3.3.1. Basic SD Model of Remanufacturing CLSC
The study constructs a comprehensive SD model of CLSC, comprising a remanufacturer (M), a retailer (R), and a recycler (T). The remanufacturer manufactures and reproduces new products while the retailer sells the finished products. The recycler recovers and delivers old products to the remanufacturer for remanufacturing. Qualified reproductions are reintegrated into the CLSC system.
While developing the basic SD model of CLSC, this study employs the Vensim DSS 6.0 simulation software. The model comprises two main components: the inventory and capital system and the evolutionary game system, which determines each member’s quality improvement decision. During modeling, certain parameters of the inventory and capital system model are established based on the research of Duan W. et al. [
46], while other parameters of the evolutionary game system model concerning quality improvement decisions are derived from the work of Duan W. et al. [
49]. The parameters related to quality control and contract coordination are defined in this study. These model parameters are set according to the real CLSC operation data, and the validity and reliability of these parameters are proven by model tests and practice tests.
The SD model of the inventory and capital system is shown in
Figure 1.
The SD model of the evolutionary game system for each member’s quality improvement decision is shown in
Figure 2.
The system flow diagram is refined according to the causal relationship, which can clearly reflect the relationship between state variables, rate variables, auxiliary variables, and constants so as to describe the system structure and simulate the system behavior. As shown in
Figure 1, the SD model of the inventory and capital system includes the inventory subsystem of the remanufacturer, the inventory subsystem of the retailer, the inventory subsystem of the recycler and the capital subsystem of the remanufacturer, the capital subsystem of the retailer, and the capital subsystem of the recycler. By connecting these six subsystems, a three-level CLSC system flow diagram reflecting material flow and capital flow can be formed.
The SD model of each member’s quality improvement decision evolutionary game system is composed of the remanufacturer evolutionary game subsystem, the retailer evolutionary game subsystem and the recycler evolutionary game subsystem.
The main parameters of the basic SD model are listed in
Table 4.
Since the basic SD model does not include quality control or revenue distribution, the corresponding parameters related to quality control and revenue sharing are 0.
Main equations of basic SD model:
M Probability of quality improvement = IF THEN ELSE((M Probability of quality improvement + M Rate of change in strategy selection) > 1, 1-M Probability of quality improvement, IF THEN ELSE((M Probability of quality improvement + M Rate of change in strategy selection) < 0, -M Probability of quality improvement, M Rate of change in strategy selection)); initial value = 0.4.
R Probability of quality improvement = IF THEN ELSE((R Probability of quality improvement + R Rate of change in strategy selection) > 1, 1-R Probability of quality improvement, IF THEN ELSE((R Probability of quality improvement + R Rate of change in strategy selection) < 0,-R Probability of quality improvement, R Rate of change in strategy selection)); initial value = 0.5.
T Probability of quality improvement = IF THEN ELSE((T Probability of quality improvement + T Rate of change in strategy selection) > 1, 1-T Probability of quality improvement, IF THEN ELSE((T Probability of quality improvement + T Rate of change in strategy selection) < 0,-T Probability of quality improvement, T Rate of change in strategy selection)); initial value = 0.4.
M Cost = MIN(M low quality production affects R high quality sales when the fine on M+M fixed charge+ M’s subsidy to R’s active promotion+ “M Additional inputs for high-quality production”+ Recovery of parts quality inspection costs+ Recycling parts procurement costs + Cost of new parts + M delivery ratio*MR Revenue sharing coordinates unit price difference, M Profit).
R Cost = MIN(MR revenue sharing+ R fixed cost+ R Potential loss of negative sales+ R positive marketing costs+ R order cost+ T recovery rate*RT Revenue sharing coordinates unit price difference, R Profit).
T Cost = MIN(T recycling cost+ T fixed cost+ “T provides a fine for M when low-quality recycled products affect M’s high-quality production”+ “T provides the additional cost of high-quality recycled products” + T revenue sharing, T Profit).
M Income = Wholesale unit price * (1 + M Government subsidy coefficient for remanufactured products manufacturing) * M delivery ratio+ “R actively promotes the gain of M enterprise image.”+ “T provides a fine for M when low-quality recycled products affect M’s high-quality production"+ M revenue sharing.
R Income = Sale unit price * (1 + R Subsidy coefficient of government subsidies for the sale of remanufactured products) * R Sales rate+ M’s subsidy to R’s active promotion+ M low quality production affects R high quality sales when the fine on M+ RT revenue sharing+ MR Revenue sharing coordinates unit price difference*M delivery ratio.
T Income = Recycling parts procurement costs * (1 + “The coefficient of government subsidy for high-quality recovery”) + ”M, R’s support for T additional input”+ T recovery rate * RT Revenue sharing coordinates unit price difference.
3.3.2. SD Model of the Quality Control Contract Scheme
The quality control contract scheme compared with the basic SD model adds the following variables: M’s subsidy to R’s active promotion; M’s low-quality production affects R’s high-quality sales when the fine is on M; M, R’s support for T’s additional input, T provides a fine for M when low-quality recycled products affect M’s high-quality production.
When R retailers actively fulfill the quality improvement commitment, M subsidizes R’s active marketing behavior. Meanwhile, if the M remanufacturer breaches the quality improvement contract and the product quality affects the R retailers’ high-quality sales, the M remanufacturer will be fined a certain amount. As for the T recycler, if T actively performs high-quality recycling, M and R will support the additional investment of T; if T does not actively perform the quality improvement contract to provide low-quality recycling, M remanufacturer will fine T.
Main equations of the SD model of quality control contract scheme:
M’s subsidy to R’s active promotions = IF THEN ELSE(M Probability of quality improvement > 0, (9500 * (R Probability of quality improvement + M Probability of quality improvement))/2, 0).
M low-quality production affects R high-quality sales when the fine on M = (1-M Probability of quality improvement) * 8000.
M, R’s support for T’s additional input = ((M Probability of quality improvement + R Probability of quality improvement)/2) * 10,000.
T provides a fine for M when low-quality recycled products affect M’s high-quality production = M Probability of quality improvement * (1-T Probability of quality improvement) * 1000.
The main parameters of the SD model of the quality control contract scheme are set as consistent with the basic SD model and are not presented here. In addition, this part of the model does not involve revenue sharing, and the relevant parameters are set to 0.
3.3.3. SD Model of the Quality Control-Revenue-Sharing Contract Scheme
The SD model of the quality control–revenue-sharing contract scheme adopts the combination contract form and adds the revenue-sharing contract on the basis of the quality control contract scheme to coordinate the benefits of each member and explore more scientific benefit distributions among the CLSC members after quality control.
Unlike the quality control contract scheme, this scheme increases the revenue sharing between upstream and downstream manufacturers. The M remanufacturer takes the initiative to reduce the wholesale price of the product and sells it to the R retailer at a lower price. Meanwhile, the R retailer shares the revenue with the M remanufacturer in proportion (C). The R retailer subsidizes the recycling of the T recycler to reduce the recycling unit price of the T recycler, and the T recycler shares the revenue with the R retailer in proportion (D).
Main equations of the SD model of the quality control–revenue-sharing contract scheme:
M revenue sharing = MR revenue sharing = R Income * C
T revenue sharing = RT revenue sharing = T Income * D
Wholesale unit price = 100-MR Revenue sharing coordinates unit price difference
Recycle unit price = 13-RT Revenue sharing coordinates unit price difference
The main parameters of the SD model of the quality control–revenue-sharing contract scheme are listed in
Table 5.
When determining the revenue-sharing ratios C and D for the tuning test, the Vensim DSS software is utilized. The assigned ranges for C and D are 0 to 1. The tuning test is conducted in the following sequence:
First, with C set to 0, the profits of the M remanufacturer, R retailer, T recycler, and CLSC are tested and recorded consecutively, with D increasing from 0 to 1 in a 0.1 step.
Second, C is increased from 0 to 1 in a 0.1 step, and the profits of the CLSC members and the total profit are recorded under different C and D values.
Subsequently, the results are compared and screened, and the smallest profit difference among all parties occurs when C = 0.2 and D = 0.5. Furthermore, all parties’ profits exceed those of the basic SD model.
Next, a more precise valuation of C and D is conducted by repeating steps 1 and 2 with an interval of 0.05. Specifically, C ranges from 0.15 to 0.25, and D ranges from 0.45 to 0.55. The profits are recorded accordingly.
Finally, the results are again compared and screened, leading to the selection of C = 0.15 and D = 0.45 with a value accuracy of 0.01.