shipment and transit time receive the con-
are formed through overtime; and change ensures diversity. signor’s cargo information, and finally, check and receive the
Over several generations of “evolution,” the population consignor’s waybill [22]. *e registry is close to receiving
slowly reaches or remains normal until the algorithm linked orders. Its main purpose is to register the purpose of
converges. *rough multiple operations, multiple chro- shipment and the delivery customer’s bill of lading number
mosomes finally obtained positive results. *e higher the and receipt received by your delivery person in the delivery
cost of exercise, the better the drug. *erefore, the method of area, and to confirm and sign the shipping register. Phone
genetic coding is very high when using the genetic algorithm calls and shipping are the most important links in inter-
to optimize the vehicle logistic distribution network system. national shipping. *e quality of the two links directly affects
Chromosomes represent the solution to the LRIP problem. the performance of the logistic industry. *e steps involved
Chromosomes are coded for natural copper. Each chro- in making a call include planning for traffic, writing traffic
mosome consists of four parts, as shown in Table 2 [19]. information, submitting support services and tracking
feedback, and completing forms using a computer. *e
shipper is the shipper who arranges the goods according to
4.1.1. Model of Genetic Algorithm. *e genetic algorithm is the final destination of the received goods, the weight of the
an intelligent algorithm, which is often used as an important goods, the volume of the goods, and the maximum carrying
method to solve VRP (transportation problem). Professor capacity of the vehicle, and informs the customers of their
Holland conducted the first international research on nat- needs. Ensure the waiting time of vehicle arrival through
ural selection and genetics in 1975. He modeled the evo- reasonable transportation planning [23].
lution of genetically engineered species, including selection,
crossover, and transformation, and used force to demon-
strate the efficiency of chromosomes. It has strong ro- 4.2. Division of Logistic Distribution Area. Before optimizing
bustness and global optimization ability. It is necessary to the logistic distribution path, we analyze the regional dis-
solve complex multipole optimization problems and com- tribution process, as shown in Figure 5.
binatorial problems, and has a wide application cost. Fig- Based on the above analysis, the logistic distribution area
ure 3 is a functional diagram of the genetic algorithm [20]. is divided by the grouping algorithm. *e distribution is as
follows:

4.1.2. Logistic Distribution Process. *e normal operation Step 1. We select the starting group, find n files in all data
process includes orders, registration, telephone, delivery, files in the group position, find n files in the file data of n � 1,
pick-and-place, in-transit, sign-in, and transportation, as 2, . . ., n information, and take n � 1, 2, . . ., n information
shown in Figure 4. that is based on group settings. According to the needs of the
 8483, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/8468438, Wiley Online Library on [07/06/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
6 Computational Intelligence and Neuroscience

Table 2: Chromosome coding design.

1 bit N-bit K-bit 1 bit
Main engine factory Vehicle number Dealer service Objective function value

Start
Consumers

Code the problem

Shipping information

Population initialization

Select operation
Enterprise distribution center

Mutation

Processing
Whether the
conditions are
met

Figure 3: Work flowchart of genetic algorithm.

Regional distribution
Order
receiving
Figure 5: Route distribution diagram.
Register
of the group entity. According to the model, the center of the
nearest data center is calculated and divided into centers.
Call to arrange *e goal is to save time for trucks on the road.

e team Pick up and

Step 3. We adjust the center position. Since the center of the
Warehouse n groups containing some of the most frequently used data
handover delivery
changes, the data in the groups are modified according to the
Figure 4: Flowchart of logistic distribution in local areas [21]. calculation results of step 2. It is calculated as follows: we
e-commerce logistic distribution center, the management of divide the longitude coordinates of all points in a set by the
N groups of office partitions can be divided into (x1, y1), (x2, number of points. According to this principle, the sum of all
y2), . . ., (xn, xn) to ensure that the original distribution areas points is divided by the number of points to get the latitude
do not overlap. and longitude coordinates of the new group and the location
of the new group. *e goal is to make the distribution as
efficient as possible and avoid inconsistencies. *erefore, the
Step 2. We use the distance model to find the distance volume of each particle must be equal.
between groups’ center and each data. *e calculation
formula is shown in formula (4)as follows: Step 4. We determine the center position, use the center
√�� point count in step 3 for the next iteration cycle, and
D(k + 1) � ga − fa+1 . (4)
compare the group in the middle position count with the
Among them, D(K+1) represents the distribution area; three steps above, and the middle position is final. *e milk
ga represents the group center; and fa+1 represents the data count is represented by formula (5)as follows:
 8483, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/8468438, Wiley Online Library on [07/06/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Computational Intelligence and Neuroscience 7

a
K • n . Path to the document
g�􏽘q f •h (5)
a

Among them, kg represents the product group of the Path to the table
logistic division; 􏽐aq represents the large group; fn repre-
sents the distance from the group; h represents the mea-
Path selection
surement comparison group; and a represents the density of
distribution points.
If the spatial positions calculated according to the model Path optimization Path
greatly differ, we recalculate so that the volume of each Path division optimization
objective
vehicle is equal.
Figure 6: Path design process.

4.3. Genetic Algorithm Logistic Distribution Path

Optimization. On the basis of the abovementioned logistic
distribution area division, the logistic division method is 1 n+1
optimized. We evaluate the design process prior to opti- Sk � 􏼠 􏼡, (10)
F hy · j
mization. *e design method is shown in Figure 6.
On this basis, the delivery method is established and where Sk represents the distance from the kth distribution
optimized for various operational purposes, usually in- point; 1/F represents the distance after shipment; hy·j is the
cluding weight measurement, time measurement, user value exchange coefficient of the whole method; and N is the total
measurement, window measurement time, and total mea- value measurement.
surement methods. *e special counting procedure is as Based on the above weights, a refinement of measure-
follows. Weight index: in the process of regional logistics and ment time, customer value time, measurement time, and
distribution, the weight of the goods must be calculated, as total measurement method, the delivery purpose is created.
shown in sections (6) and (7) for weights as follows: Since shipping companies charge different amounts for
Sg � 􏽘 f(g − i)tu , export purposes, the weight is set for distribution purposes.
(6) *e above operational goals are not broad until set. *e
i
formula is expressed as formulas (11) and (12)as follows:
Xi
Zi � . (7) x−n
Xi Xe � , (11)
M−m
Among them, Sg represents the delivery place number; f x � AF + ε,
represents the weight of the delivered goods; g is the regular (12)
source of logistic distribution; i is the weight of the goods X � X1 X2 X3 􏼁, F � F1 F2 Fm 􏼁, ε � ε1 ε2 ε3 􏼁,
shipped from the ith place; and cleanliness is a measure of where xe represents the value of the measurement in each
weight. input; M represents the minimum value in m; and x rep-
Based on the above weight scale, the aging scale is resents the maximum value in n.
calculated. As the e-commerce export market is more and On the basis of the above dimensionless processing, the
more cold and new, and the requirements for delivery time weight is determined, and the relative importance is mea-
are getting higher and higher, the aging index is designed to sured by comparing the two. *e analysis is shown in
affect the demand for products in a timely manner. *e Table 3.
formula is expressed as formulas (8) and (9)as follows: According to the functions and weights of the above
1 ta − ti logistic distribution indicators, the distribution path is op-
St � 􏼠 􏼡, (8)
N Tfx timized, as shown in equation (13).
G � η ∗ W1 , W2 , W3 , W4 , W5 􏼁. (13)
x1 � a1 F1 + a2 F2 + a1m + Fm + ε
􏼢 􏼣, (9) *ey are as follows: W1, W2, W3, W4, and W5 represent
X2 � a2 F2 + a2 F2 + a2m + Fm + ε2
the total weight of the weight indicator, time indicator,
where ti represents the storage time of the goods at the ith customer saliency indicator, time window indicator, and
distribution point; Tfx represents the time required to deliver total traffic indicator, respectively; η is the proportional
the goods; ta is the time to exit the distribution; 1/N rep- profit margin.
resents the components involved in the distribution; and St is
the estimated time of arrival of the product.
Finally, all measures for export logistics are included. 4.4. Test Data. *e test data are provided by the logistic
Due to different logistic distribution methods, in order to provider. A total of 7 items need to be delivered in the
improve distribution efficiency, we include the overall shipment from the fulfillment center. *e main data of these
measures. *e formula is shown in formula (10)as follows: seven projects are shown in Table 4.
 8483, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/8468438, Wiley Online Library on [07/06/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
8 Computational Intelligence and Neuroscience

Table 3: Relative column scale values. 5

Weight
Different levels of importance
assignment
4
i and j two elements are equally important 2
Element i is slightly more important than element j 4
Element i is more important than element j 6
3
Element i is much more important than element j 8

path
2

Table 4: Logistic good information.

Serial number Weight/kg Importance/% Unloading time/h 1

1 0.6 4 0.4
2 0.4 5 0.7
0
3 0.1 6 0.6
1.0 1.5 2.0 2.5 3.0 3.5 4.0
4 0.5 4 0.6
TIME
5 0.67 2 0.64
6 0.29 1 0.36 Figure 7: Comparison between optimization algorithm and tra-
7 0.66 2 0.2 ditional algorithm.

According to the analysis in Table 4, the weights, values,

and loading times of the seven distribution zones are dif- Table 5: Measurement error of travel path time.
ferent. *e best way to find the seven products is to use the Traditional Algorithm in
Error index
two methods, and the time to find the best way is compared algorithm this study
with the two methods. Relative error 95.20 77.03
Average relative error 9.54 7.44
Mean absolute relative error 7.84 5.98
4.5. Analysis of Experimental Results. *e time results of the Maximum absolute relative error 0.28 0.07
logistic distribution path optimization algorithm and the Square root of mean square
0.07 0.02
traditional algorithm to find the optimal path are shown in sum of relative error
Figure 7. Equalization coefficient 0.70 042
From the analysis in Figure 7, it can be seen that in this
project, the time to find the best solution is less than 2
5. Conclusions
minutes, while in the normal process, the time to find the
best method is more than 3 minutes. In contrast, the time *is work has become increasingly important as the pace
required for traditional methods to find the optimal method of global market integration accelerates, and as the
is five times that of the algorithm developed this time. technology develops rapidly, its share of the global market
From the above experiments, it can be demonstrated that grows. It has gradually become an important indicator for
the derived algorithm optimization method developed in domestic and foreign enterprises to measure their com-
this form in the context of big data has a shorter time to find prehensive competitiveness. *erefore, choosing the
a better method than the usual process and can meet the correct, appropriate, and effective research can save costs
needs of good delivery. and improve the overall competitiveness of enterprises.
To clearly demonstrate the effectiveness of the different Because the logistic distribution optimization process
algorithms, the best performance of the different algorithms involves many disciplines and many influencing factors, it
is compared to the best method and the errors are tabulated. is necessary to use various algorithms to compare and
*e results are shown in Table 5. identify different models. Classified logistics is an im-
From the analysis in Table 5, it can be seen that among portant part of the combinatorial optimization process
the two different optimization algorithms, the best way to and is robust. However, the optimization algorithm itself
derive the design time of the optimization algorithm is lower is rich in content and is in a stage of rapid development
than the actual one, and each error is lower than the per- both in practice and theory.
formance of the result, that is, the traditional algorithm. *e
above experimental results show that the best way to derive Data Availability
the optimization algorithm is to follow the best way, thus
confirming the accuracy of designing the logistic distribu- *e data used to support the findings of this study are
tion optimization algorithm. available from the corresponding author upon request.
 8483, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/8468438, Wiley Online Library on [07/06/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Computational Intelligence and Neuroscience 9

Conflicts of Interest [12] X. Zhu, R. Zhang, F. Chu, Z. He, and J. Li, “A flexsim-based
optimization for the operation process of cold-chain logistics
*e authors declare that they have no conflicts of interest. distribution centre,” Journal of Applied Research and Tech-
nology, vol. 12, no. 2, pp. 270–278, 2014.
[13] H. Xiong, “Research on cold chain logistics distribution route
Acknowledgments based on ant colony optimization algorithm,” Discrete Dy-
namics in Nature and Society, vol. 2021, Article ID 6623563,
(1) Sichuan e-commerce and modern logistics research 1–10pages, 2021.
center funded the “Research on Sichuan Rural e-Commerce [14] H. Gao, X. U. Guang-Hui, Z. R. Wang, and H. Yang, “Ap-
and Logistics Helping Rural Revitalization and Development plication of clonal selection algorithm on competitive location
(no. dswl21-11).” (2) Sichuan Rural Development Research model of logistics distribution center,” Journal of Highway
Center funded “Sichuan Rural e-Commerce and Logistics and Transportation Research and Development, vol. 132, no. 6,
Development Research (no. cr1507).” pp. 529–538, 2007.
[15] Y. P. Wang, “Adaptive ant colony algorithm for the vrp so-
lution of logistics distribution,” Research Journal of Applied
References Sciences, Engineering and Technology, vol. 6, no. 5, pp. 807–
811, 2013.
[1] T. Wang and Y. Wu, “Study on optimization of logistics [16] J. B. Sheu and T. Kundu, “Forecasting time-varying logistics
distribution route based on chaotic pso,” Computer Appli- distribution flows in the one belt-one road strategic context,”
cations in Engineering Education, vol. 47, no. 29, pp. 218–221, Transportation Research Part E: Logistics and Transportation
2011. Review, vol. 117, pp. 5–22, 2018.
[2] P. Asokan, R. Saravanan, and K. Vijayakumar, “Machining [17] Z. Zhang, J. Li, and N. Xu, “Robust optimization based on ant
parameters optimisation for turning cylindrical stock into a colony optimization in the data transmission path selection of
continuous finished profile using genetic algorithm (ga) and WSNs,” Neural Computing & Applications, vol. 33, no. 24,
simulated annealing (sa),” International Journal of Advanced pp. 17119–17130, 2021.
Manufacturing Technology, vol. 21, no. 1, pp. 1–9, 2003. [18] T. Ning, J. Wang, and Y. Han, “Logistics distribution de-
[3] Wang, F. Ying, Z. Teng, and Liyi, “Low carbon logistics carbonization pathways and effect in China: a systematic
distribution route optimization research based on chaos ant analysis using vrpsdp model,” International Journal of Low
colony algorithm,” Journal of Investigative Medicine, vol. 62, Carbon Technologies, vol. 16, no. 4, pp. 1404–1411, 2021.
no. 1, p. S105, 2014. [19] N. Li, Z. Mao, and Dbntfpo, “ANN-based approach for lo-
[4] B. Liu, “Logistics distribution route optimization model based gistics distribution optimization,” Mathematical Problems in
on recursive fuzzy neural network algorithm,” Computational Engineering, vol. 2022, Article ID 5068748, 10 pages, 2022.
Intelligence and Neuroscience, vol. 2021, Article ID 3338840, [20] S. Yang, “Optimization of urban logistics distribution path
1–10pages, 2021. under dynamic traffic network,” International Core Journal of
[5] R. A. N. Wenxue, S. H. I. Xinling, F. U. Huasen, and Engineering, vol. 6, no. 1, pp. 243–248, 2020.
Y. A. N. G. Guomin, “Application research on ant colony [21] H. Yan, “Optimization of regional logistics distribution in-
algorithm in logistic distribution route-optimization of fresh formation network platform,” Security and Communication
agricultural products,” International Journal of Digital Con- Networks, vol. 2022, Article ID 6729247, 7 pages, 2022.
tent Technology and its Applications, vol. 7, no. 6, pp. 391–399, [22] J. Kong, Y. Zhao, F. Wang, and Q. Xiang, “Research on in-
2013. novation of logistics terminal distribution service mode,”
[6] A. Martinson and X. QiangQiang, “Route optimization in Open Access Library Journal, vol. 8, no. 12, p. 8, 2021.
logistics distribution based on particle swarm optimization,” [23] C. L. Wang, Y. Wang, Z. Y. Zeng, C. Y. Lin, and Q. L. Yu,
International Journal of Computer Application, vol. 178, “Research on logistics distribution vehicle scheduling based
no. 30, pp. 23–27, 2019. on heuristic genetic algorithm,” Complexity, vol. 2021, no. 11,
[7] P. J. Daugherty, T. P. Stank, and A. E. Ellinger, “Leveraging pp. 1–8, 2021.
logistics/distribution capabilities: the effect of logistics service
on market share,” Journal of Business Logistics, vol. 19, no. 2,
pp. 35–51, 1998.
[8] H. Sun, Z. Gao, and J. Wu, “A bi-level programming model
and solution algorithm for the location of logistics distribu-
tion centers,” Applied Mathematical Modelling, vol. 32, no. 4,
pp. 610–616, 2008.
[9] L. Yang, X. Ji, Z. Gao, and K. Li, “Logistics distribution centers
location problem and algorithm under fuzzy environment,”
Journal of Computational and Applied Mathematics, vol. 208,
no. 2, pp. 303–315, 2007.
[10] J. B. Sheu, “A hybrid fuzzy-optimization approach to cus-
tomer grouping-based logistics distribution operations,”
Applied Mathematical Modelling, vol. 31, no. 6, pp. 1048–1066,
2007.
[11] W. Fan, H. Xu, and X. XuXu, “Simulation on vehicle routing
problems in logistics distribution,” COMPEL: 2e Interna-
tional Journal for Computation & Mathematics in Electrical &
Electronic Engineering, vol. 28, no. 6, pp. 1516–1531, 2009.