Journal of Mathematics

Journal of Mathematics / 2020 / Article

Research Article | Open Access

Volume 2020 |Article ID 7176548 | https://doi.org/10.1155/2020/7176548

Jiayu Shen, "Uncertain Single-Machine Scheduling with Deterioration and Learning Effect", Journal of Mathematics, vol. 2020, Article ID 7176548, 10 pages, 2020. https://doi.org/10.1155/2020/7176548

Uncertain Single-Machine Scheduling with Deterioration and Learning Effect

Academic Editor: Ada Che
Received03 Feb 2020
Accepted06 Apr 2020
Published07 May 2020

Abstract

A single-machine scheduling problem with deterioration and learning effect is studied in the present paper. The processing time and due date are considered uncertain variables due to lack of historical data. The aim is to minimize the makespan, total completion time, total weight completion time, and maximum lateness under an uncertain environment. To address the problem in an uncertain environment, the expected value model and pessimistic value model are developed. These models can be converted into equivalent models based on the inverse distribution method. It is proved that the corresponding dispatching rules can solve the problem optimally under different objective criteria. Finally, sensitivity analysis is used to illustrate the effectiveness of these rules.

1. Introduction

Single-machine scheduling is the basis of all scheduling problems, and many complex scheduling problems can be simplified into several single-machine problems. In recent years, many literature studies on scheduling are mainly concerned with the single scheduling problem. The traditional scheduling problem usually assumes that the processing time is a constant. This assumption may be impractical in many cases because the processing time may be prolonged or shortened as time passes. Some examples about deterioration or learning effect were mentioned in [1]. Raw material in the queue waiting to be processed deteriorates as time passes, which may lead to a certain loss (job deterioration). On an assembly line, workers become more skilled over time (learning effect). Many scholars had studied the scheduling problem with deterioration and learning effect [217].

Since environmental impact and resource constraint, emergencies can lead to uncertainty in processing time and due date. Many researchers studied the scheduling problems under stochastic or fuzzy environment. Optimal permutation policies in single-machine stochastic scheduling problems with learning effects were investigated [18]. A sequence-dependent single-machine scheduling problem with different learning effects was considered [19]. The processing times, setup times, and reliabilities/unreliabilities were considered as random variables. Li [20] studied a single-machine scheduling problem with random processing time and job-based learning effect recently. Ahmadizar and Hosseini [21, 22] considered single-machine scheduling problems with a position-based learning effect and fuzzy processing times. Toksari and Arık [23] considered single-machine scheduling problems under position-dependent fuzzy learning effect with fuzzy processing times, and polynomially solvable algorithms for different confidence levels for these problems were proposed.

Most people believe that probability distribution is easy to obtain from the historical data, and then, we should use probability theory. However, the distribution function obtained in most practical problems is not close enough to the actual frequency. In this case, we have to invite the domain expert to assess the degree of belief that each event will occur. For example, a worker can only assess the processing time of a job based on his past experience. To address personal belief degrees, uncertainty theory was founded by Liu [24] in 2007 and refined it in 2010 [25]. Uncertainty theory is a branch of axiomatic mathematics. It has been developed in many fields such as uncertain calculus [15, 2628], uncertain risk analysis [2931], uncertain differential equation [3235], and uncertain programming [3640].

This paper studies the impacts of both deterioration and learning effect in a single-machine scheduling problem under an uncertain environment. The processing time and due date are considered uncertain variables. For example, in the rough machining operation process, due to machine failure, operation error, environmental impact, and many other uncertainties, workers usually evaluate the processing time according to their own experience. Much asymmetric information in the production and logistics process of the supply chain may lead to uncertainty and cause delays of the due date. It is more suitable for production practice to assume these parameters to be uncertain variables, which can ensure the feasibility and effectiveness of scheduling. The objective functions are makespan, total completion time, total weight completion time, and maximum lateness, respectively. Under different decision criteria, the expected value model and pessimistic value model are developed. Once uncertain distributions are given, the deterministic equivalence of the models is obtained by the uncertainty theory. To address the models effectively, corresponding dispatching rules are proposed.

The rest of the paper is organized as follows. To understand this paper better, basic definitions and results about uncertainty theory are introduced in Section 2. In Section 3, the problem under an uncertain environment is described and the mathematical models are constructed. In Section 4, to facilitate computation, the equivalence of the models is investigated and dispatching rules for the optimal solutions are presented. Section 5 gives an example of sensitivity analysis.

2. Preliminary

To describe an uncertain variable which refers to human uncertainty, Liu [24] established the uncertainty theory and it has been developed well up to now. Some necessary preliminary concepts about uncertainty theory are given.

Let be a nonempty set, is a -algebra over , and each element in is called an event. A set function from to is called an uncertain measure if it satisfies normality axiom, duality axiom, subadditivity axiom, and product axiom [24, 41].

The uncertain distribution of an uncertain variable is defined by for any real number . The uncertain variables are said to be independence (Liu [41]) iffor any Borel sets of real numbers.

Uncertain measure has the following two useful properties: (i) for any events , we have , and (ii) for any events and , we have

Definition 1. (see [24]). Let ξ be an uncertain variable, and . Thenis called the -optimistic value to , andis called the -pessimistic value to .

Definition 2 (see [24]). An uncertain distribution is said to be regular if its inverse function exists and is unique for each . Then, the inverse function is called the inverse uncertainty distribution of .

Theorem 1 (see [24]). Assume the constraint function is strictly increasing with respect to and strictly decreasing with respect to . If are independent uncertain variables with uncertainty distributions , respectively, then the chance constraintholds if and only if

Theorem 2 (see [24]). Let be an uncertain variable. Then, the expected value of is defined byprovided that at least one of the two integrals is finite.

If and are independent uncertain variables, thenholds for any real numbers and .

3. Mathematical Models

The definition of the uncertain single-machine scheduling problem with deterioration and learning effect is given as follows.

Assume that independent and nonpreemptive jobs are processed on a single machine. All jobs are available at time 0. Only one job can be processed at a time. The basic processing time of job is , the due date of job is , and the weight of job is . The deterioration function is an increasing function and is a differentiable convex function. We consider the similar actual processing time proposed by Lee [42]. The actual processing time of job is , where is the learning coefficient and is the start processing time of job .

The processing time, due date, and learning coefficient are assumed to be uncertain variables. In the real operating environment, the forgotten effect leads to job deterioration due to the frequent rotation of the task. With the increase of time and the accumulation of experience, workers become more skilled in operating the machine. Before constructing the mathematical model of the uncertain single-machine problem, the parameters for models are given as follows (Table 1 lists notations for the proposed models):


NotationsDefinitions

The index of job
The index of position
Inherent processing time of job
The due date of job
The weight of job
Deterioration function
The processing time of the th job on the machine
Learning coefficient
The makespan
The completion time of job
The completion time of the th job on the machine
The total completion time
The total weighted completion time
The maximum lateness

Decision variables:

The relationship of variables is as follows:

Equation (11) define the actual processing time of the th job on the machine. Clearly, the processing time changes with the position. Equation (12) define the completion time of the th job on the machine. Equation (13) reveal the relationship between the completion times.

When the processing time, due date, and learning coefficient are uncertain variables, how can we obtain an optimal solution in such an uncertain environment? First, we can use the uncertainty expectation to measure the uncertain variables. The expected value is the average of the uncertain variables on the uncertainty measure.

The problem is modeled under the expected value criterion:

This model minimizes the expected values of these objective functions subject to some constraints, respectively. Equation (14) ensure that each job can only be assigned to one position on the machine. Equation (14) ensure that each position can only process one job and there is no spare time on the machine. Constraint (14) gives the range of the decision variables.

In reality, the decision maker always tries to find the optimal solution within a small deviation. A preset target is determined such that there exists a solution satisfying . For example, let , the decision maker should determine a target and then choose a solution which satisfies . It means that if the decision maker chooses the solution , the objective function will be less than or equal to at least . A pessimistic value model is conceived:where are preset confidence levels.

This model is to calculate the pessimistic value to . The detailed proof will be obtained in the next section. Constraint (15) ensure that the conditions hold under confidence levels , respectively.

4. Equivalence of Uncertain Models

It is noted that the above two models contain many uncertain variables, so it is difficult to solve them directly. In most uncertain programming literature studies, GA and PSO are used to find the approximate optimal solution. Although the effectiveness of these approaches is often illustrated by numerical experiments, the preciseness and generality of these approaches are not satisfactory. The uncertain models can be converted into deterministic forms with the help of uncertainty theory.

Theorem 3. The expected value model can be converted into the following model equivalently:where represents the inverse uncertainty distribution of .

Proof. Since the expectation operator is linear and according to Jensen’s inequality, it yieldsSimilarly, other objective functions can be obtained.

Theorem 4. The pessimistic value model can be converted into the following model equivalently:where are preset confidence levels and represents the inverse uncertainty distribution of .

Proof. Let be independent uncertain variables with regular uncertainty distributions , respectively. Assume is strictly increasing with respect to , , …, and strictly decreasing with respect to . Let , and according to Definition 2, is an uncertain variable with the inverse uncertainty distributionwhich yields .
is equivalent to . Then, it yieldsSince the uncertainty distribution is an increasing function, it yields is equivalent towhere is a given confidence level.
According to the above conclusion, we can obtain the equivalent form of the pessimistic value model:where are preset confidence levels and represents the inverse uncertainty distribution of .
According to the definition of optimistic value (Definition 1), equation (23) is equivalent to equation (18), thus (23) are equivalent to equation (18), equations (23) are equivalent to equation (18), equations (23) are equivalent to equation (18). Thus, the theorem is proved.

Theorem 5. The optimal schedule of the makespan problem under the uncertain environment is obtained by the shortest processing time (SPT) rule.

Proof. An optimal schedule which consists of two adjacent jobs ( precedes ) is considered. Assume the start processing time of job is . Let and represent the completion times of the jobs and , respectively.(1)Under the expectation value criterion :By exchanging the position of two jobs, a schedule is obtained:By comparing their sizes and according to Lagrange mean value theorem, it yieldswhere and .Because , , and , it yields . It implies that the expected makespan under is smaller than that under . This result contradicts the optimality of .(2)Under the pessimistic value criterion , , and :Because , and according to the two useful properties of uncertain measure, it yieldsIt implies that the pessimistic value of makespan under is smaller than that under . This result contradicts the optimality of . Hence, the SPT rule gets the optimal solution. The theorem is proved.

Theorem 6. The optimal schedule of problem under the uncertain environment is obtained by the SPT rule.

Proof. Theorem 5 implies that the start processing time of jobs after and can be advanced by changing the order. Therefore, the in is less than that in . This result contradicts the optimality of . The theorem is proved.

Theorem 7. The optimal schedule of problem under the uncertain environment is obtained by the WSPT rule.

Proof. An optimal schedule which consists of two adjacent jobs ( precedes ) is considered, and the jobs are not arranged according to the WSPT rule. Hence, we have .(1)Under the expectation value criterion:By exchanging the position of two jobs, a schedule is obtained:Similar to Theorem 5, it yieldsBecause , , , , and , it yields . It implies that the expected weighted sum under is smaller than that under . This result contradicts the optimality of .(2)Under the pessimistic value criterion , , , and .Because , it yields :Becausethen according to the two useful properties of uncertain measure, it yieldsIt implies that the pessimistic value of under is smaller than that under . This result contradicts the optimality of . Hence, the SPT rule gets the optimal solution. The theorem is proved.

Theorem 8. The optimal schedule of problem under uncertain environment is obtained by the EDD rule.

Proof. An optimal schedule which consists of two adjacent jobs ( precedes ) is considered, and the jobs are not arranged according to the EDD rule.(1)Under the expectation value criterion, we have . It implies that :By exchanging the position of two jobs, a schedule is obtained:Because and , it yieldsIt implies that EDD is optimal.(2)Under the pessimistic value criterion .Because , it yields .And becauseit yieldsIt implies that the optimal solution can be obtained by the EDD rule. The theorem is proved.Note that these theorems have a higher dependence on the confidence level in the case of pessimistic value. For instance, the condition that the confidence level is greater than 0.75 is sufficient for the result of Theorem 7. If the confidence level is less than 0.75, the required result may not be deduced based on the current method in the proof of Theorem 7. Also the confidence level greater than 0.75 shows that the conclusion holds at a high confidence level. It is reasonable since we generally do not consider the situation that a result holds at a lower confidence level in reality. The pessimistic value case also shows that the decision maker is more inclined to believe that the rule can effectively obtain the optimal solution when faced with making decisions.

5. Sensitivity Analysis

Numerical examples are conducted to investigate the sensitivity of these confidence levels. Assume that there are 30 jobs; ; processing time of job is generated from uncertain normal distribution: for ; due date of job is generated from uncertain normal distribution: for ; ; obeys normal distribution ;and deterioration function . To examine the sensitivity of confidence levels , , Figures 14 indicate that the objective values are nondecreasing with respect to these confidence levels. It can be seen that these confidence levels must be greater than 0.6. Since the confidence level is too low, there is no guarantee that the conclusion of the theorem will be established in the case of pessimistic value.

6. Conclusions

An uncertain single-machine scheduling problem with deterioration and learning effect was studied in this paper. The processing time, due date, and learning coefficient were considered uncertain variables. Two different criteria (expected value and pessimistic value) multiobjective programming models were proposed. The equivalence of the two models was proved based on the uncertainty inverse distribution method. Due to the complexity of the problem, different dispatching rules was used to solve the problem. It was proved that the SPT rule can solve the makespan and total completion time problems under different uncertain criteria optimally, and the total weight completion time and maximum lateness can be solved by WSPT and EDD rules optimally, respectively. The results also suggested that the decision maker will choose the scheme with relatively high confidence level. At the last, the sensitivity analysis revealed that the objective function values are nondecreasing with respect to the confidence levels. For future work, it is an interesting research to investigate the influence of these uncertain variables on other workshop environment (e.g., parallel machine, flow shop, and job shop).

Data Availability

The data used to support the findings of this study are included within the article.

Conflicts of Interest

The author declares that there are no conflicts of interest regarding the publication of this paper.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (no. 61673011) and Research Foundation of NIIT (YK18-10-02 and YK18-10-03).

References

  1. K. McKay, M. Pinedo, and S. Webster, “Practice-focused research issues for scheduling systems,” Production & Operations Management, vol. 11, pp. 249–258, 2002. View at: Google Scholar
  2. J. Bai, Z.-R. Li, and X. Huang, “Single-machine group scheduling with general deterioration and learning effects,” Applied Mathematical Modelling, vol. 36, no. 3, pp. 1267–1274, 2012. View at: Publisher Site | Google Scholar
  3. T. C. E. Cheng, S.-C. Tseng, P.-J. Lai, and W.-C. Lee, “Single-machine scheduling with accelerating learning effects,” Mathematical Problems in Engineering, vol. 2013, Article ID 816235, 7 pages, 2013. View at: Publisher Site | Google Scholar
  4. X. Huang, G. Li, Y. Huo, and P. Ji, “Single machine scheduling with general time-dependent deterioration, position-dependent learning and past-sequence-dependent setup times,” Optimization Letters, vol. 7, no. 8, pp. 1793–1804, 2013. View at: Publisher Site | Google Scholar
  5. W.-H. Kuo and D.-L. Yang, “Single-machine scheduling problems with the time-dependent learning effect,” Computers & Mathematics with Applications, vol. 53, no. 11, pp. 1733–1739, 2007. View at: Publisher Site | Google Scholar
  6. W.-C. Lee, “Single-machine scheduling with past-sequence-dependent setup times and general effects of deterioration and learning,” Optimization Letters, vol. 8, no. 1, pp. 135–144, 2014. View at: Publisher Site | Google Scholar
  7. M. Toksarı, D. Oron, and E. Güner, “Single machine scheduling problems under the effects of nonlinear deterioration and time-dependent learning,” Mathematical and Computer Modelling, vol. 50, no. 3-4, pp. 401–406, 2009. View at: Publisher Site | Google Scholar
  8. J.-B. Wang, “Single-machine scheduling problems with the effects of learning and deterioration,” Omega, vol. 35, no. 4, pp. 397–402, 2007. View at: Publisher Site | Google Scholar
  9. X. Wang and T. C. Edwin Cheng, “Single-machine scheduling with deteriorating jobs and learning effects to minimize the makespan,” European Journal of Operational Research, vol. 178, no. 1, pp. 57–70, 2007. View at: Publisher Site | Google Scholar
  10. J.-B. Wang, Y. Jiang, and G. Wang, “Single-machine scheduling with past-sequence-dependent setup times and effects of deterioration and learning,” The International Journal of Advanced Manufacturing Technology, vol. 41, no. 11-12, pp. 1221–1226, 2009. View at: Publisher Site | Google Scholar
  11. J.-B. Wang, “Single machine scheduling with a time-dependent learning effect and deteriorating jobs,” Journal of the Operational Research Society, vol. 60, no. 4, pp. 583–586, 2009. View at: Publisher Site | Google Scholar
  12. J.-B. Wang and C. Wang, “Single-machine due-window assignment problem with learning effect and deteriorating jobs,” Applied Mathematical Modelling, vol. 35, no. 8, pp. 4017–4022, 2011. View at: Publisher Site | Google Scholar
  13. L.-Y. Wang and E.-M. Feng, “A note on single-machine scheduling problems with the effects of deterioration and learning,” The International Journal of Advanced Manufacturing Technology, vol. 59, no. 5–8, pp. 539–545, 2012. View at: Publisher Site | Google Scholar
  14. C.-C. Wu and W.-C. Lee, “Single-machine group-scheduling problems with deteriorating setup times and job-processing times,” International Journal of Production Economics, vol. 115, no. 1, pp. 128–133, 2008. View at: Publisher Site | Google Scholar
  15. D.-L. Yang and W.-H. Kuo, “Single-machine scheduling with both deterioration and learning effects,” Annals of Operations Research, vol. 172, no. 1, pp. 315–327, 2009. View at: Publisher Site | Google Scholar
  16. Y. Yin and D. Xu, “Some single-machine scheduling problems with general effects of learning and deterioration,” Computers & Mathematics with Applications, vol. 61, no. 1, pp. 100–108, 2011. View at: Publisher Site | Google Scholar
  17. X. Zhang and G. Yan, “Single-machine group scheduling problems with deteriorated and learning effect,” Applied Mathematics and Computation, vol. 216, no. 4, pp. 1259–1266, 2010. View at: Publisher Site | Google Scholar
  18. Y. Zhang, X. Wu, and X. Zhou, “Stochastic scheduling problems with general position-based learning effects and stochastic breakdowns,” Journal of Scheduling, vol. 16, no. 3, pp. 331–336, 2013. View at: Publisher Site | Google Scholar
  19. H. M. Soroush, “Stochastic bicriteria single machine scheduling with sequence-dependent job attributes and job-dependent learning effects,” European Journal of Industrial Engineering, vol. 8, no. 4, pp. 421–456, 2014. View at: Publisher Site | Google Scholar
  20. H. Li, “Stochastic single-machine scheduling with learning effect,” IEEE Transactions on Engineering Management, vol. 64, no. 1, pp. 94–102, 2017. View at: Publisher Site | Google Scholar
  21. F. Ahmadizar and L. Hosseini, “Single-machine scheduling with a position-based learning effect and fuzzy processing times,” The International Journal of Advanced Manufacturing Technology, vol. 56, no. 5–8, pp. 693–698, 2011. View at: Publisher Site | Google Scholar
  22. F. Ahmadizar and L. Hosseini, “Minimizing makespan in a single-machine scheduling problem with a learning effect and fuzzy processing times,” The International Journal of Advanced Manufacturing Technology, vol. 65, no. 1–4, pp. 581–587, 2013. View at: Publisher Site | Google Scholar
  23. M. D. Toksari and O. A. Arık, “Single machine scheduling problems under position-dependent fuzzy learning effect with fuzzy processing times,” Journal of Manufacturing Systems, vol. 45, pp. 159–179, 2017. View at: Publisher Site | Google Scholar
  24. B. Liu, Uncertainty Theory, Springer-Verlag, Berlin, Germany, 2nd edition, 2007.
  25. B. Liu, Uncertainty Theory: A Branch of Mathematics for Modeling Human Uncertainty, Springer-Verlag, Berlin, Germany, 2010.
  26. X. Chen and D. A. Ralescu, “Liu process and uncertain calculus,” Journal of Uncertainty Analysis and Applications, vol. 1, no. 1, 2013. View at: Publisher Site | Google Scholar
  27. X. Chen, “Uncertain calculus with finite variation processes,” Soft Computing, vol. 19, no. 10, pp. 2905–2912, 2015. View at: Publisher Site | Google Scholar
  28. K. Yao, “Multi-dimensional uncertain calculus with liu process,” Journal of Uncertain Systems, vol. 8, no. 4, pp. 244–254, 2014. View at: Google Scholar
  29. Z. Chen, Y. Lan, and R. Zhao, “Impacts of risk attitude and outside option on compensation contracts under different information structures,” Fuzzy Optimization and Decision Making, vol. 17, no. 1, pp. 13–47, 2018. View at: Publisher Site | Google Scholar
  30. Y. Liu and D. A. Ralescu, “Value-at-risk in uncertain random risk analysis,” Information Sciences, vol. 391-392, pp. 1–8, 2017. View at: Publisher Site | Google Scholar
  31. J. Zhou, Y. Liu, X. Zhang, X. Gu, and D. Wang, “Uncertain risk aversion,” Journal of Intelligent Manufacturing, vol. 28, no. 3, pp. 615–624, 2017. View at: Publisher Site | Google Scholar
  32. X. Chen and Y. Ning, “The th moment exponential stability of uncertain differential equation,” Journal of Intelligent & Fuzzy Systems, vol. 33, no. 2, pp. 725–732, 2017. View at: Publisher Site | Google Scholar
  33. Y. Feng, X. Yang, and G. Cheng, “Stability in mean for multi-dimensional uncertain differential equation,” Soft Computing, vol. 22, no. 17, pp. 5783–5789, 2018. View at: Publisher Site | Google Scholar
  34. H. Liu and W. Fei, “Neutral uncertain delay differential equations,” Information: An International Interdisciplinary Journal, vol. 16, no. 2, pp. 1225–1232, 2013. View at: Google Scholar
  35. K. Yao, J. Gao, and Y. Gao, “Some stability theorems of uncertain differential equation,” Fuzzy Optimization and Decision Making, vol. 12, no. 1, pp. 3–13, 2013. View at: Publisher Site | Google Scholar
  36. Q. Cui and Y. Sheng, “Uncertain programming model for solid transportation problem,” Information: An International Interdisciplinary Journal, vol. 16, no. 2, pp. 1207–1214, 2013. View at: Google Scholar
  37. R. Mu, Y. Lan, and W. Tang, “An uncertain contract model for rural migrant worker’s employment problems,” Fuzzy Optimization and Decision Making, vol. 12, no. 1, pp. 29–39, 2013. View at: Publisher Site | Google Scholar
  38. J. Shen and Y. Zhu, “Uncertain flexible flow shop scheduling problem subject to breakdowns,” Journal of Intelligent & Fuzzy Systems, vol. 32, no. 1, pp. 207–214, 2017. View at: Publisher Site | Google Scholar
  39. J. Shen, “An uncertain parallel machine problem with deterioration and learning effect,” Computational and Applied Mathematics, vol. 38, no. 1, pp. 1–17, 2019. View at: Publisher Site | Google Scholar
  40. X. Zhang, L. Li, and G. Meng, “A modified uncertain entailment model,” Journal of Intelligent & Fuzzy Systems, vol. 27, no. 1, pp. 549–553, 2014. View at: Publisher Site | Google Scholar
  41. B. Liu, “Some research problems in uncertainty theory,” Journal of Uncertain Systems, vol. 3, no. 1, pp. 3–10, 2009. View at: Google Scholar
  42. W. Lee, “A note on deteriorating jobs and learning in single-machine scheduling problems,” International Journal of Business and Economics, vol. 3, pp. 83–89, 2004. View at: Google Scholar

Copyright © 2020 Jiayu Shen. 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.


More related articles

 PDF Download Citation Citation
 Download other formatsMore
 Order printed copiesOrder
Views123
Downloads165
Citations

Related articles

We are committed to sharing findings related to COVID-19 as quickly as possible. We will be providing unlimited waivers of publication charges for accepted research articles as well as case reports and case series related to COVID-19. Review articles are excluded from this waiver policy. Sign up here as a reviewer to help fast-track new submissions.