1. bookAHEAD OF PRINT
Detalles de la revista
License
Formato
Revista
eISSN
2444-8656
Primera edición
01 Jan 2016
Calendario de la edición
2 veces al año
Idiomas
Inglés
access type Acceso abierto

Optimal Model Combination of Cross-border E-commerce Platform Operation Based on Fractional Differential Equations

Publicado en línea: 15 Jul 2022
Volumen & Edición: AHEAD OF PRINT
Páginas: -
Recibido: 08 Feb 2022
Aceptado: 31 Mar 2022
Detalles de la revista
License
Formato
Revista
eISSN
2444-8656
Primera edición
01 Jan 2016
Calendario de la edición
2 veces al año
Idiomas
Inglés
Abstract

This article takes cross-border e-commerce platform operation receipts as the research object. It uses fractional differential equations to construct a market demand function jointly affected by channel radiation range, price, and price differences. The study results show that the heterodyne effect of cross-border e-commerce guarantees the competitive position of the market incumbents of the cross-border e-commerce platform. The article also illustrates the effectiveness and practicality of the cross-border e-commerce supply and demand fractional differential matching decision-making method.

Keywords

MSC 2010

Introduction

The product's short shelf life and its high risk of spoilage are the typical characteristics of fresh products that are different from normal temperature products. This also makes the cold chain dual-channel optimization unable to follow the previous channel optimization strategies directly. On the one hand, consumers have higher requirements for the quality and safety of the cold chain as food. The differences between online and physical sales in terms of product quality visualization and return and exchange services can easily affect the distribution of channel demand [1]. On the other hand, the excessive overlap of the target markets of different channels, the cross-effects of the quality of cold chain products, and channel prices have made channel conflicts and channel competition increasingly severe. The difficulty of dual-channel management optimization has increased.

The article analyzes the three cold chain and dual-channel models of non-converged dual-channel, converged dual-channel decentralized decision-making, and converged dual-channel integrated decision-making. We use profit maximization as the basis for decision-making to jointly optimize channel pricing and preservation investment levels. Based on this example, we explore the scope of application of the three modes [2]. The article expects to ensure channel food safety, ease channel conflicts, and promote consumption upgrade based on enriching cold chain dual-channel integration theoretical research. This assists in realizing healthy competition in the consumer market and the sustainable development of the multi-channel cold chain under the background of the mobile Internet.

Model description

We use superscript NI, IT to denote non-integrated dual-channel, integrated dual-channel decentralized decision-making, and integrated dual-channel integrated decision-making. Use the subscript r,m to denote the relevant parameters of the retailer and the manufacturer, respectively. Other symbols and their meanings are shown in Table 1.

Model symbols and meanings.

Symbol meaning Symbol meaning
A Potential market demand C2 Additional service costs for retailers serving online consumers
X The total area of system area Ce Additional unit costs are borne by network channels when non-uniform pricing
Y Physical channel radiation area D1 Market demand generated outside the Y range (network demand)
β The proportion of consumers shopping online within the physical radiation range 0<β<1 D2 Network demand generated in the Y range
α Demand impact factor of physical channel fresh-keeping input, a>0 D3 Entity demand generated in Y range
b1 Channel price elasticity, b1≥b2 π profit
b2 The demand influencing factor of the price difference between channels P1 Online channel selling price (decision variable)
k Fresh-keeping input cost coefficient P2 Physical channel selling price (decision variable)
θ Deterioration rate of cold chain products s Retailer's fresh-keeping investment level (decision variable), s ∈ [1,100]
w Physical channel commodity wholesale price P Commodity selling price when the channel price is consistent (decision variable)
c1 Manufacturer’s basic unit cost (including production and inventory costs, etc.) μ Unit compensation was given to retailers by the manufacturer (decision variable)
Establishment and solution of dual-channel profit function model
Non-converged dual-channel optimization

Non-converged dual channels implement non-uniform pricing, and consumers within the physical radiation range Y have the right to choose channels [3]. When the online price is higher than the physical price, some consumers may flow from physical channels to online channels, while consumers are more inclined to purchase physical. Consumers outside the scope of Y do not have the right to choose channels, so the market demand outside of Y can only be satisfied by D1N=AX(XA)b1.P1 D_1^N = {A \over X}\left({X - A} \right) - {b_1}.{P_1} through online channels. The network channel and physical channel requirements within the scope of Y are: D2N=AXYβb1,P1b2(P2P1),D3N=AXY(1β)b1P2P2(P1P2)+αsN. \matrix{{D_2^N = {A \over X}Y \cdot \beta - {b_1},\,{P_1} - {b_2} \cdot \left({{P_2} - {P_1}} \right),} \hfill \cr {D_3^N = {A \over X}Y \cdot \left({1 - \beta} \right) - {b_1} \cdot \,{P_2} - {P_2} \cdot \left({{P_1} - {P_2}} \right) + \alpha {s^N}} \hfill \cr}.

From the hypothesis, it can be seen that there is no crossover of inventory and service between the non-converged dual channels, and the retailer's profit comes from physical sales [4]. The profits of manufacturers are mainly composed of online sales profits and entity wholesale profits. The profits of each entity in the cold chain are as follows: πrN=p2D3ωD3(1+θ)12k(sN)2 \pi _r^N = {p_2} \cdot {D_3} - \omega \cdot {D_3} \cdot \left({1 + \theta} \right) - {1 \over 2}k \cdot {\left({{s^N}} \right)^2} πmN=[ωc1(1+θ)]D3(1+θ)+[P1(c1+ce)(1+θ)](D1+D2) \pi _m^N = \left[{\omega - {c_1} \cdot \left({1 + \theta} \right)} \right] \cdot {D_3} \cdot \left({1 + \theta} \right) + \left[{{P_1} - \left({{c_1} + {c_e}} \right) \cdot \left({1 + \theta} \right)} \right] \cdot \left({{D_1} + {D_2}} \right)

The cold chain dual-channel optimization is an Stackelberg game dominated by manufacturers. We use the reverse induction method to find the first derivative of P2 with respect to πrN \pi _r^N , and we can get: πrNP2=AXY(1β)b1P2b2(P1P2)+asN+[P2ω(1+θ)](b2b1) {{\partial \pi _r^N} \over {\partial {P_2}}} = {A \over X}Y \cdot \left({1 - \beta} \right) - {b_1} \cdot {P_2} - {b_2}\left({{P_1} - {P_2}} \right) + a{s^N} + \left[{{P_2} - \omega \left({1 + \theta} \right)} \right]\left({{b_2} - {b_1}} \right)

2πrNP22=2(b2b1)<0 {\because{{\partial ^2}\pi _r^N} \over {\partial P_2^2}} = 2({b_2} - {b_1}) < 0 . Therefore, in the case of the non-fusion type, the concave function of P2 has a maximum value. Therefore, the reaction function of the retailer's physical channel selling price P2 available to πrNP2=0 {{{\partial \pi _r^N} \over {\partial {P_2}}}} = 0 is: P2*=P1b1asNω(1+θ)(b1b2)+AXY(β1)2(b2b1) P_2^* = {{{P_1} \cdot {b_1} - a{s^N} - \omega \left({1 + \theta} \right)\left({{b_1} - {b_2}} \right) + {A \over X} \cdot Y \cdot \left({\beta - 1} \right)} \over {2\left({{b_2} - {b_1}} \right)}}

We can get πrNsN=a2(b2b1)D3+[P2ω(1+θ)]aksN {{\partial \pi _r^N} \over {\partial {s^N}}} = - {a \over {2\left({{b_2} - {b_1}} \right)}} \cdot {D_3} + \left[{{P_2} - \omega \cdot \left({1 + \theta} \right)} \right] \cdot a - {ks}^N by substituting formula (4) into formula (1). Since 2πrNsN2=a22(b1b2)k {{{\partial ^2}\pi _r^N} \over {\partial {s^{N2}}}} = {{{a^2}} \over {2\left({{b_1} - {b_2}} \right)}} - k is and only when k>a22(b1b2) k > {{{a^2}} \over {2\left({{b_1} - {b_2}} \right)}} is πrN \pi _r^N is a concave function with respect to sN. Suppose πrNsN=0 {{\partial \pi _r^N} \over {\partial {s^N}}} = 0 can obtain the response function of retailer's fresh-keeping investment sN level as: sN*=a{AY(β1)+P1b2X+Xω[b1b2(1θ)]}X[a22k(b1b2)] {s^{N*}} = {{a\left\{{A \cdot Y \cdot \left({\beta - 1} \right) + {P_1} \cdot {b_2} \cdot X + X \cdot \omega \left[{{b_1} - {b_2} \cdot \left({1 - \theta} \right)} \right]} \right\}} \over {X\left[{{a^2} - 2k\left({{b_1} - {b_2}} \right)} \right]}}

We substitute equations (4) and (5) into equation (2) and find the first derivative of πmN \pi _m^N to obtain: πmNP1=b12+a2b22[a22k(b1b2)]+D1+D2+[P1(c1+ce)(1+θ)]{2b1+b2[1b12(b2b1)+a2b22(b2b1)[a22k(b1b2)]]}2πmNP12=4b1+2b1[1b12(b2b1)+a2b22(b2b1)[a22k(b1b2)]] \matrix{{{{\partial \pi _m^N} \over {\partial {P_1}}} = - {{{b_1}} \over 2} + {{{a^2} \cdot {b_2}} \over {2\left[{{a^2} - 2k\left({{b_1} - {b_2}} \right)} \right]}} + {D_1} + {D_2} + \left[{{P_1} - \left({{c_1} + {c_e}} \right) \cdot \left({1 + \theta} \right)} \right] \cdot \left\{{- 2{b_1} + {b_2} \cdot} \right.} \hfill \cr {\left. {\left[{1 - {{{b_1}} \over {2\left({{b_2} - {b_1}} \right)}} + {{{a^2} \cdot {b_2}} \over {2\left({{b_2} - {b_1}} \right)\left[{{a^2} - 2k\left({{b_1} - {b_2}} \right)} \right]}}} \right]} \right\}} \hfill \cr {{{{\partial ^2}\pi _m^N} \over {\partial P_1^2}} = - 4{b_1} + 2{b_1} \cdot \left[{1 - {{{b_1}} \over {2\left({{b_2} - {b_1}} \right)}} + {{{a^2} \cdot {b_2}} \over {2\left({{b_2} - {b_1}} \right)\left[{{a^2} - 2k\left({{b_1} - {b_2}} \right)} \right]}}} \right]} \hfill \cr}

If and only if k>a2b222(b2b1)[4b12+2b227b1b2) k > {{{a^2} \cdot b_2^2} \over {2\left({{b_2} - {b_1}} \right)\left[{\left. {4b_1^2 + 2b_2^2 - 7{b_1} \cdot {b_2}} \right)} \right.}} is 2πmNP12<0 {{{\partial ^2}\pi _m^N} \over {\partial P_1^2}} < 0 , that is, πmN \pi _m^N is a concave function with respect to P1. We assume that πmNP1=0 {{\partial \pi _m^N} \over {\partial {P_1}}} = 0 can obtain the best network price decision of the manufacturer under the non-converged dual channel as: P1*=M1X+(c1+c2)(1+θ)M2+b2k(b1b2)(1+θ)[c1(1+θ)ω]2M2 P_1^* = {{{{{M_1}} \over X} + \left({{c_1} + {c_2}} \right) \cdot \left({1 + \theta} \right) - {M_2} + {b_2} \cdot k \cdot \left({{b_1} - {b_2}} \right)\left({1 + \theta} \right)\left[{{c_1} \cdot \left({1 + \theta} \right) - \omega} \right]} \over {2{M_2}}}

In,

M1 = Aa2 · (Y · (1−β) − X ] + Xωb2 · (a2kb2 · (a2kb2) + 2 AXk · (b1b2) + AYk · (2b1b2)(β−1) − Xb2 ·(b1b2θ+ b1θ)

M2=a2(b2b1)+k(4b126b1b2) {M_2} = {a^2} \cdot \left({{b_2} - {b_1}} \right) + k\left({4b_1^2 - 6{b_1} \cdot {b_2}} \right)

We bring formula (7) into formula (5) and formula (4) to obtain the retailer's best physical pricing P2* P_2^* and fresh-keeping investment level sN*. We put them into equations (1) and (2) respectively to finally obtain the best profit function πrN* \pi _r^{N*} , πmN* \pi _m^{N*} for cold chain producers and retailers under the non-integrated dual channel.

Integrated dual-channel decentralized decision-making

Converged dual channels implement unified pricing, so the price difference between channels is zero. Therefore, online channels do not need to bear additional costs. The three types of consumer demand under P1P2 = 0, Ce = 0 time-integrated dual-channel decentralized decision-making are expressed as follows: D1I=AX(XY)b1PI;D2IAXYβb1PI;D3I=AXY(1β)b1PI+asI D_1^I = {A \over X}\left({X - Y} \right) - {b_1} \cdot {P^I};D_2^I - {A \over X} \cdot Y \cdot \beta - {b_1} \cdot {P^I};\,D_3^I = {A \over X} \cdot Y \cdot \left({1 - \beta} \right) - {b_1} \cdot {P^I} + a{s^I}

In the converged dual channel, the network demand in the range of Y is diverted to offline retailers who need to pay a certain additional cost when serving such consumers. Manufacturers will give retailers a certain amount of compensation based on the amount of drainage. Therefore, the producer-dominated Stackelberg game is still the case [5]. The manufacturer first decides the compensation strategy, and then the retailer decides on the price of the product and the investment level of preservation according to the manufacturer's decision. The profit functions of the retailer and the manufacturer are: πrl=[PIω(1+θ)]D3I+(μc2)D2I(1+θ)12k(sI)2 \pi _r^l = \left[{{P^I} - \omega \cdot \left({1 + \theta} \right)} \right] \cdot D_3^I + \left({\mu - {c_2}} \right) \cdot D_2^I \cdot \left({1 + \theta} \right) - {1 \over 2}k \cdot {\left({{s^I}} \right)^2} πml=[ωc1(1+θ)]D3I(1+θ)+[PIc1(1+θ)](D1I+D2I)μ(1+θ)D2I \pi _m^l = \left[{\omega - {c_1} \cdot \left({1 + \theta} \right)} \right] \cdot D_3^I \cdot \left({1 + \theta} \right) + \left[{{P^I} - {c_1} \cdot \left({1 + \theta} \right)} \right] \cdot \left({D_1^I + D_2^I} \right) - \mu \cdot \left({1 + \theta} \right) \cdot D_2^I

We use the reverse induction method to find that πrI \pi _r^I is a concave function concerning price PI has a maximum value [6]. We assume that the reaction function of the integrated dual-channel decentralized decision-making cold chain product uniform pricing of πrIPI=0 {{\partial \pi _r^I} \over {\partial {P^I}}} = 0 is: PI*=AXY(1β)+asI+b1(1+θ)(ωμ+c2)2b1 {P^{I*}} = {{{A \over X}Y \cdot \left({1 - \beta} \right) + a{s^I} + {b_1}\left({1 + \theta} \right)\left({\omega - \mu + {c_2}} \right)} \over {2{b_1}}}

We can take Eq. (11) into Eq. (9) and find the derivative of πrI \pi _r^I for sI, πrIsI=a2b1[AXY(1β)b1PI+asI]+a2[PIω(1+θ)]ksI,2πrIsI2=a22b1k {{\partial \pi _r^I} \over {\partial {s^I}}} = {a \over {2{b_1}}} \cdot \left[{{A \over X} \cdot Y \cdot \left({1 - \beta} \right) - {b_1} \cdot {P^I} + a{s^I}} \right] + {a \over 2}\left[{{P^I} - \omega \left({1 + \theta} \right)} \right] - k{s^I},\,{{{\partial ^2}\pi _r^I} \over {\partial {s^{I2}}}} = {{{a^2}} \over {2{b_1}}} - k

If and only if k>a22b1 k > {{{a^2}} \over {2{b_1}}} , πrI \pi _r^I is a concave function concerning the fresh-keeping input level sI. We assume that the response function of the retailer's fresh-keeping investment level when πrIsI=0 {{\partial \pi _r^I} \over {\partial {s^I}}} = 0 is obtained when the integrated dual-channel decentralized decision-making is: sI*=a[AY(1β)Xb1(1+θ)(ω+μc2)]X(2b1ka2) {s^{I*}} = {{a\left[{AY \cdot \left({1 - \beta} \right) - X{b_1} \cdot \left({1 + \theta} \right)\left({\omega + \mu - {c_2}} \right)} \right]} \over {X\left({2{b_1}k - {a^2}} \right)}}

We bring Eqs. (13) and (12) into Eq. (11) and find the derivative of πmI \pi _m^I for μ. At the same time, we assume that πmIμ=0 {{\partial \pi _m^I} \over {\partial \mu}} = 0 can obtain the best compensation strategy for the manufacturer under the integrated dual-channel decentralized decision-making: μ*=AYβG1+Ab1kG2+2Xb13k2G3+Xa4b1(c12ω+c1θ2)+Xa2b12kG42Xb12k(a24b1k)(1+θ) {\mu ^*} = {{AY\beta \cdot {G_1} + A{b_1}k \cdot {G_2} + 2Xb_1^3{k^2} \cdot {G_3} + X{a^4}{b_1}\left({{c_1} - 2\omega + {c_1}{\theta ^2}} \right) + X{a^2}b_1^2k \cdot {G_4}} \over {2Xb_1^2k \cdot \left({{a^2} - 4{b_1}k} \right) \cdot \left({1 + \theta} \right)}}

In,

G1 = a4 + 6b1k(2b1ka2)

G2 = 2b1k(X − 4Y) − a2 (X − 2Y).

G3 = 3(c1c2ω) + θ(4c1 − 3c2 − 4ω + c1θ),

G4 = 10ω + c2 − 5c1 + θ(10ω + c2 − 8c1 − 3c1θ)

We bring equation (14) into equation (13) and equation (12) successively to obtain the retailer's best pricing and preservation investment decision under the integrated dual-channel decentralized decision-making. We bring the above parameters into equations (8) and (9), respectively, to obtain the optimal profit πrI* \pi _r^{I*} , πmI* \pi _m^{I*} of retailers and manufacturers in the integrated dual-channel decentralized decision-making.

Integrated dual-channel integrated decision-making

The integrated dual-channel integrated decision-making aims to maximize the overall profit of the cold chain system [7]. During integration, the decision-making variables are determined by three factors: channel pricing, fresh-keeping investment level, and producer compensation strategy. The three market requirements for integrated decision-making are: D1IN=D1I(PIN) D_1^{IN} = D_1^I\left({{P^{IN}}} \right) , D2IN=D2I(PIN) D_2^{IN} = D_2^I\left({{P^{IN}}} \right) , D3IN=D3I(PIN) D_3^{IN} = D_3^I\left({{P^{IN}}} \right) . Therefore, the total profit of the integrated dual-channel cold chain is: πIT=πrI+πmI=PIT(D1IN+D2IN+D3IN)c1(1+θ).[D1IN+D2IN+D3IN(1+θ)]c2D2IN(1+θ)12k(sIT)2 \matrix{{{\pi ^{IT}} = \pi _r^I + \pi _m^I = {P^{IT}}\left({D_1^{IN} + D_2^{IN} + D_3^{IN}} \right) - {c_1}\left({1 + \theta} \right).} \hfill \cr {\left[{D_1^{IN} + D_2^{IN} + D_3^{IN}\left({1 + \theta} \right)} \right] - {c_2}D_2^{IN}\left({1 + \theta} \right) - {1 \over 2}k{{\left({{s^{IT}}} \right)}^2}} \hfill \cr}

We find the first and second derivatives of PIT,sIT for πIT, and we can see that 2πITPIT2<0 {{{\partial ^2}{\pi ^{IT}}} \over {\partial {P^{IT2}}}} < 0 , 2πITsIT2<0 {{{\partial ^2}{\pi ^{IT}}} \over {\partial {s^{IT2}}}} < 0 , πIT has a maximum value. The optimal channel pricing and preservation investment levels in the integrated dual-channel integrated decision-making process obtained by solving the simultaneous equations πITPIT=0 {{{\pi ^{IT}}} \over {\partial {P^{IT}}}} = 0 , πITsIT=0 {{{\pi ^{IT}}} \over {\partial {s^{IT}}}} = 0 are: PIT*=A+asIT+b1c2(1+θ)+b1c1(θ2+4θ+3)6b1 {P^{IT*}} = {{A + a{s^{IT}} + {b_1}{c_2}\left({1 + \theta} \right) + {b_1} \cdot {c_1}\left({{\theta ^2} + 4\theta + 3} \right)} \over {6{b_1}}} sIT*=a[Ab1c1(5θ2+8θ+3)+b1c2(1+θ)a2+6b1k {s^{IT*}} = {{a\left[{A - {b_1} \cdot {c_1}\left({5{\theta ^2} + 8\theta + 3} \right) + {b_1} \cdot {c_2}\left({1 + \theta} \right)} \right.} \over {- {a^2} + 6{b_1}k}}

We bring equation (16) and equation (17) into equation (15) to obtain the optimal profit πIT* of the cold chain system under the integrated dual-channel integrated decision-making.

Example analysis
Calculation example data and optimization results

A certain large cherry planting/production and processing company was initially limited to online channels. Later, it expanded to enter the offline market at the same time. Based on the investigation of various data that occurred in big cherry circulation, the relevant parameters are simulated as follows [8]. The radiation range of the physical store within 25km covered by the system is roughly 18km. The potential market size is 200. The deterioration rate of fresh agricultural products is 0.05. Within the scope of physical radiation, 50% of consumers choose online channels to purchase fresh agricultural products. Other parameter settings are a = 0.8, b1 =1.5, b2 = 0.5, k =1.2, ω=15, c1 = 8, c2 = ce =1 respectively. We bring the parameters into the model and use Matlab software to calculate retailers’ and manufacturers’ best channel strategies and profits under non-converged dual-channel, converged dual-channel decentralized decision-making, and centralized decision-making (Table 2 and Figure 1).

Optimization results of different dual-channel types.

Non-converged dual-channel Integrated dual-channel decentralized decision-making Integrated dual-channel integrated decision-making
P1 25.89 32.35 27.79
P2 45.28
s 19.49 11.07 12.65
μ / 5.72 /
πr 639.39 579.39 /
πm 872.69 824.39 /
Total cold chain profit 1512.08 1403.78 150.4.60

Figure 1

Comparison of market demand under different dual-channel types

By comparing the optimization results of non-integrated dual-channel, integrated dual-channel decentralized decision-making, and integrated dual-channel integrated decision-making in Table 2, the following results can be obtained:

1) The non-converged dual-channel system has the largest total profit. The second is integrated decision-making. The system of integrated decentralized decision-making has the lowest profit. 2) There is a significant difference in commodity prices between physical and online channels in a non-converged dual-channel supply chain. 3) The sales of converged integrated decision-making products are higher than the prices of converged decentralized decision-making products. P1 < PI < PIN < P2. 4) The fresh-keeping investment level of fusion integrated decision-making is slightly higher than that of fusion decentralized decision-making sI < sIN s.

Combining the electric field demand distribution in Figure 1, it can be seen that the physical channel demand of fusion decentralized decision-making is slightly higher than that of a non-converged entity when the cost/cost is the same. Still, the physical price of non-converged unit commodities is much higher than that of converged decentralized decision-making. This caused the profit of the non-converged system to be higher than that of the converged system [9]. In addition, although the profit difference between the two is relatively small, the total demand for an integrated electric field for integrated decision-making is 85. This is higher than the total demand of the non-converged market 70 and is also significantly higher than the total demand of the convergent decentralized decision-making market 63. Therefore, the analysis results in Table 2 and Figure 1 are non-integrated and integrated decision-making through competitive prices. Non-converged dual channels use the high fresh-keeping investment to satisfy the physical consumption experience. The significant price difference between physical channels and online channels stimulates online consumption to achieve high profitability. The performance of integrated decentralized decision-making in terms of price, preservation level, and cold chain profit indirectly proves that the integrated development of dual channels should be based on highly integrated decision-making.

Sensitivity analysis

There are many new varieties involved in actual production and life. There are obvious differences in the decay holdings, even for consumer groups of the same type of goods in different geographical areas [10]. The model selects three parameters: commodity deterioration rate θ, physical channel radiation range and physical radiation range Y, consumer online shopping ratio β for sensitivity analysis.

The impact of fluctuations in commodity deterioration rate θ

We select a floating value of 0.05 units to analyze the impact of the difference in new product types on dual-channel optimization. We compare the differences in the optimization results of the three dual-channel decision-making modes at θ ∈ [0.05, 0.25] the time (Figure 2).

Figure 2

The impact of deterioration rate on channel prices, total profits, and fresh-keeping input levels

It is found from Figure 2 that the increase in the spoilage rate of the three types of dual-channel decision-making commodities will cause an increase in channel prices and a decrease in the total profit of the system [11]. The increase in spoilage rate means an increase in the number of goods spoiled in the same environment. To ensure the market supply, retailers and manufacturers have to order/produce more goods to cope with the high spoilage rate. This increases the cost of corruption. To maintain the profit of the system, the cold chain dual channels will appropriately increase the unit price and reduce the level of investment in preservation.

The influence of fluctuations in radiation range Y of physical channels

When the area of the system is determined, the change in the radiation range of the physical channel will directly affect the fluctuation of the market demand of each channel. This may trigger changes in the dual-channel optimization strategy [12]. Therefore, analyze the optimization results of the three decision-making modes of dual channels when Y is internally changed (Figure 3).

Figure 3

Impact on channel price and total profit

With the increase in the radiation range of physical channels, the differences between the three optimization channels are significant. In the case of non-convergence, the online price decreases with the increase of Y, while the physical price increases. The price difference between channels gradually increases. Retailers’ fresh-keeping investment also increases significantly with Y, and the channel price of integrated decentralized decision-making increases with Y. The increase has been increasing slowly [13]. Its investment in preservation has also increased slightly, while the channel price and investment level in preserving the integrated decision-making remain unchanged. When the radiation range of the physical channel does not exceed 50% of the total area of the area where the system is located, the profitability of the integrated dual-channel integrated decision-making system is relatively strong. When the radiation range of the physical channel exceeds 50%, the non-converged dual-channel decision-making is more conducive to the increase of system profits. The increase of Y in Figure 3 makes the difference between P1 and A gradually increase, which leads to retailers’ profits gradually being higher than those of manufacturers. This also verifies from the side that even some consumers have a preference for online channels. The expansion of physical sales radiation will also reduce the “barrier-free” advantage of online sales in geographic space.

Conclusion

The model has not well reflected the difference in the storage cost of fresh products per unit and the superiority of the integrated dual-channel in intensive inventory management. Therefore, considering inventory cost differences, cold chain dual-channel optimization will become one of the future research directions.

Figure 1

Comparison of market demand under different dual-channel types
Comparison of market demand under different dual-channel types

Figure 2

The impact of deterioration rate on channel prices, total profits, and fresh-keeping input levels
The impact of deterioration rate on channel prices, total profits, and fresh-keeping input levels

Figure 3

Impact on channel price and total profit
Impact on channel price and total profit

Model symbols and meanings.

Symbol meaning Symbol meaning
A Potential market demand C2 Additional service costs for retailers serving online consumers
X The total area of system area Ce Additional unit costs are borne by network channels when non-uniform pricing
Y Physical channel radiation area D1 Market demand generated outside the Y range (network demand)
β The proportion of consumers shopping online within the physical radiation range 0<β<1 D2 Network demand generated in the Y range
α Demand impact factor of physical channel fresh-keeping input, a>0 D3 Entity demand generated in Y range
b1 Channel price elasticity, b1≥b2 π profit
b2 The demand influencing factor of the price difference between channels P1 Online channel selling price (decision variable)
k Fresh-keeping input cost coefficient P2 Physical channel selling price (decision variable)
θ Deterioration rate of cold chain products s Retailer's fresh-keeping investment level (decision variable), s ∈ [1,100]
w Physical channel commodity wholesale price P Commodity selling price when the channel price is consistent (decision variable)
c1 Manufacturer’s basic unit cost (including production and inventory costs, etc.) μ Unit compensation was given to retailers by the manufacturer (decision variable)

Optimization results of different dual-channel types.

Non-converged dual-channel Integrated dual-channel decentralized decision-making Integrated dual-channel integrated decision-making
P1 25.89 32.35 27.79
P2 45.28
s 19.49 11.07 12.65
μ / 5.72 /
πr 639.39 579.39 /
πm 872.69 824.39 /
Total cold chain profit 1512.08 1403.78 150.4.60

Yang, Y. Research on the optimization of the supplier intelligent management system for cross-border e-commerce platforms based on machine learning. Information Systems and e-Business Management., 2020; 18(4): 851–870 YangY. Research on the optimization of the supplier intelligent management system for cross-border e-commerce platforms based on machine learning Information Systems and e-Business Management. 2020 18 4 851 870 10.1007/s10257-019-00402-1 Search in Google Scholar

Luo, Y., Ma, J., & Li, C. Entity name recognition of cross-border e-commerce commodity titles based on TWs-LSTM. Electronic Commerce Research., 2020; 20(2): 405–426 LuoY. MaJ. LiC. Entity name recognition of cross-border e-commerce commodity titles based on TWs-LSTM Electronic Commerce Research. 2020 20 2 405 426 10.1007/s10660-019-09371-6 Search in Google Scholar

Miao, Y., Du, R., Li, J., & Westland, J. C. A two-sided matching model in the context of B2B export cross-border e-commerce. Electronic Commerce Research., 2019; 19(4): 841–861 MiaoY. DuR. LiJ. WestlandJ. C. A two-sided matching model in the context of B2B export cross-border e-commerce Electronic Commerce Research. 2019 19 4 841 861 10.1007/s10660-019-09361-8 Search in Google Scholar

Li, G., & Li, N. Customs classification for cross-border e-commerce based on text-image adaptive convolutional neural network. Electronic Commerce Research., 2019; 19(4): 779–800 LiG. LiN. Customs classification for cross-border e-commerce based on text-image adaptive convolutional neural network Electronic Commerce Research. 2019 19 4 779 800 10.1007/s10660-019-09334-x Search in Google Scholar

Gençoğlu, M. & Agarwal, P. Use of Quantum Differential Equations in Sonic Processes. Applied Mathematics and Nonlinear Sciences., 2021; 6(1):21–28 GençoğluM. AgarwalP. Use of Quantum Differential Equations in Sonic Processes Applied Mathematics and Nonlinear Sciences. 2021 6 1 21 28 10.2478/amns.2020.2.00003 Search in Google Scholar

Zhou, P., Fan, Q. & Zhu, J. Empirical Analysis on Environmental Regulation Performance Measurement in Manufacturing Industry: A Case Study of Chongqing, China. Applied Mathematics and Nonlinear Sciences., 2020; 5(1): 25–34 ZhouP. FanQ. ZhuJ. Empirical Analysis on Environmental Regulation Performance Measurement in Manufacturing Industry: A Case Study of Chongqing, China Applied Mathematics and Nonlinear Sciences. 2020 5 1 25 34 10.2478/amns.2020.1.00003 Search in Google Scholar

Cheng, X., Su, L., & Zarifis, A. Designing a talents training model for cross-border e-commerce: a mixed approach of problem-based learning with social media. Electronic Commerce Research., 2019; 19(4): 801–822 ChengX. SuL. ZarifisA. Designing a talents training model for cross-border e-commerce: a mixed approach of problem-based learning with social media Electronic Commerce Research. 2019 19 4 801 822 10.1007/s10660-019-09341-y Search in Google Scholar

Wu, Y., Wang, Q., Zhou, Y., & Huang, M. Strategic decisions on the establishment of an offshore warehouse for manufacturers with an e-commerce channel. International Journal of Industrial and Systems Engineering., 2021; 38(2): 239–262 WuY. WangQ. ZhouY. HuangM. Strategic decisions on the establishment of an offshore warehouse for manufacturers with an e-commerce channel International Journal of Industrial and Systems Engineering. 2021 38 2 239 262 10.1504/IJISE.2021.115323 Search in Google Scholar

Song, Y. Application of Big Data Analysis Technology in Cross-Border E-Commerce. Journal of Electronic Research and Application., 2021; 5(4): 1–3 SongY. Application of Big Data Analysis Technology in Cross-Border E-Commerce Journal of Electronic Research and Application. 2021 5 4 1 3 10.26689/jera.v5i4.2495 Search in Google Scholar

Tan, C. S. Understand China’s cross-border e-commerce industry: a market entry mode. Systematic Literature Review and Meta-Analysis Journal., 2021; 2(2): 74–85 TanC. S. Understand China’s cross-border e-commerce industry: a market entry mode Systematic Literature Review and Meta-Analysis Journal. 2021 2 2 74 85 10.54480/slrm.v2i2.23 Search in Google Scholar

Zhao, W. X., & Fang, L. Collaborative Strategy Optimization of Cross-Border e-Commerce Coordination Along the Belt and Road Initiative. Journal of Frontiers in Engineering Technology., 2021; 1(1): 71–78 ZhaoW. X. FangL. Collaborative Strategy Optimization of Cross-Border e-Commerce Coordination Along the Belt and Road Initiative Journal of Frontiers in Engineering Technology. 2021 1 1 71 78 10.23977/infse.2021.020206 Search in Google Scholar

Niu, B., Wang, J., Lee, C. K., & Chen, L. “Product+ logistics” bundling sale and co-delivery in cross-border e-commerce. Electronic Commerce Research., 2019; 19(4): 915–941 NiuB. WangJ. LeeC. K. ChenL. “Product+ logistics” bundling sale and co-delivery in cross-border e-commerce Electronic Commerce Research. 2019 19 4 915 941 10.1007/s10660-019-09379-y Search in Google Scholar

Zhu, Q., Liu, B., Han, F., & Lee, M. The optimization effect of fuzzy factional-order ordinary differential equation in block chain financial cross-border E-commerce payment mode. Alexandria Engineering Journal., 2020; 59(4): 2839–2847 ZhuQ. LiuB. HanF. LeeM. The optimization effect of fuzzy factional-order ordinary differential equation in block chain financial cross-border E-commerce payment mode Alexandria Engineering Journal. 2020 59 4 2839 2847 10.1016/j.aej.2020.06.031 Search in Google Scholar

Artículos recomendados de Trend MD

Planifique su conferencia remota con Sciendo