Decision Model of Double Channel Recycling Closed Loop Supply
Chain Considering Service Level
Hongsheng Sun, Hao Zhang and Baoyu Li
*
School of Traffic and Transportation, Beijing Jiaotong University, Beijing, China
Keywords: Dual-Channel Recycling, Closed-Loop Supply Chain, Game Theory.
Abstract: With the development of economy and the iteration and upgrading of science and technology, the upgrading
rate of electronic products has been gradually accelerated. Improving the recycling volume of waste electronic
products and forming a standardized recycling system are the key issues facing the current recycling link.
Under this background, based on Stackelberg game model, the author establishes a closed loop supply chain
model of dual channel recycling of waste electronic products, and takes profit maximization as the decision
objective to conduct a numerical analysis. The paper proves that under certain conditions, consumers' online
channel preference will promote the profits and recycling capacity of the supply chain system, and improve
the service level of recycling participants, which has a positive effect on the system profits.
1 INTRODUCTION
With the gradual improvement of the upgrading rate
of electronic products, the recycling of waste
electronic products has attracted more attention.
Electronic product recycling can effectively reduce
environmental pollution and production costs. At
present, domestic and foreign scholars mainly
consider pricing decisions and contract coordination
under such influencing factors as consumer behavior,
demand, channel dominance, and fairness concerns.
There is less research on service level decisions. At
the same time, there is also insufficient research on
reverse supply chains where different entities provide
different types of recovery services. Therefore, this
paper mainly constructs a dual channel recycling
model of waste electronic products considering
service level, analyzes the game relationship between
different players in the supply chain, and judges the
impact of service level on decision-making.
*
Corresponding author
2 PROBLEM DESCRIPTION AND
MODEL ASSUMPTION
2.1 Problem Description
The main body of this paper is a dual channel
recycling closed-loop supply chain, which includes
two links: forward logistics and reverse logistics. In
the positive process, manufacturers wholesale their
goods to retailers at wholesale prices, and retailers
sell their products to consumers at retail prices. In the
process of reverse logistics, third-party recyclers and
Internet recycling platforms respectively recycle
waste electronic products from consumers at
different offline and online recycling prices, and
manufacturers recycle them from both at recycling
transfer prices. The closed-loop supply chain
structure of dual channel recycling is shown in Figure
1.
730
Sun, H., Zhang, H. and Li, B.
Decision Model of Double Channel Recycling Closed Loop Supply Chain Considering Service Level.
DOI: 10.5220/0012042900003620
In Proceedings of the 4th International Conference on Economic Management and Model Engineering (ICEMME 2022), pages 730-735
ISBN: 978-989-758-636-1
Copyright
c
 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

manufacturer retailer consumer
Online recycling
platform
Forward logis tics Reverse logistics
Offline recyclers
Offline recycling channel
Online recycling channel
ω p
g
g
p
off
p
on
Figure 1. Double channel recycling closed-loop supply chain
In the supply chain system, the manufacturer and
retailer, offline recycler and online recycling
platform are Stackelberg game relations, the
manufacturer is the leader of the game, and the other
three are followers. The decision-making process of
the system is as follows:
(1) the manufacturer determines the wholesale
price
ω
and recycling price
g
.
(2) Retailers set the selling price
p
based on
ω
,
offline recyclers and online recycling platforms set
the offline recycling price
off
p
and online recycling
price
on
p
based on
p
and
g
respectively.
2.2 Model Assumption
In order to accurately describe the operation mode of
the supply chain system model, this paper proposes
the following assumptions:
(1) The decision-making objective of each
subject is to maximize economic benefits, without
considering the influence of irrational factors.
(2) In the game relationship, the decision is made
under the condition of complete information, the
manufacturer gives priority to the decision, then
others follow.
(3) The quality of remanufactured products is the
same as that of new products, the sales price of both
is the same, and the consumer acceptance is the same.
The production cost of remanufactured products is
r
c
, and the production cost of new products is
n
c
,
0
nr
ccΔ= >−
. (Giri, 2017)
(4) The damage degree of WEEE is consistent,
and the recovery conversion rate is 100%, that is,
Δ
remains unchanged. (Wu, 2019)
(5) Because the service level of traditional offline
recycling channels is very low, this paper does not
consider the service level of this channel, but only
considers the service level of online recycling
channels.
(6) Discuss a single cycle decision-making
model, assuming that the products of the previous
cycle can be recycled in the current research cycle.
(7) The recycling function of waste products is
assumed to be a linear function of online recycling
price, offline recycling price and service level. A
large number of literatures have used this function
(Wu, 2019; Wu, 2012; Huang, 2012). Manufacturers
and Internet recycling platforms need to pay
corresponding service costs to provide recycling
services. This paper assumes that the service cost is
positively correlated with the square of the service
level. (Wang, 2017; Guo, 2019)
3 MODEL BUILDING
3.1 Parameter Definition
Two dual-channel supply chain inventory models are
constructed based on the multi-agent method. The
parameters and relevant symbols of the model are
shown in Table 1.
Decision Model of Double Channel Recycling Closed Loop Supply Chain Considering Service Level
731

Table 1: Symbols and parameter definitions.
S
y
mbols Definition S
y
mbols Definition
ω
Manufacturer wholesale price
p
Retailer selling price
n
c
Unit new product manufacturing cost
r
c
Unit recovered manufacturing cost
γ
Price elasticity of product demand
q
Market demand in sales. Maximum
demand is Q
t
q
Recycling volume of offline recycling channels
e
q
Recycling volume of online recycling
channels
θ
Consumer channel preference
β
Recovery elastic coefficient of
com
p
etitive channel rec
y
clin
g
p
rice
α
Recovery elastic coefficient of operation
channel rec
y
clin
g
p
rice
e
p
Unit recycling price of online recycling
p
latfor
m
t
p
Unit recycling price of offline recyclers
g
Unit recycling price of manufacturer
1
s
Convenience service level of online recycling
p
latfor
m
2
s
Safety service level of manufacturer
1
s
c
Convenience service cost of online recycling
p
latfor
m
2
s
c
Safety service cost of manufacturer
1
μ
Service cost coefficient of online recycling
p
latfor
m
2
μ
Service cost coefficient of manufacturer
a
Market recycling volume
i
Recovery elastic coefficient of
operation channel recycling volume
j
Recovery elastic coefficient of competitive
channel recycling volume
Π
m
Profit of manufacturer
Π
e
Profit of onling recycling platform
Π
t
Profit of offline recyclers
Π
r
Profit of retailer
Π
Profit of Supply chain system
The model construction takes into account the
system differences between the provision of
recycling services and the provision of recycling
services. A superscript S indicates that the provision
of recycling services is a parameter and variable of
the system.
3.2 Model Construction
A dynamic game model is built for the situation that
manufacturers provide security services and online
recycling platforms provide convenience services, in
order to explore the impact of the existence of
recycling services and the improvement of service
levels on supply chain system decisions and profits.
The following quantity relations exist in the model:
()()()
ees2
2
111 2 22 22
Π
) ) ] ( )( ( [ ( ) ( )
S
mn t t
nte
cq q q gq q c
cQ p gap p i js i js s
ω
ω
γμ
=− +Δ+− +−
=− − +Δ− +++− +− −
(1)
() ()( )
2
es1 1122 11
Π2
S
ee e et
g
pq c g p a p p is is s θ
μ
=− −=− + −++ −
(2)
()( )()
11 2 2
Πg 1 2
S
tttt te
p
qgp appjsjs θ

=− =− − + −− −

(3)
()()( )
Π
r
S
p
qp Qp ωω
γ
=− =− −
(4)
Under the condition of complete information,
manufacturers can estimate the impact of their own
decisions on the decisions of other followers, and
make optimal decisions based on understanding the
follower's decision response function. Therefore,
reverse induction can be used to solve the model, that
is, the optimal profit decision of retailers, offline
recyclers and online recycling platforms can be
solved first, and then the manufacturer make a
decision.
Retailers need to make decisions on price
p
, and
take the partial derivative of
Π
S
r
to
p
. It can be seen
that there is
p
to maximize
Π
S
r
.
2
Q
p
γ
ω
γ
+
=
(5)
Offline recyclers need to make decisions on price
t
p
, and take the partial derivative of
Π
S
t
to
t
p
. It
can be seen that there is
t
p
to maximize
Π
S
t
.
()
11 2 2
1
2
4
te
pagapjsjs θ =−+ +++ +
(6)
Online recovery needs to make decisions on
e
p
and
1
s
, and take partial derivative of
Π
S
e
to
e
p and
1
s
respectively. Through the judgment of
Hessianmatrix of
e
p
and
1
s
, it can be seen that there
are
e
p
and
1
s
to maximize
Π
S
e
.
()
11 2 2
1
2
4
et
pgapisis θ =−+−−
(7)
ICEMME 2022 - The International Conference on Economic Management and Model Engineering
732

()
1
1
1
2
e
ig p
s
μ
−
=
(8)
Solve by simultaneous formula (5)~(8) and
substitute the result into formula (1).
() ()
()()
[]
()()
22
1 222 22
2
12 2 22
2
11 2 2 2 2
2
111 1
10 4 2 3 3 3 4 6
42
224
Π
82 60
n
n
n
S
m
Da g Q Ec Di j s s
isDj s Daga EEc
ij Ec s s Di D g a Q
iij
μωγω μ
μθω
μθωγω
μ
+− + + + − +
+−−+−+−
++ −−+−+
=
−−




(9)
2
111 1
430Aiij μ −−=
(10)
222
3(4 )Ba a i js θ =+ + −
(11)
222
410 3 ( 4)Ca ga i js θ =− − +− (12)
Dg=−Δ
(13)
EQ γω =−
(14)
After knowing the above information, the
manufacturer could make a decision on
g
,
ω
and
2
s
. Through the judgment of Hessianmatrix of
g
,
ω
and
2
s
, it can be seen that there are
g
,
ω
and
2
s
to
maximize
Π
S
m
.
*
2
S
n
Qc γ
ω
γ
+
=
(15)
()
2
11 11
*
2
10
s
KJiLijF
s
G
μ −−
=
−
(16)
()() ()
22
21 11 1
*
10 2
S
FA aHi aij a
g
G
μθμ

Δ− Δ+ +Δ− + −Δ

=
(17)
()
2
12 121 2 2 1
10Fij iij i j μ =− + −
(18)
2
11 1 1 2
)2[ ](20 GF Aii j μμ =− +−
(19)
2Ka=+Δ (20)
La θ =Δ+ (21)
4 EXAMPLE ANALYSIS
4.1 Parameter Setting
n this paper, Mathematica software is used for
simulation analysis. Because some data related to the
interests of enterprises are difficult to obtain, this
study chooses to design parameters based on existing
literature. Set the parameters as follows:
20000Q =
,
2 γ =
,
1500
n
c =
,
500
r
c =
, 500a = , 2 α = ,
1 β =
,
12
2ii==
,
12
1jj==
.
4.2 Influence of Consumer Channel
Preference on Decision-Making
This section assumes that the service cost coefficient
12
8 μμ ==
, the remaining parameters remain
unchanged, and the consumer channel preference is
gradually increased from 0 to 1 in steps of 0.1. The
model results are shown in Figure 2.
(
a
)
Decision variable curve
(
b
)
Profit curve
Figure 2: System simulation results on
θ .
As consumers prefer online channels, online
recycling platforms can continuously reduce
recycling prices, while offline recyclers need to
increase recycling prices to strengthen
competitiveness, and manufacturers can slowly
reduce recycling prices. Online recycling platforms
and manufacturers should improve their service
levels and retain more consumers. In addition, with
the increase of x, the profits of offline recyclers
continue to decrease, while the profits of
manufacturers and online recycling platforms
increase, and the total profits of the system show a
Decision Model of Double Channel Recycling Closed Loop Supply Chain Considering Service Level
733

trend of first decreasing and then increasing. This
shows that there are difficulties in the initial
development of online recycling platform
construction, and there are negative benefits for the
system as a whole. However, with the increase of
platform customer groups, the system benefits will
gradually increase.
4.3 Influence of Service Cost
Coefficient on Decision-Making
In this section, it is assumed that consumer channel
preference remains unchanged, 𝜃0.4 , other
parameters remain unchanged, 𝜇

and 𝜇

change
from 10 to 3, and the step size is -1. The model results
are shown in Figure 3.
(a)
(
b
)
(
c
)
(
d
)
(
e
)
(
f
)
Figure 3: System simulation results on 𝜇

and 𝜇

.
𝜇

has a particularly significant impact on the
profits of online recycling platforms and offline
service providers. Optimizing
2
μ
will improve the
profits of online recycling platforms, but reduce the
profits of offline service providers. The impact of
optimizing
1
μ
on both is the same as optimizing 𝜇

,
but the impact is smaller than 𝜇

. The influence of
optimizing 𝜇

on the profit of manufacturer and
supply chain system is significantly higher than that
of optimizing 𝜇

. To sum up, optimizing the
convenience service cost coefficient is more
beneficial to the improvement of the supply chain
profit, and the profit loss of the offline recyclers is
also the smallest in this state. Based on this, when
pursuing the profit stability of each entity and
ICEMME 2022 - The International Conference on Economic Management and Model Engineering
734

maximizing the system profit, priority can be given
to optimizing the convenience service level.
5 CONCLUSION
This paper establishes a closed-loop supply chain
game model consisting of manufacturers, retailers,
third-party recyclers and Internet recycling platform,
and considers the impact of service level on the
decision-making of supply chain participants.
Through the analysis of an example, it is verified that
the improvement of consumers' online recycling
channel preference has a more positive role in
promoting the pricing and profits of online recycling
platforms, but has a more negative role in weakening
offline recyclers. At the same time, it can also
promote manufacturers and online recycling
platforms to improve service levels, which has a
positive impact on the supply chain system. The
service cost also plays a positive role in the system
benefits, but in the pursuit of environmental
protection effects, priority can be given to optimizing
the security service level, while in the pursuit of
profit maximization, priority can be given to
optimizing the convenience service level.
This paper also has limitations. It believes that the
service level of online recycling services has become
worse, and its influencing factors have not been taken
into account in the model. However, with the renewal
of policies such as dual carbon, the service mode of
offline service providers will also change, and the
service level will be improved. Relevant factors can
be considered in future research.
REFERENCES
Giri B C, Chakraborty A, Maiti T. Pricing and Return
Product Collection Decisions in a Closed-Loop Supply
Chain with Dual-Channel in Both Forward and Reverse
Logistics [J]. Journal of Manufacturing Systems,
2017,42:104-123.
Guo Li, Lin Li, Jiasen Sun. Pricing and Service Effort
Strategy in a Dual-Channel Supply Chain with
Showrooming Effect [J]. Transportation Research Part
E, 2019,126.
Huang S, Yang C, Zhang X. Pricing and Production
Decisions in Dual-Channel Supply Chains with
Demand Disruptions[J]. Computers & Industrial
Engineering, 2012,62(1):70–83.
Wang L, Song H, Wang Y, et al. Pricing and Service
Decisions of Complementary Products in a Dual-
Channel Supply Chain[J]. Computers & Industrial
Engineering, 2017:223-233.
Wu D, Chen J, Yan R, et al. Pricing Strategies in Dual-
Channel Reverse Supply Chains Considering Fairness
Concern [J]. International Journal of Environmental
Research and Public Health, 2019, 16(9): 1657.
Wu C H. Price and Service Competition between New and
Remanufactured Products in A Two-echelon Supply
Chain [J]. International Journal of Production
Economics, 2012,140(1):496-507.
Decision Model of Double Channel Recycling Closed Loop Supply Chain Considering Service Level
735