Research on Partner Selection of High-Speed Railway Dynamic Logistics Alliance Based on the Dynamic Programming Model

Wu, Haixia; Wang, Yinghui; Zhang, Pei; Yu, Jiang; Du, Yuxuan

doi:https://doi.org/10.1155/2023/6987845

Journal of Advanced Transportation

On this page

Abstract Introduction Literature Review Methods Conclusion Appendix Data Availability Conflicts of Interest Acknowledgments References Copyright Related Articles

Research Article | Open Access

Volume 2023 | Article ID 6987845 | https://doi.org/10.1155/2023/6987845

Research on Partner Selection of High-Speed Railway Dynamic Logistics Alliance Based on the Dynamic Programming Model

Haixia Wu,¹Yinghui Wang,²Pei Zhang,²Jiang Yu,³and Yuxuan Du²

Academic Editor: Luca D’Acierno

Received24 May 2022

Revised11 Dec 2022

Accepted14 Dec 2022

Published14 Jan 2023

Abstract

Traditional railway transportation can no longer meet people’s demand for logistics services. This paper takes advantage of high speed to propose a high-speed railway dynamic logistics alliance based on a cloud platform to make up for the lack of transport capacity at both ends of the high-speed railway logistics trunk line. Selecting partners is crucial to the high-speed rail logistics alliance. This paper uses the methods of multiobjective fuzzy optimization and dynamic programming to conduct multistage optimization of high-speed railway dynamic logistics alliance partners. When the market demand changes, in order to optimize the overall interests of the alliance, this paper uses the efficiency profit field method to achieve the dynamic selection of alliance partners or potential partners. The case study shows that the establishment of the high-speed railway dynamic logistics alliance can optimize the interests of the members of the alliance, verify the effectiveness of the method, and provide a reference for the better development of high-speed railway logistics.

1. Introduction

Railway freight transport is an important part of the modern transportation system. It plays a key role in the entire transportation field. China’s railway freight transport is one of the basic modes of land transportation. Between 2010 and 2021, the total annual freight volume reached over 3.9 billion tons. As a railway transportation mode with Chinese characteristics, high-speed rail logistics is constantly developing with the growth of the freight market.

The first attempt at high-speed rail express transportation in China was implemented by the Guangzhou Railway Group on the Wuhan high-speed rail line. However, it was not until the promulgation of the interim measures for the administration of high-speed rail express transportation in 2013 that the high-speed rail enterprises involved in the express delivery business were equipped with corresponding document guidance. In the same year, the Beijing-Guangzhou line and the Beijing-Shanghai line also launched high-speed rail express trains. The partners included high-quality express enterprises such as SF Express and UPS. In 2016, China Railway Corporation provided high-speed rail express services. By 2020, the high-speed rail express business volume in China will increase by 40% year over year. There are 980 lines in China, covering more than 80 cities [1]. Such a vast freight market has created business opportunities for high-speed rail logistics. High-speed rail logistics has the outstanding advantages of “quickness, accurateness, and stability,” mainly transporting high-value-added products such as medical machinery and fresh products. Since the modern logistics system is a long industrial chain, high-speed rails can only realize part of the functions of this chain and cannot complete the logistics services at both ends of high-speed rails. It is difficult for high-speed rails to take the lead in the modern logistics market by doing it alone [1]. Therefore, based on its own advantages and disadvantages, the establishment of a high-speed rail dynamic logistics alliance led by high-speed rails is a great breakthrough against the status quo.

A high-speed rail dynamic logistics alliance means that high-speed rail logistics enterprises temporarily cooperate with other enterprises to build a dynamic logistics alliance organization to grasp a certain market opportunity. Based on information network technology and through the integration of high-speed rail equipment, infrastructure, and superior logistics resources of cooperative enterprises, a temporary logistics network structural organization with complementary advantages, risks, and resource sharing is formed. The alliance responds to market challenges with its overall advantages, with a view to seizing market opportunities and increasing economic returns. It is a feasible and effective measure to form a high-speed rail dynamic logistics alliance, integrate the logistics resources of multiple enterprises, realize the complementary advantages between enterprises, and enhance market competitiveness, so as to achieve seizing market opportunities and increasing economic returns.

2. Literature Review

The preparation of a high-speed rail dynamic alliance based on a cloud platform has attracted some scholars’ attention. The construction of the alliance based on a cloud platform is essential to integrate the high-quality service resources of various enterprises into the cloud so that the reasonable scheduling and effective supervision of resources can be realized, thereby reducing the cost of related services and effectively improving the resource efficiency, safety, and timeliness of the service. The selection of alliance partners is essential to the survival and development of the alliance and is a key step in the alliance’s life cycle. Scholars have conducted a lot of research on logistics enterprise alliances and partner selection and have achieved rich results.

For the research on the construction of dynamic alliances by logistics enterprises, Du and Qu [2] proposed a cross-border logistics alliance model based on information collaboration, built a three-way evolutionary game model of information collaboration between cross-border e-commerce platforms, domestic logistics enterprises, and foreign logistics enterprises, and conducted dynamic simulation analysis on the three-way information collaboration strategy; Luo et al. [3] built a quadratic programming model based on the square distance and least square method of interval numbers to minimize the sum of the square excess of the interval numbers of all players in the largest cooperative cold chain logistics alliance and then determined the income distribution value of each player in the alliance; Li et al. [4] proposed to apply “expressway + high-speed railway” as a fast intermodal transport mode to the multimodal transport of fresh products in view of the transport characteristics of small batches and the high timeliness of fresh products. Taking the lowest total cost as the goal, a high-speed rail service time window, and customer satisfaction as constraints, a combination model of road-rail intermodal transport route selection and transport mode for fresh products based on fuzzy demand is constructed; Jauhar et al. [5] proved that the efficiency of the third-party reverse logistics (3PRLP) may have an impact on the selection of third-party reverse logistics partners and the orders assigned to them through the two-stage method. The first stage includes a method combining data envelopment analysis and a differential evolution algorithm to improve discrimination and distinguish 3PRLPs according to efficiency. In the second stage, use the multiobjective model to allocate orders to 3PRLP. Two solutions are adopted to solve the proposed multiobjective model and find the Pareto optimal solution.

Railway transportation is one of the main modern transportation modes. In order to transform traditional logistics into modern logistics, many experts have conducted research on high-speed rail logistics services. D’Acierno et al. [6, 7] and other scholars analyzed how to select the best strategy to minimize the impact on users when the railway system fails; Gallo et al. [8], taking into account the costs of all transportation systems (cars, buses, and railways) and external costs, proposed four different objective functions that can be used to assume that demand is elastic and study the design of regional railway frequency; Zhang et al. [9] proposed to build a modern logistics service system, strengthen the construction of freight brands, optimize the window period of transportation supply quality, embed the railway logistics park into the social logistics system, provide diversified logistics services, accelerate the construction of railway informatization, improve the soft power of railway logistics, and accelerate the transformation of railway freight to modern logistics; Rao et al. [10] used the double difference model to prove that the city where the enterprise is located has a significant impact on the distribution of suppliers after the high-speed railway is opened; Ren et al. [11] used the double difference tendency score matching method to prove that the opening of high-speed rail has promoted the development of the logistics industry in cities along the line on the whole; Han and Chao [12] analyzed the indirect freight effect of high-speed rail construction using the enterprise inventory cost as the starting point. According to the different transportation modes of enterprises, they built a triple-difference model. The results showed that the impact of high-speed rail on the inventory level increased with the increase in enterprises’ dependence on railway transportation.

How to optimize the selection of dynamic alliance partners is the key to determining the quality of the alliance. There are many research methods for optimization in academia. This paper mainly discusses the literature review from the following two aspects: freight optimization and database use optimization.

In freight optimization research, Wang et al. [13] established a multiobjective optimization model for a high-speed railway station emergency evacuation path with the goal of crowding degree and total evacuation time in the process of high-speed railway station emergency evacuation and solved it through an improved adaptive quantum ant colony algorithm. Ma and Liao [14] used a heuristic iterative algorithm to solve the integer programming model of high-speed rail express base location and scale optimization based on the service network. Wang and Wu [15] built an emergency logistics model based on cloud platforms and optimized the emergency logistics path based on cloud platforms through a genetic algorithm. Li [16] proposed a hybrid genetic algorithm combining the k-means algorithm and cluster analysis to solve the problem of multiparty collaborative logistics distribution path optimization. Camur et al. [17] used the new heuristic algorithm of the rolling time range to achieve the optimization frame of an efficient and sustainable logistics operation through transportation mode optimization and shipment integration. Song et al. [18] handled the optimization of emergency logistics through an in-depth learning model based on big data.

On the optimization of database use, Crispim et al. [19] proposed using the TOPSIS method to sort and select candidate partners; Wang Daoqu [20] combined AHP and TOPSIS to achieve partner selection; Gong et al. [21] used the adaptive dynamic programming method to research and solve the online solution of the network multiagent pursuit and evasion game so that each agent can obtain the strategy to achieve Nash equilibrium in real time; Liang and Xu [22] established a finite-time domain Markov decision process model with the goal of maximizing the benefits of the hospital in terms of inspection equipment, and combined it with the dynamic programming theory to obtain the optimal reservation scheduling strategy of the system. In order to adapt the warehouse to the increasing variety and quantity of storage products, Djurdjević et al. [23] used the dynamic programming method to obtain the optimal allocation of products in different order-picking areas. In order to solve the optimal control problem in the full driving scenario of hybrid electric vehicles, Bao et al. [24] proposed an adaptive dynamic programming method to solve the interpolation error problem and obtained the solution without theoretical error at the cost of small driving cycle accuracy. Lan et al. [25] developed a heuristic algorithm combined with the dynamic programming algorithm to generate a task attribute list for each doctor when studying the doctor scheduling problem, so as to ensure the optimization of the solution. Hu et al. [26] proposed a multiobjective planning algorithm based on pruning rules to solve the problem of the dimension disaster of dynamic programming algorithms: prune the dominated state by pruning rules and reduce the search space to improve the calculation efficiency.

Existing literature has conducted meaningful research and exploration on the general dynamic alliances of logistics enterprises and their partners selection of dynamic alliances. However, there is a lack of research on high-speed rail logistics enterprises as the core of dynamic alliances. It can be seen from the literature that the dynamic programming method has the advantage of being able to obtain the global optimal solution and is often used to solve optimization problems, but there are few articles on the optimization of the selection of alliance partners. Due to the problem of “dimension disaster” in dynamic programming, this paper proposes to conduct multistage optimization of high-speed railway dynamic logistics alliance partners with the goal of maximizing the overall profit of the alliance based on the realistic background that big data cloud platforms have been widely used, using the method of combining multiobjective fuzzy optimization and dynamic programming. When the market demand changes, in order to optimize the overall interests of the alliance, this paper uses the efficiency profit field method to achieve the dynamic selection of alliance partners or potential partners. In this way, the dynamic selection of partners and the dynamic entry and exit mechanisms of alliances can be formed.

3. Analysis of Partners Selection of High-Speed Rail Dynamic Logistics Alliance

3.1. The Formation of High-Speed Rail Dynamic Logistics Alliance

High-speed rail logistics dynamic alliance is task-oriented. It uses cloud platforms as the infrastructure. The specific formation process of the alliance is that after the cloud platform owned by the high-speed rail logistics enterprise obtains the customer’s transportation demands, based on the characteristics of the task, among all the partners that the high-speed rail logistics enterprise can obtain, the enterprise that best meets the overall profit maximization will be selected and formed into the alliance to serve the transportation. Through the formation process, it can be seen that the alliance with high-speed rail logistics enterprises as the core emerges with tasks. Different partners are determined according to different cooperation goals, and sometimes it is necessary to run multiple alliances simultaneously at the same time. This requires the use of cloud platforms to efficiently and orderly process large amounts of data and realize information sharing and collaboration among alliance members. The cloud platforms have powerful distributed processing and storage capabilities, so they can quickly process the information between the partners of the dynamic logistics alliance and realize real-time information sharing between alliance members. Meanwhile, because of the large amount of storage, the stored information is more comprehensive and specific and contributes to the overall optimization of the alliance, which avoids the problem that traditional information systems can only issue decisions based on one-sided information [27].

3.2. The Partners Selection Process of High-Speed Rail Dynamic Logistics Alliance

In the formation of a high-speed rail dynamic logistics alliance, partner selection is extremely important as a link throughout. This paper divides the partner selection for the alliance based on a cloud platform into three phases.

Phase I is the classification of partners. Before the transportation task is launched, the cloud platform will classify the partners based on the information collected from high-speed rail logistics enterprise partners combined with various quantitative and qualitative indicators so that when the demand for the alliance’s construction arises, feasible enterprises can be selected efficiently.

Phase II is the selection of partners. To allow the alliance partners to collaborate to achieve the effect of “1 + 1 > 2,” a comprehensive evaluation was carried out on the basis of the classification of the proposed cooperative enterprises according to the logistics and transportation stages in the classification of partners so as to achieve the globally optimal goal [28, 29]. Phase II is divided into the following two steps:

Step one is to comprehensively consider the quantitative and qualitative indicators that need to be evaluated in the partners’ selection of the alliance and determine the relative superiority of each candidate enterprise in each link based on the multiobjective fuzzy optimization model.

Step two is to use the basic idea of the dynamic programming method: first, divide the target logistics tasks completed by the alliance’s plan into stages according to the various links that need to be completed and transform the partners’ selection into a multistaged optimal combination decision; second, by adopting the method of maximizing the sum of the relative superiority of the decision sequence, calculate and solve the optimal combination of the alliance partners.

Phase III is the dynamic selection of partners. When the market demand changes and the selected partner combination is not sufficient to meet the market demand, the alliance members are dynamically replaced through the efficiency profit field model by setting the critical point for entering and exiting the alliance. Enterprises that meet the conditions will enter the alliance organization, and those that do not meet the conditions will exit.

4. Methods and Models for Partners Selection of the High-Speed Rail Dynamic Logistics Alliance

4.1. Multiobjective Fuzzy Optimization Method

Suppose that the high-speed rail dynamic logistics alliance needs to go through s links to complete a target logistics task and an appropriate partner must be selected in each link to complete the task. Suppose that the collection of candidate partners for all links is , where is the collection of candidate partners for the link. Suppose that there are a total of indicators in the evaluation indicator system selected by the partner and is the collection of the qualitative and quantitative evaluation targets. Let 1 to h be the quantitative evaluation indicator and to be the qualitative evaluation indicator. The m target feature matrices of collection in the link are

In equation (1), ; ; ; element represents the evaluation value of j^th candidate partner t_sj in the s^th link with the i^th quantitative target .

4.1.1. Analysis of Relative Superiority of Qualitative Targets [27, 30]

According to fuzziness and relativity, set as the relative superiority of the eigenvalues of the sample indicator with respect to the fuzzy concept [31]. The equations are as follows:(1)If the indicator belongs to the smaller the better, then(2)If the indicator belongs to the higher the better, then The nondimensional relative superiority matrix of the quantitative targets is as follows: In equation (4), ; ; ; ; represents the relative superiority matrix of the candidate plan in the s^th link with the quantitative target.

4.1.2. Analysis of Relative Superiority of Qualitative Targets [27, 30]

Through binary comparison of the superiority of each qualitative target of different candidate partners, sort and compare in turn, determine the mood operator, and then obtain the relative superiority matrix of qualitative targets according to Tables 1. The specific steps are as follows.

The qualitative ranking scale of superiority obtained by binary comparison takes values between 0, 0.5, and 1.0. The following equation can be obtained:

Equation (5) represents the qualitative ranking scale of the superiority binary comparison between the candidate partner and the candidate partner in the link with the quantitative target. If the superiority of the partner for the qualitative target is greater than the superiority of the partner for the qualitative target and then, if both superiorities are the same, then . The binary comparison matrix can be derived as follows:

When , is the superiority ranking consistency scale matrix. If , the ranking is required to be recalculated. When meets the criteria, sort the rows and columns from the largest to the smallest, and then obtain the relative superiority matrix of qualitative indicators according to the mood operator in Table 2, as shown in the equation (5).

In the abovementioned equation, represents the relative superiority matrix of the candidate plan in the link with the quantitative target.

Through and , the relative superiority matrix of all quantitative and qualitative evaluation targets of the candidate partners in the link is as follows:

In equation (8), ; ; ; ; represents the relative superiority matrix of the j^th candidate plan in the link with the quantitative target.

4.1.3. Analysis of Target Weight Vector

Different factors have different effects on target selection. To ensure that target selection is reliable and evidence-based, imitate the abovementioned method of determining the superiority of qualitative and quantitative targets to determine the proportion of each factor. That is, the “importance” of the fuzzy concept replaces the concept’s “excellence” in order to obtain the relative superiority degree of each target to the importance. After normalization, the weight vector of the target can be obtained as follows:

is the importance scale calculated according to the relationship between the mood operator and the quantitative scale, and the relative superiority degree is obtained after the qualitative indicator uses the binary comparison method to obtain the superiority ranking consistency scale matrix.

4.1.4. Determination of the Relative Superiority of Candidate Partners [27, 30]

To facilitate comparison, the reference benchmark must be unified. According to the relativity of optimization, the maximum value of the relative superiority of each indicator of candidate partners in all S links can be taken as the superior solution of the decision vector, and the minimum value as the inferior solution, which are represented, respectively, as and . Therefore, in the link, for the determined comprehensive relative superiority matrix and the target weight vector, according to the basic theory of fuzzy optimization [32], the comprehensive relative superiority of all evaluation objectives of the candidate partner in the link can be obtained as follows:

In the given equation ; .

The main purpose of the fuzzy optimization stage of candidate partners is to select candidates in a unified and comparable form through the fuzzy optimization method and to determine the relative superiority of candidate enterprises in each link.

4.2. Dynamic Programming Method

(1)Construction of a Decision Matrix [27, 30]. According to the dynamic programming method, a problem needs to be considered as being composed of different stages. Therefore, logistic links are regarded as stages; is defined as the state collection in dynamic programming, representing the collection of candidate partners in the stage, in which; represents the decision of the stage; represents the relative superiority of the candidate partners in the stage, which is determined by the decision of this stage and the state of the previous stage. Using the multiobjective fuzzy optimization method in 4.1, the relative superiority of candidate partners’ decision matrix in the s^th stage is obtained as follows:In the given equation, represents the relative superiority of the candidate partners when the state in the stage is and the decision is in the stage.(2)Solution of Optimal Partner Combination [33]. According to the dynamic programming theory and the principle of maximum superiority, determine the objective function of the dynamic programming model and select the objective function of the fuzzy optimization dynamic programming with the largest sum of the relative superiority in stages. The function is as follows:According to the abovementioned dynamic programming objective function and the dynamic programming optimization principle, the recursive equations of the relative superiority of candidate partners are obtained as follows:In the given equation, represents the maximum value of the sum of relative superiority of candidate partners in s stages; represents the relative superiority of the candidate in the stage; and represents the maximum value of the sum of relative superiority of the candidate partner in the previous stage of the stage.

Through continuous recursion of equation (14), the optimal solution (or optimal partner combination) at each stage is .

The dynamic programming method is used to select partners, which corresponds to the second stage of partner selection, namely, the optimal selection stage, that is, to compare all the combinations of all proposed cooperative enterprises one by one. The purpose is to select the globally optimal plan through comparison, so as to determine the partners of the alliance.

4.3. Construction of Efficiency Profit Field Model

Enterprises construct a high-speed rail dynamic logistics alliance to meet the needs of the modern logistics market and to obtain profits while ensuring the efficient completion of logistics tasks. In the dynamic partners’ selection, this paper makes innovations based on the concept and model of the profit field. Since the outstanding advantage of the alliance is high efficiency, this paper adds the “unit logistics service time” indicator that reflects efficiency when constructing an efficiency profit field model to show the different proportions of various influencing factors in partners selection. Influencing factors are reflected in the number of indicators. The specific indicators are shown in Tables 3 and 4.

4.3.1. Profit Distance

The profit distance D is mainly affected by the comprehensive service strength (SC) and the enterprises’ operation status (RP). The specific values of SC and RP are derived from the indicator system in Tables 3 and 4, where each indicator is divided into 5 grades using a 5-point system. The 1 to 5 grades correspond to 1 to 5 points. The full score of the SC including 12 measurement indicators is 70 points, and the full score of the RP including 9 measurement indicators is 45 points. The higher the score is, the more popular it is. The grade value of each measurement indicator is determined by the threshold value of the superiority of each target of the enterprise calculated above. Therefore, the equation for calculating the enterprise’s profit distance is

In the given equation, represents the comprehensive service distance and represents the distance of the enterprise’s operation status.

4.3.2. Profit Attraction [34]

The quality of the alliance and the proposed cooperative enterprises is determined by the types of logistics resources they own and market demand. This paper uses the vector L to represent the set of n logistics resources. In the equation, () represents the number of resources i; vector Y represents the set of resource utilization. The resource utilization rate is expressed by whether the enterprise can provide the resource. means the enterprise cannot provide resources, and means the enterprise can provide resources. Since market demand is dynamically changing, the alliance and the proposed cooperative enterprises have different resource supply capabilities in different time periods. Thus, the time vector is introduced.

The equation of the resource quality of the alliance is

Suppose the collection of resource income of the alliance is. Different resources provide different incomes. Therefore, resources provide kinds of incomes, which is expressed by a vector as

In the equation, represents the resource income corresponding to resource. The profit quality is

The calculation of the profit quality of the proposed cooperative enterprise is similar to that of the alliance:

In the given equation, represents the resource quality of the proposed cooperative enterprises, represents the profit quality of the proposed cooperative enterprises, and represents the collection of the resource income of the proposed cooperative enterprise.

4.3.3. Willingness Resistance

The willingness resistance is expressed by the sum of risk cost and the opportunity cost of the proposed cooperative enterprise joining the alliance as

Suppose the profit increase of the proposed cooperative enterprises after joining the alliance is , and represents the increase in profit per unit resource of the enterprise after joining the alliance under current market condition. Then,

The strength of the profit field is

The calculation formula of the profit attraction FG is

4.3.4. Construction of Efficiency Profit Field Model

Based on the abovementioned analysis, the efficiency profit field model of the alliance is determined, as shown in Figure 1. The dotted circles in Figure 1 represent the radiation range of different profit field strengths; is the profit radius, and represents the enterprises in different profit circles. The construction of the efficiency profit field model is as follows:

The constraints are

The objective function equation (24) indicates that the alliance has the greatest profit quality. The constraint equation (25) shows that the profit distance of the enterprise should be smaller than the profit distance threshold of the alliance, where represents the profit distance threshold. The constraint equation (26) illustrates that the profit attraction of the alliance is greater than the enterprise’s willingness resistance. The dual constraints of equations (25) and (26) form the trigger point for the proposed cooperative enterprises to enter into and exit from the alliance, which reflects the two-way choice between the alliance and the enterprise.

4.3.5. Mechanism of Dynamic Selection

When market demand changes, to ensure the alliance’s advantage in the market, the alliance will adjust its internal alliance members. During the whole process, the profit quality and the strength of the alliance will change correspondently. The change in profit quality is

The profit field, then, turns into

The willingness resistance of the enterprise is given as follows:

After entering the algorithm program of the above method into the cloud platform’s background system, the leader of the alliance only needs to use the client side of the cloud platform to complete the automated selection of partners.

5. Case Analysis

5.1. Case I

Assuming that a high-speed rail logistics enterprise collects a batch of supply information through the cloud platform. To seize the opportunity, the enterprise prepares to form a high-speed rail dynamic logistics alliance. Through analysis, it is learned that, in addition to high-speed rail trunk line transportation, three stages are needed to complete this logistics task. The enterprise plans to select one company to cooperate with at each stage. The selected cooperative companies are SF Express, JD.com, YTO Express, and Yunda Express. The first-stage companies collected from society through the cloud platform are represented by ; the second-stage companies are represented by ; the third-stage companies are represented by . The specific evaluation indicators for selection are information technology level (the information technology input is expressed by the amount of capital input), operation efficiency (expressed by the task completion time), logistics service level, enterprise compatibility, and business performance, which are expressed by . The level of input resources and work efficiency are quantitative indicators; the smaller the better, and the rest are qualitative indicators. The specific evaluation indicator data are shown in Table 5.

Table 5 shows the data information of the simulation candidate company and uses the model in the previous chapter to select partners.

Take this case as a dynamic programming problem and divide the stages according to the links. It is divided into three stages.

The first stage is to determine the relative superiority of the quantitative target using formula (2); the relative superiority of the quantitative target matrix of the four companies in the first stage is obtained as follows:

The second stage is to determine the relative superiority of the qualitative target, taking the logistics service level as an example. After binary comparison among the four companies, the following consistent scaling matrix is obtained:

According to the superiority ranking , which means that T₁₁ is better than T₁₃, T₁₃ is better than T₁₄, and T₁₄ is better than T₁₂. According to Table 5, the relative superiority of the logistics service levels of the four companies is

Similarly, the relative superiority of the other two qualitative indicators can be obtained as follows:

Based on the abovementioned results, the relative superiority matrix of all quantitative and qualitative evaluation targets in the first stage can be obtained as follows:

Take the target weight vector. Assuming and , through the equation (11), the relative superiority of the four enterprises in the first stage is , , , and .

According to the above steps, the relative superiority matrix of the second and third stages is obtained as follows:

By using equation (14) for recursive calculation, the best optimal combination is (see appendix) as follows:

That is, the first stage selects enterprise T₁₁, the second stage selects enterprise T₂₂, and the third stage selects enterprise T₃₄ to form the partner combination of the alliance.

Through the calculation, it can be concluded that the combination of the optimal candidate enterprises at each stage is not the globally optimal combination: in the first stage, (0.9779); in the second stage, (0,8186); and in the third stage, (0.836 0). The combined superiority of the three stages is , which is not the optimal combination. Therefore, the choice of the alliance should be based on the whole process rather than the optimal combination of various stages. This will maximize the benefits of the alliance.

5.2. Case II

Assuming that the market changes at time , and the partners in the first link of the alliance operation in case I should be updated. High-speed rail logistics enterprises have collected the resource information of the three enterprises through the client side of the cloud platform. It is known that the resource status of the cooperative enterprises and the proposed external cooperative enterprises in the first link of the alliance are shown in Table 6, among which E1 is the enterprise in the alliance and E2, E3, and E4 are the proposed external cooperative enterprises. To ensure that the resources of the proposed external enterprises that have cooperated are reasonably matched with the needs of the alliance, the profit distance to meet the requirements is set to be 7.

The resource utilization rate of the alliance and the proposed cooperative enterprise is shown in Table 7.

Using the random scoring criteria, the SC and RP of the enterprises in the alliance and external partners are obtained as shown in Table 8. At the same time, equation (15) is used to calculate the profit distance between the enterprises in the alliance and the external ones at time t, as shown in Table 9.

As shown in Table 10, the profit attraction and willingness resistance of E2, E3, and E4 are calculated by the equations (16), (17), and (20).

Combining the results of Tables 9 and 10, according to the equations (21) and(22), the change of the partner at time t can be obtained, if the profit distance of enterprise is greater than the ideal profit distance of 7, will exit from the alliance; if the profit distances of enterprises and are greater than the ideal profit distance of 7, they will not choose to enter into the alliance; if the profit distance of enterprise is less than the ideal profit distance of 7 and its profit attraction is greater than willingness resistance, E4 will choose to enter into the alliance, which will realize the dynamic replacement of partners in the alliance.

The dynamic replacement of partners at time t is shown in Figure 2. At time t, the alliance dynamically selects partners; exits the alliance while enters.

(a)

(b)

6. Conclusion

With the rapid development of e-commerce, the demand for high-value-added products is growing, which makes the development prospect of high-speed railway logistics broad. Based on the perspective of the cloud platform, this paper establishes the target system, confirms the weight through the multiobjective fuzzy optimization method, and then determines the order of each partner in a unified and comparable form through the quantitative calculation of the degree of membership. It uses the dynamic programming method to achieve the second stage of partner optimization, the optimization stage, which selects the overall optimal scheme through comparison. In the face of a rapidly changing market, this paper uses the efficiency profit field model to dynamically update partners, set the critical point for enterprises to enter and exit the alliance, and form a profit field-based dynamic entry and exit mechanism for the alliance partners, so that the high-speed rail alliance becomes more flexible. In the future, it can be further considered to add other evaluation indicators to the model to better provide a scientific decision-making basis for the selection of high-speed railway dynamic logistics alliance partners.

Appendix

Example A.1. applies Formula (4−18), recursion, and the recursion results are as follows: Phase I: Phase II: Phase III: Final: To sum up, the optimal combination is

Data Availability

Some or all data, models, or code that support the findings of this study are available from the corresponding author upon reasonable request. The data are from reference 34. The following website is the reference data address: https://kns.cnki.net/kcms/detail/detail.aspx?dbcode=CMFD&dbname=CMFD2010&filename=2010120923.nh&uniplatform=NZKPT&v=LI_klHAPI5cBhErG4sXcbn_6YbapLlGPR8eXf1n-dyZ9Hw2AhQ-YuwNGYi5_V0IU.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

This work was supported by Research on the Design of Beijing Tianjin Hebei Logistics Service Network Based on High-Speed Rail Highway Combined Transport (SCT202003).

References

Y. Zu, “Analysis on the “high-speed rail+e-commerce” mode of high-speed rail freight trains in the context of digital economy,” Urban Rail Transit Research, vol. 24, no. 10, pp. 296-297, 2021.
View at: Google Scholar
Z. Du and Y. Qu, “Research on information collaboration mechanism of cross-border logistics alliance based on tripartite evolutionary game,” China Management Science, vol. 19, pp. 1–12, 2022.
View at: Google Scholar
Y. Luo, P. Cheng, B. Fan, and J. Liu, “The interval number average residual value based on the player’s square excess and its application in the cooperation income distribution of cold chain logistics alliance,” Science Technology and Engineering, vol. 22, no. 12, pp. 4707–4714, 2022.
View at: Google Scholar
Y. Li, M. Qiu, Heiberlin, and F. Zengli, “Road rail intermodal transport route selection of fresh products considering high-speed rail time window and customer satisfaction,” Industrial Engineering, vol. 24, no. 2, pp. 49–56, 2021.
View at: Google Scholar
S. K. Jauhar, S. H. Amin, and H. Zolfagharinia, “A proposed method for third-party reverse logistics partner selection and order allocation in the cellphone industry,” Computers & Industrial Engineering, vol. 162, Article ID 107719, 2021.
View at: Publisher Site | Google Scholar
L. D’Acierno, M. Gallo, B. Montella, and A. Placido, “Evaluation of travel demand impacts in the case of rail system failure,” Procedia - Social and Behavioral Sciences, vol. 87, pp. 75–84, 2013.
View at: Publisher Site | Google Scholar
A. Placido, V. D. Martinis, B. Montella, M. Gallo, and L. D’Acierno, “Effects of travel demand levels on optimal strategies for metro system management in failure contexts,” Procedia - Social and Behavioral Sciences, vol. 111, pp. 819–828, 2014.
View at: Publisher Site | Google Scholar
M. Gallo, B. Montella, and L. D’Acierno, “The transit network design problem with elastic demand and internalisation of external costs: an application to rail frequency optimisation,” Transportation Research Part C: Emerging Technologies, vol. 19, no. 6, pp. 1276–1305, 2011.
View at: Publisher Site | Google Scholar
C. Zhang, Z. Yu, and M. Liu, “Motivation and measures for accelerating railway freight’s transformation toward modern logistics in the new era,” Journal of Beijing Jiaotong University, vol. 18, no. 01, pp. 10–18, 2019.
View at: Google Scholar
P. Rao, D. Wang, and X. Li, “High speed railway opening and supplier distribution decision,” China Industrial Economy, no. 10, pp. 137–154, 2019.
View at: Google Scholar
X. Ren, X. Guo, W. Wang, and M. Xu, “Impact of high-speed rail opening on the development of logistics industry - evidence from 280 prefecture level cities nationwide,” Economic Review of Production, vol. 11, no. 05, pp. 104–121, 2020.
View at: Google Scholar
Li Han and Li Chao, “Evaluation of indirect freight effect of high-speed railway: empirical analysis based on enterprise inventory cost,” Finance and Trade Economy, vol. 42, no. 10, pp. 102–115, 2021.
View at: Google Scholar
F. Wang, H. Xie, Z. Lin, and J. Wang, “Multi objective optimization model and adaptive quantum ant colony algorithm for emergency evacuation of high-speed railway stations under emergencies,” China Management Science, vol. 15, pp. 1–12, 2012.
View at: Google Scholar
C. Ma and W. Liao, “A method for optimizing the location and scale of high-speed rail express base based on service network,” Journal of Railway Science and Engineering, vol. 4, pp. 1–10, 2019.
View at: Google Scholar
Y. Wang and J. Wu, “Research on optimization of emergency logistics path under public security emergencies based on cloud platform,” Technical Economy, vol. 40, no. 12, pp. 162–173, 2021.
View at: Google Scholar
X. Li, “Multiparty coordinated logistics distribution route optimization based on data analysis and intelligent algorithm,” Journal of Sensors, vol. 2022, Article ID 6053332, 9 pages, 2022.
View at: Publisher Site | Google Scholar
M. C. Camur, S. Bollapragada, and A. E. Thanos, “An optimization framework for efficient and sustainable logistics operations via transportation mode optimization and shipment consolidation: a case study for ge gas power,” 2022, https://arxiv.org/abs/2212.03662.
View at: Google Scholar
A. Song, X. Yang, L. Ni, C. Liu, Y. Yao, and L. Pan, “Optimization analysis of the emergency logistics identification method based on the deep learning model under the background of big data,” Wireless Communications and Mobile Computing, vol. 2022, Article ID 2463035, 8 pages, 2022.
View at: Publisher Site | Google Scholar
J. A. Crispim and J. Pinho, Muhieritia Partner Selection in Virtual Interprise, Springer, Boston, Massachusetts, 2017.
W. Daoqu and L. Yun, “Partner selection of technology alliance: research based on AHP-TOPSIS method,” Journal of Xi'an University of Electronic Science and Technology (Social Science Edition), vol. 20, no. 06, pp. 1–10, 2010.
View at: Google Scholar
Z. Gong, B. He, C. Hu, X. Zhang, and W. Kang, “Online adaptive dynamic programming-based solution of networked multiple-pursuer and single-evader game,” Electronics, vol. 11, no. 21, p. 3583, 2022.
View at: Publisher Site | Google Scholar
F. Liang and P. Xu, “Research on optimization method of medical examination appointment scheduling based on MDP and dynamic programming,” Operations Research and Management, vol. 29, no. 05, pp. 17–25, 2020.
View at: Google Scholar
D. Djurdjević, N. Bjelić, D. Popović, and M. Andrejić, “A combined dynamic programming and simulation approach to the sizing of the low-level order-picking area,” Mathematics, vol. 10, no. 20, p. 3733, 2022.
View at: Publisher Site | Google Scholar
S. Bao, P. Sun, J. Zhu, Q. Ji, and J. Liu, “Improved multi-dimensional dynamic programming energy management strategy for a vehicle power-split hybrid powertrain,” Energy, vol. 256, Article ID 124682, 2022.
View at: Publisher Site | Google Scholar
S. Lan, W. Fan, S. Yang, N. Mladenović, and P. M. Pardalos, “Solving a multiple-qualifications physician scheduling problem with multiple types of tasks by dynamic programming and variable neighborhood search,” Journal of the Operational Research Society, vol. 73, no. 9, pp. 2043–2058, 2022.
View at: Publisher Site | Google Scholar
L. Hu, Z. Li, X. Liu, and C. Peng, “Energy saving scheduling method for material handling of assembly line based on dynamic programming,” System Engineering Theory and Practice, vol. 38, no. 9, pp. 2424–2433, 2018.
View at: Google Scholar
X. Tian, D. Wang, and S. Fu, “Research on the construction and operation of cross-border e-commerce logistics alliance cloud platform,” Price Monthly, vol. 5, pp. 68–73, 2018.
View at: Google Scholar
A. Baidu and K. Feng, “Cloud and Taiyuan Railway Bureau officially signed a contract to jointly build a smart logistics cloud platform,” China Logistics and Procurement, vol. 14, p. 18, 2016.
View at: Google Scholar
L. Wang and H. Tian, “Research on the income distribution of the multi entity dynamic alliance of the digital publishing industry chain,” Publishing Wide Angle, vol. 7, pp. 22–27, 2022.
View at: Google Scholar
X. Liu, X. Luo, P. Ji, and Z. Zhang, “Research on dynamic planning of tourists’ journey based on real time information,” China Management Science, vol. 3, pp. 1–11, 2021.
View at: Google Scholar
J. B. Gurav and D. G. Regulwar, “Multi-objective fuzzy optimization for sustainable irrigation planning,” H2Open Journal, vol. 3, no. 1, pp. 373–389, 2020.
View at: Publisher Site | Google Scholar
Y. Qi and S. Lu, “Research on decision support model of electronic resource procurement based on dynamic programming,” Digital Library Forum, vol. 11, pp. 27–32, 2020.
View at: Google Scholar
Y. Shan, Research on the Construction and Key Problems of Railway Logistics Dynamic Alliance, Beijing Jiaotong University, Beijing, China, 2010.
M. Guo, G. Nie, and D. Chen, “Research on dynamic selection mechanism of cloud alliance partners based on profit field,” Computer Application Research, vol. 5, pp. 1367–2136, 2017.
View at: Google Scholar

Copyright

Copyright © 2023 Haixia Wu 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.

PDF Download Citation

Download other formats

Order printed copies