CN112085271B - A Crowdsourcing Model-Based Optimization Method for Collecting Goods in Traditional Industry Clusters - Google Patents

Abstract

The invention discloses a traditional industry cluster cargo collection path optimization method based on a crowdsourcing mode. And matching the position of the available truck in the crowdsourcing mode with the position and the demand of a logistics demand party, performing cluster selection on the vehicle and the customer by applying cluster analysis and a greedy algorithm, and determining the vehicle and the route applied by the path optimization. And (3) calculating the solution of the previous step as an initial population by using a genetic algorithm, optimizing the path, and finally verifying and comparing the obtained different schemes to obtain an optimal path scheme. The method takes the genetic algorithm as the core, combines the basic concepts of the cluster analysis model and the greedy algorithm, and improves the operation efficiency, thereby utilizing the logistics resources of the traditional industrial cluster to the maximum extent, improving the logistics efficiency and reducing the logistics cost.

Description

A method for optimizing the collection path of traditional industrial clusters based on crowdsourcing mode

technical field

The invention relates to a traditional industrial cluster collection path optimization method based on crowdsourcing mode, and belongs to the technical field of logistics path optimization.

Background technique

After my country's reform and opening up, a number of traditional industrial clusters with greater influence have been formed, such as China (Changshu) Zhaoshang Mall and Nantong Dieshiqiao Home Textiles Market. These traditional industrial clusters have become an important fulcrum for leveraging domestic economic growth and modernization. However, these spontaneously formed traditional industrial clusters have relatively weak awareness of socialized logistics, high proportion of self-operated logistics, and lagging logistics supply chain management. At the same time, there are many main bodies of logistics enterprises in the cluster, which are small in scale and independent, which leads to the confusion of logistics lines and waste of logistics resources in traditional industrial clusters. These problems greatly limit the pace of development of the transformation and upgrading of traditional industrial clusters, and the logistics system of traditional industrial clusters needs to be optimized and upgraded urgently.

Crowdsourcing mode logistics distribution refers to the logistics mode in which the distribution work that should be completed by professional institutions is transferred to the social public groups outside the enterprise to complete. Generally, the logistics demander publishes the logistics demand to a special public information platform. Logistics practitioners who are willing to serve also enter their information into the information platform. The platform integrates resources, dispatches orders and dispatches vehicles in a unified manner. Finally, logistics practitioners settle with the platform and get paid, as shown in Figure 2. The crowdsourcing model can make full use of the idle logistics resources of the society and improve the efficiency of logistics distribution.

At present, domestic and foreign researches on capacity resource allocation, terminal distribution mode, and scheduling optimization have been relatively sufficient, but there are few studies on the application of crowdsourcing mode to distribution. The idea of crowdsourcing can effectively integrate the existing small and medium-sized logistics enterprises in traditional industrial clusters, and optimize vehicle scheduling based on the crowdsourcing model. has important guiding significance.

SUMMARY OF THE INVENTION

The technical problem to be solved by the present invention is: to provide a traditional industrial cluster cargo collection path optimization method based on crowdsourcing mode, which can reduce the calculation steps in the path optimization process, improve the computing efficiency, and obtain the optimal collection of goods more quickly. path. In this way, the logistics resources of traditional industrial clusters can be more fully utilized, the logistics efficiency can be improved, and the logistics cost can be reduced.

The present invention adopts the following technical solutions for solving the above-mentioned technical problems:

A method for optimizing the collection path of traditional industrial clusters based on crowdsourcing mode, comprising the following steps:

Step 1, analyze the cost in the process of collecting goods in traditional industrial clusters, and establish an objective function with the lowest total cost as the goal;

The objective function is as follows:

表示物流需求方i是否是货车r第一个服务的客户，

为0-1变量，i＝1,…,n，n表示物流需求方的数量；d_ij表示物流需求方i到物流需求方j的距离，

表示货车r是否为物流需求方i服务后立刻为物流需求方j服务，

为0-1变量，j＝1,…,n+1，j≠i，j＝n+1时表示货车将货运回集群物流中心；C_w表示单个货车将单位重量运输单位距离的成本，W_ri表示货车r到达物流需求方i取货后车上货物的总重量；C_i表示物流需求方i的惩罚成本，X_i表示物流需求方i是否在约定的时间内得到服务，X_i为0-1变量；Among them, C represents the total cost; C ¹ , C ² , C ³ , and C ⁴ represent the fixed cost of trucks, truck driving costs, truck weight costs, and time window transportation penalty costs, respectively; C _g represents the fixed use and maintenance costs of a single truck , X _r represents whether the truck r is used, X _r is a 0-1 variable, r=1,...,m, m represents the number of available trucks; C _d represents the cost of traveling a unit distance when a single truck is empty, d _ri represents The distance from the truck r to the logistics demand side i,

Indicates whether the logistics demander i is the first customer served by the truck r,

is a 0-1 variable, i=1,...,n, n represents the number of logistics demanders; d _ij represents the distance from logistics demander i to logistics demander j,

Indicates whether the truck r serves the logistics demander j immediately after serving the logistics demander i,

is a 0-1 variable, j=1,...,n+1, j≠i, j=n+1 means that the truck will return the freight to the cluster logistics center; C _w means the cost of a single truck transporting the unit weight per unit distance, W _ri represents the total weight of the goods on the truck after the truck r arrives at the logistics demander i to pick up the goods; C _i represents the penalty cost of the logistics demander _i , Xi indicates whether the logistics demander _i is served within the agreed time, and Xi is 0 -1 variable;

Step 2, under the condition that the location of the logistics demander, the freight demand of the logistics demander, the location of the truck, and the rated cargo capacity of the truck are known, set constraints;

Step 3: Using the location of the trucks as the clustering center, use the cluster analysis method to initially cluster the logistics demanders, and at the same time apply the greedy idea to determine the minimum number of trucks and the corresponding trucks, and at the same time determine the logistics demanders corresponding to each truck. ;

Step 4: According to the minimum number of trucks and the corresponding trucks determined in Step 3, secondary clustering is performed by using the cluster analysis method, and the logistics demander is allocated to the trucks determined above;

Step 5: According to the clustering result of Step 4, determine the path of the truck with the shortest total distance as the goal, and use the above-mentioned path of the truck as the initial population of the genetic algorithm. Based on the clustering result of Step 4, use the lowest total cost as the goal. The genetic algorithm optimizes the path of the trucks. During the optimization process, the number of trucks and the logistics demand side of the corresponding service of each truck are adjusted, so as to obtain the number of trucks corresponding to the lowest total cost and the path of each truck, that is, the lowest total cost. Program;

Step 6: Use the constraints set in Step 2 to discriminate the lowest total cost solution obtained in Step 5. When the lowest total cost solution obtained in Step 5 satisfies all constraints, the solution is the optimal solution; otherwise, adjust Each truck corresponds to the logistics demand side of the service and the path of the truck until the optimal solution is obtained.

As a preferred solution of the present invention, the penalty cost C _i of the logistics demander i described in step 1 is expressed as:

Among them, T _i represents the time required for the truck to arrive at the logistics demander i, γ ₁ represents the penalty coefficient of the truck arriving in advance to provide services, γ ₂ represents the penalty coefficient of the truck arriving late to provide services, LT _i and ET _i represent the logistics demand, respectively The upper and lower limits of the ideal service time window of party i, lT _i and eT _i respectively represent the upper and lower limits of the maximum service time window that the logistics demander i can accept.

As a preferred solution of the present invention, the constraints described in step 2 are as follows:

Constraint 1: At least one truck is selected to serve all logistics demanders, that is

Q_r表示货车r的额定载货量，Q_i表示物流需求方i的货运需求量；Constraint 2: The finalized truck can complete all cargo collection services, that is

Q _r represents the rated cargo capacity of the truck r, and Qi represents the freight demand of the logistics demander _i ;

Constraint 3: The total weight W _ri of the goods on the truck after the truck r arrives at the logistics demander i to pick up the goods is equal to the total weight of the goods on the truck after the truck r picks up the goods from the previous logistics demander plus the freight demand of the logistics demander i Q _i ;

表示物流需求方i是否由货车r提供服务，

为0-1变量；Constraint 4: The actual cargo volume of the truck r does not exceed its rated cargo capacity, that is

Indicates whether the logistics demander i is served by the truck r,

is a 0-1 variable;

Constraint 5: The total amount of goods collected by all trucks finally determined is equal to the total amount of freight demand of all logistics demanders, namely

Constraint 6: Determine the penalty cost C _i of the logistics demander i according to the time when the truck provides services and the upper and lower limits of the maximum service time window that the logistics demander i can accept.

Compared with the prior art, the present invention adopts the above technical scheme, and has the following technical effects:

The invention refers to the solving steps of the LRP problem, divides the path optimization problem into a plurality of sub-problems and solves it step by step, takes the genetic algorithm as the core, combines the cluster analysis model with the basic concept of the greedy algorithm, obtains a high-quality initial population, and shortens the evolution process. The calculation steps in the iteration are reduced, the calculation efficiency is improved, and the optimal cargo collection path can be obtained more quickly. In this way, the logistics resources of traditional industrial clusters can be more fully utilized, the logistics efficiency can be improved, and the logistics cost can be reduced.

Description of drawings

FIG. 1 is a flow chart of the method for optimizing the collection path of traditional industrial clusters based on the crowdsourcing mode of the present invention.

Figure 2 is a flow chart of mass-contracting logistics for traditional industries.

FIG. 3 is a flow chart of the collection of goods in a traditional industry cluster based on the crowdsourcing mode of the present invention.

FIG. 4 is a basic flow chart of the present invention, which takes genetic algorithm as the core and combines cluster analysis and greedy algorithm to solve the optimal collection path.

Detailed ways

Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, but not to be construed as a limitation of the present invention.

As shown in Figure 1, it is a flow chart of the method for optimizing the collection path of traditional industrial clusters based on the crowdsourcing mode of the present invention, and the method includes the following steps:

S1. Analyze the cost in the process of collecting goods in traditional industrial clusters based on the crowdsourcing model, and take the lowest total cost as the objective function.

货车行驶成本

货车载重成本

时间窗约束惩罚成本

四类，Costs in the process of collecting goods in traditional industrial clusters based on the crowdsourcing model, including: fixed cost of trucks

truck driving cost

Truck weight cost

time window constraint penalty cost

four categories,

in,

Therefore, the minimum total cost objective function is as follows:

为0-1变量，表示客户i是否是货车r第一个服务的客户；d_ij表示客户i到客户j的距离；

为0-1变量，表示货车r是否为客户i服务后立刻为客户j服务；C_w代表货车将单位重量运输单位距离的成本；W_ri代表货车r到达客户i取货后车上货物的总重量；C_i表示客户i的惩罚成本；X_i为0-1变量，表示客户i是否在约定的时间内得到服务；T_i表示货车抵达客户i点所需的时间；γ₁表示货车提前到达提供服务的惩罚系数；γ₂表示货车延后到达提供服务的惩罚系数；[ET_i,LT_i]为客户i理想的服务时间窗；[eT_i,lT_i]为客户i可以接受服务的最大时间窗。i为当下有物流需求的客户，j为当下有物流需求的客户或集群物流中心E，当j＝n+1时表示货车最后将货运回集群物流中心，i与j不可代表同一客户，j≠i，i∈N_i，j∈N_i∪E。Among them, M _r is the set of crowdsourced trucks available in the current cluster, r=1,...,m; C _g represents the fixed use and maintenance cost of the truck; X _r is a 0-1 variable, indicating whether the vehicle r is used; Ni is the set of customers who generate freight demand, _i =1,...,n; C _d represents the cost of traveling a unit distance when the truck is empty; d _ri represents the distance from truck r to customer i;

is a 0-1 variable, indicating whether customer i is the first customer served by truck r; d _ij represents the distance from customer i to customer j;

It is a 0-1 variable, indicating whether truck r serves customer j immediately after serving customer i; C _w represents the cost of transporting unit weight by the truck per unit distance; W _ri represents the total amount of goods on the truck after truck r arrives at customer i to pick up the goods. weight; _Ci represents the penalty cost of customer i; X _i is a 0-1 variable, indicating whether customer i is served within the agreed time; Ti represents the time it takes for the truck to reach customer _i ; γ ₁ represents the truck arrives ahead of schedule Penalty coefficient for providing services; γ ₂ represents the penalty coefficient for trucks arriving late to provide services; [ET _i ,LT _i ] is the ideal service time window for customer i; [eT _i ,lT _i ] is the maximum service time that customer i can accept Time Window. i is a customer with current logistics needs, j is a customer with current logistics needs or cluster logistics center E, when j=n+1, it means that the truck will finally return the goods to the cluster logistics center, i and j cannot represent the same customer, j≠ i, i∈N _i , j∈N _i ∪E.

S2. When customer demand, customer location, and crowdsourced vehicle location are known, take customer demand, vehicle capacity, and time window requirements as constraints.

1) Select at least one truck to serve all logistics demanders,

其中Q_i表示客户i的货运需求量；Q_r表示车辆r的额定载货量。2) The final selected truck has the ability to meet all customer needs and complete all cargo collection services,

Where Q _i represents the freight demand of customer i; Q _r represents the rated cargo capacity of vehicle r.

3) After the truck r arrives at the customer i to pick up the goods, the total weight W _ri of the goods on the truck is equal to the total weight of the previous point plus the freight demand quantity Q _{i of the customer i} .

其中

为0-1变量，表示客户i是否由货车r提供服务。4) The actual collection volume of the vehicles on each route does not exceed the rated load capacity, that is to ensure the rationality of the distribution of goods collection tasks for each truck,

in

is a 0-1 variable indicating whether customer i is served by truck r.

5) The total collection weight of each vehicle is equal to the freight demand of all customers, that is to ensure that all goods are collected,

6) Based on the constraint condition of the penalty cost of the time window, determine the penalty cost C _i of the customer i according to the time when the truck provides service.

S3. According to the location of the available trucks in the crowdsourcing mode and the location and demand of the logistics demander, use the cluster analysis and greedy algorithm to select the vehicle and the customer, determine the vehicle used in this route optimization and arrange the route .

First, use cluster analysis to perform initial clustering, that is, to cluster the customers near the truck with the truck, and according to the principle of limited distance, cluster all the vehicles and customer nodes, and generally assign customers with similar distances to them. near the truck. At the same time, greedy thinking is applied to determine the minimum number of trucks and determine the choice of vehicles. Next, secondary clustering is performed on the determined trucks and customers, and the customers are clustered according to the distance and the freight capacity of the trucks and are allocated to the above determined trucks.

S4. Using the genetic algorithm, calculate the solution of the previous step as the initial population, optimize the path, and save the plan.

Secondly, select the vehicle path, according to the clustering scheme obtained in the previous steps, and determine the path of the truck with the shortest total distance as the goal, which is the initial population of the genetic algorithm, and then based on the current clustering scheme, with the lowest cost as the target application The genetic algorithm performs path optimization. During the optimization process, the number of vehicles and the vehicle selection scheme are adjusted, the lowest cost is calculated, and the lowest cost scheme is saved.

S5, verifying and comparing the obtained different schemes, and obtaining an optimal path scheme.

Finally, the obtained scheme is verified, that is, all constraints are judged one by one. When all constraints are met, it is the optimal solution. Otherwise, the truck selection scheme and driving path need to be changed. Finally, through the optimization process, different schemes are evaluated The optimal solution is obtained by the comparison and selection of the solutions, and the final required truck selection and route optimization scheme are obtained.

As shown in Figure 3, in order to meet the logistics needs of traditional industrial clusters, goods are collected based on the crowdsourcing model.

According to the logistics needs of the D home textile industry cluster at a certain moment, the collection path is optimized. Taking D logistics park as the center point, the positions of 20 customers are placed in the coordinate system of 100*100, and the distance between two adjacent coordinates is 1 km. The customer's location, logistics demand and maximum service time window are known, such as shown in Table 1.

Table 1

Randomly generate 5 trucks with different positions in the coordinate system. The trucks are of the same model and have a rated cargo capacity of 5 tons. The average vehicle speed is 50 km/h. Based on the actual situation, set the fixed use and maintenance cost C of the truck. _g is 90 yuan/vehicle; the cost C _d of the unit distance traveled by the truck is 3 yuan/km when the truck is empty; the cost C _w per unit distance of the truck transporting the unit weight is 0.9 yuan/ton kilometer; the penalty cost C _i for customer i is set 20 yuan/minute.

The randomly generated vehicle locations are shown in Table 2, where the point numbered 0 is the D logistics park. Taking the lowest total cost of collection as the objective function:

Table 2

Set constraints from the perspective of customer demand, vehicle capacity, and time windows.

As shown in Figure 4, the improved genetic algorithm is used to solve the problem by combining cluster analysis and greedy algorithm. The initial clustering is performed first, and then the path optimization is performed on the initial scheme. The scale of the genetic race is set to 100 and the number of iterations to 200. The final path optimization results are shown in Table 3.

table 3

It can be seen from Table 3 that the trucks numbered 1, 2, 4, and 5 were finally selected to serve 20 customers. The goods loaded by each crowdsourcing truck did not exceed the rated cargo capacity, and the time window requirements of each customer were obtained. Satisfaction, the total service cost after path optimization is 2705.52 yuan. That is to say, the above result is the optimal collection path of the traditional industrial cluster.

The above embodiments are only to illustrate the technical idea of the present invention, and cannot limit the protection scope of the present invention. Any modification made on the basis of the technical solution according to the technical idea proposed by the present invention falls within the protection scope of the present invention. Inside.

Claims (3)

1. A traditional industry cluster collection path optimization method based on a crowdsourcing mode is characterized by comprising the following steps:

step 1, analyzing the cost in the traditional industrial cluster cargo collection process, and establishing a target function with the lowest total cost as a target;

the objective function is as follows:

wherein C represents the total cost; c¹、C²、C³、C⁴Respectively representing the truck fixed cost, the truck running cost, the truck loading cost and the time window transportation punishment cost; c_gRepresenting fixed use and maintenance costs, X, of a single truck_rIndicating whether truck r is in use, X_rIs a variable from 0 to 1, r is 1, …, m represents the number of available trucks; c_dRepresenting the cost per unit distance travelled by a single truck when empty, d_riIndicating the distance of the truck r from the logistics consumer i,

indicating whether the logistics consumer i is the first customer serviced by truck r,

is a variable from 0 to 1, i is 1, …, n, n represents the number of logistics demand side; d_ijIndicating the distance from the logistics demand side i to the logistics demand side j,

indicating whether the truck r serves the logistics demand side i and then serves the logistics demand side j,

the variable is 0-1, j is 1, …, n +1, j is not equal to i, j is n +1, which indicates that the truck returns the freight to the cluster logistics center; c_wRepresenting the cost per unit distance a single truck will transport per unit weight, W_riThe total weight of goods on the truck after the truck r arrives at the logistics demand party i for taking goods is represented; c_iPenalty cost, X, representing the logistics demander i_iIndicating whether the logistics demand party i gets the service in the appointed time, X_iIs a variable from 0 to 1;

step 2, setting constraint conditions under the condition of knowing the position of a logistics demander, the freight demand of the logistics demander, the position of a truck and the rated freight volume of the truck;

step 3, taking the positions of the trucks as a clustering center, performing initial clustering on logistics demanders by adopting a clustering analysis method, simultaneously applying greedy thought, determining the lowest truck number and the corresponding truck, and simultaneously determining the logistics demander corresponding to the service of each truck;

step 4, performing secondary clustering by adopting a clustering analysis method according to the lowest truck number and the corresponding truck determined in the step 3, and distributing logistics demand parties to the determined trucks;

step 5, determining the path of the truck by taking the shortest total distance as a target according to the clustering result of the step 4, taking the path of the truck as an initial population of a genetic algorithm, optimizing the path of the truck by adopting the genetic algorithm by taking the lowest total cost as a target based on the clustering result of the step 4, and adjusting the number of the trucks and logistics demand parties corresponding to services of each truck in the optimization process, so as to obtain the number of the trucks corresponding to the lowest total cost and the path of each truck, namely a scheme with the lowest total cost;

step 6, judging the lowest total cost scheme obtained in the step 5 by adopting the constraint conditions set in the step 2, wherein when the lowest total cost scheme obtained in the step 5 meets all the constraint conditions, the scheme is the optimal solution; otherwise, adjusting the logistics demand side of the corresponding service of each truck and the path of the truck until the optimal solution is obtained.

2. The crowd-sourcing mode-based conventional industry cluster cargo gathering path optimization method as claimed in claim 1, wherein the penalty cost C of the logistics demand party i in step 1_iExpressed as:

wherein, T_iIndicates the time, gamma, required for the truck to arrive at the logistics party i₁Penalty factor, gamma, representing the advance arrival of a truck to provide service₂Penalty factor, LT, indicating the delayed arrival of a truck to provide service_i、ET_iRespectively representing the upper limit and the lower limit, lT, of the ideal service time window of the logistics demand party i_i、eT_iRespectively representing the upper limit and the lower limit of the maximum service time window which can be accepted by the logistics demand party i.

3. The conventional industry cluster cargo gathering path optimizing method based on crowdsourcing mode as claimed in claim 1, wherein the constraint conditions in step 2 are as follows:

constraint 1: at least one truck is selected to serve all the logistics consumers, i.e.

Constraint 2: the final determined truck being able to perform all the collecting services, i.e.

Q_rIndicating the nominal cargo capacity, Q, of the truck r_iRepresenting the freight demand of the logistics demand party i;

constraint 3: the total weight W of goods on the truck after the truck r arrives at the logistics demand party i for taking goods_riEqual to the total weight of goods on the truck after the truck r takes goods from the previous logistics demand party plus the freight demand Q of the logistics demand party i_i；

Constraint 4: the actual cargo-collecting capacity of the truck r not exceeding its nominal capacity, i.e.

Indicating whether the logistics customers i are serviced by the trucks r,

is a variable from 0 to 1;

constraint 5: the final determined total cargo collection amount of all trucks is equal to the total cargo demand amount of all logistics demand parties, namely

Constraint 6: determining the punishment cost C of the logistics demander i according to the service providing time of the truck and the upper limit and the lower limit of the maximum service time window which can be accepted by the logistics demander i_i。

Images