Skip to main content

Advertisement

Log in

Optimization of preventive maintenance for a multi-state degraded system by monitoring component performance

  • Published:
Journal of Intelligent Manufacturing Aims and scope Submit manuscript

Abstract

This study aims to construct an optimal preventive maintenance model for a multi-state degraded system under the condition that individual components or sub-systems can be monitored in real time. Given the requirement of minimum system availability, the total maintenance cost is minimized by determining the maintenance activities of components in degraded states. The general non-homogeneous continuous-time Markov model (NHCTMM) and its analogous Markov reward model (NHCTMRM) are used to quantify the intensity of state transitions during the degradation process, allowing the determination of various performance indicators. The bound approximation approach is applied to solve the established NHCTMMs and NHCTMRMs, thus obtaining instantaneous system state probabilities to overcome their inherent computational difficulties. Furthermore, this study utilizes a genetic algorithm to optimize the proposed model. A simulation illustrates the feasibility and practicability of the proposed approach.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10

Similar content being viewed by others

Abbreviations

NHCTMM:

Non-homogeneous continuous time Markov model

NHCTMRM:

Non-homogeneous continuous time Markov reward model

GA:

Genetic algorithm

MTTF:

Mean time to failure

CTMC:

Continuous-time Markov chain

TER:

Total expected reward

\(X(t)\) :

System state at time \(t\)

\(S\) :

Set of all possible states of system

\(p_{i}(t)\) :

Probability of state \(i\) occurring at time \(t\)

\(\upalpha _{ij}\) :

Intensity transiting from state \(i\) to state \(j\)

\(\mathbf{A}\) :

Transition intensity matrix

\(\mathbf{p(t)}\) :

Set of probabilities associated with different states at time \(t\)

\(\lambda _{i,j}(t)\) :

Failure rate transiting from state \(i\) to state \(j\) at time \(t\)

\(g_{k}\) :

Performance rate of the component under state \(k\)

\(G(t)\) :

Random variable representing the performance rate at time \(t\)

\(r_{ii}\) :

Reward staying in any state \(i\) in cost units per time unit

\(r_{ij}\) :

Reward transiting from state \(i\) to state \(j\) in cost units

\(\mathbf{r}\) :

Reward matrix

\(V_{i}(t)\) :

Expected total reward in state \(i\) accumulated until to time \(t\)

\(\Delta t\) :

Duration of each time interval

T:

System lifetime

N:

The number of time intervals

\(t_{n}\) :

Each time interval of the bound approximation approach

\(\lambda ^{n-}\) :

System failure rate at the end of the nth time interval

\(\lambda ^{n+}\) :

System failure rate at the beginning of the nth time interval

\(P_{j}^{n-}\) :

Probability in state \(j\) at the end of the nth time interval

\(P_{j}^{n+}\) :

Probability in state \(j\) at the beginning of the nth time interval

\(\alpha _{ij}^{n-}\) :

Intensity transiting from state \(i\) to state \(j\) at the end of the nth time interval

\(\alpha _{ij}^{n+}\) :

Intensity transiting from state \(i\) to state \(j\) at the beginning of the nth time interval

\(V_{i}^{n-}\) :

Expected reward in state \(i\) at the end of the nth time interval

\(V_{i}^{n+}\) :

Expected reward in state \(i\) at the beginning of the nth time interval

\(V^{n-}\) :

Lower bound expected total reward of the nth time interval

\(V^{n+}\) :

Upper bound expected total reward of the nth time interval

\( TER ^{n-}\) :

Lower bound expected total reward

\( TER ^{n+}\) :

Upper bound expected total reward

\(C_{ tot }\) :

Total maintenance cost

\(m\) :

Total number of the components constituting the system

\(C_{pm,l}\) :

The cost for implementing preventive maintenance of the \(l\)th component

\(A(t)\) :

System availability at time \(t\)

\(As\) :

Minimum allowable system availability

\(K\) :

Best performance state of the system

\(g_{i}\) :

System performance at state \(i\)

\(w\) :

User demand

\(1(g_{i} \ge w)\) :

Unit function that takes a value of 1 when \(g_{i}\) is greater than or equal to \(w\), and a value of 0 otherwise

\(\lambda _{i,j}^l (t)\) :

Transition intensity of component \(l\) when degraded from state \(i\) to \(j\) at time \(t\)

\(\mu _{j,i}^l \) :

Transition intensity of component \(l\) when restored from state \(j\) back to state \(i\)

References

  • Arab, A., Ismail, N., & Lee, L. S. (2013). Maintenance scheduling incorporating dynamics of production system and real-time information from workstations. Journal of Intelligent Manufacturing, 24(4), 695–705.

    Article  Google Scholar 

  • Ding, Y., Lisnianski, A., Frenkel, I., & Khvatskin, L. (2009). Optimal corrective maintenance contract planning for aging multi-state system. Applied Stochastic Models in Business and Industry, 25(5), 612–631.

    Article  Google Scholar 

  • Howard, R. A. (1972). Dynamic programming and Markov process. Cambridge, MA: MIT.

    Google Scholar 

  • Huang, C. C., & Yuan, J. (2010). A two-stage preventive maintenance policy for a multi-state deterioration system. Reliability Engineering & System Safety, 95, 1255–1260.

    Article  Google Scholar 

  • Goldberg, D. (1989). Genetic algorithms in search, optimization and machine learning. Reading, MA: Addison-Wesley.

    Google Scholar 

  • Khatab, A., Rezg, N., & Ait-Kadi, D. (2011). Optimum block replacement policy over a random time horizon. Journal of Intelligent Manufacturing, 22(6), 885–889.

    Article  Google Scholar 

  • Levitin, G. (2005). The universal generating function in reliability analysis and optimization. London: Springer.

  • Levitin, G., & Lisnianski, A. (1999). Joint redundancy and maintenance optimization for multi-state series-parallel systems. Reliability Engineering & System Safety, 64(1), 33–42.

    Article  Google Scholar 

  • Levitin, G., & Lisnianski, A. (2000). Optimization of imperfect preventive maintenance for multi-state systems. Reliability Engineering & System Safety, 67(2), 193–203.

    Article  Google Scholar 

  • Liao, W., Pan, E., & Xi, L. (2010). Preventive maintenance scheduling for repairable system with deterioration. Journal of Intelligent Manufacturing, 21(6), 875–884.

    Article  Google Scholar 

  • Lin, T. W., & Wang, C. H. (2012). A hybrid genetic algorithm to minimize the periodic preventive maintenance cost in a series-parallel system. Journal of Intelligent Manufacturing, 23(4), 1225–1236.

    Article  Google Scholar 

  • Lisnianski, A. (2007). Extended block diagram method for a multi-state system reliability assessment. Reliability Engineering & System Safety, 92(12), 1601–1607.

    Article  Google Scholar 

  • Lisnianski, A., Frenkel, I., Khvatskin, L., & Ding, Y. (2008). Maintenance contract assessment for aging systems. Quality & Reliability Engineering International, 24(5), 519–531.

    Article  Google Scholar 

  • Lisnianski, A., Frenkel, I., & Ding, Y. (2010). Multi-state system reliability analysis and optimization for engineers and industrial managers. London: Springer.

    Book  Google Scholar 

  • Liu, Y., & Huang, H. (2010). Optimal replacement policy for multi-state system under imperfect maintenance. Transactions on Reliability, 59(3), 483–495.

  • Liu, Y., & Kapur, K. C. (2007). Customer’s cumulative experience measures for reliability of non-repairable aging multi-state systems. Quality Technology & Quantitative Management, 4(2), 225–234.

    Article  Google Scholar 

  • Liu, Y., & Kapur, K. C. (2008). New model and measurement for reliability of multi-state systems. In Handbook of performability engineering (pp. 431–445). London: Springer.

  • Nagayama, S., Sasao, T., & Butler, J. T. (2014). EVMDD-based analysis and diagnosis methods of multi-state systems with multi-state components. Journal of Multiple-Valued Logic and Soft Computing, 22(1–2), 59–78.

    Google Scholar 

  • Nahas, N., Khatab, A., Ait-Kadi, D., & Nourelfath, M. (2008). Extended great deluge algorithm for the imperfect preventive maintenance optimization of multi-state systems. Reliability Engineering & System Safety, 93(11), 1658–1672.

    Article  Google Scholar 

  • Pascual, R., & Ortega, J. H. (2006). Optimal replacement and overhaul decisions with imperfect maintenance and warranty contracts. Reliability Engineering and System Safety, 91, 241–248.

    Article  Google Scholar 

  • Sheikhalishahi, M., Ebrahimipour, V., & Farahani, M. H. (2013). An integrated GA-DEA algorithm for determining the most effective maintenance policy for a \(k\)-out-of-\(n\) problem. Journal of Intelligent Manufacturing. doi:10.1007/s10845-013-0752-z.

  • Soro, I. W., Nourelfath, M., & Aït-Kadi, D. (2010). Performance evaluation of multi-state degraded systems with minimal repairs and imperfect preventive maintenance. Reliability Engineering and System Safety, 95, 65–69.

    Article  Google Scholar 

  • Tan, C. M., & Raghavan, N. (2008). A framework to practical predictive maintenance modeling for multi-state systems. Reliability Engineering & System Safety, 93(8), 1138–1150.

    Article  Google Scholar 

  • Tian, Z., Lin, D., & Wu, B. (2012). Condition based maintenance optimization considering multiple objectives. Journal of Intelligent Manufacturing, 23(2), 333–340.

    Article  Google Scholar 

  • Trivedi, K. S. (2008). Probability and statistics with reliability, queuing and computer science application (2nd ed.). India: Wiley.

  • Voisin, A., Levrat, E., Cocheteux, P., & Iung, B. (2010). Generic prognosis model for proactive maintenance decision support: Application to pre-industrial e-maintenance test bed. Journal of Intelligent Manufacturing, 21(2), 177–193.

    Article  Google Scholar 

  • Wang, C. H., & Tsai, S. W. (2012). Optimizing bi-objective imperfect preventive maintenance model for series-parallel system using established hybrid genetic algorithm. Journal of Intelligent Manufacturing. doi:10.1007/s10845-012-0708-8.

  • Zequeira, R. I., & Be’renguer, C. (2006). Periodic imperfect preventive maintenance with two categories of competing failure modes. Reliability Engineering and System Safety, 91, 460–468.

    Article  Google Scholar 

Download references

Acknowledgments

This work was supported by the National Science Council of Taiwan (Grant No. NSC 102-2221-E-606-010).

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Chao-Hui Huang.

Appendices

Appendix 1: Failure-rate function \(\lambda _{i,j}(t)\) and repair rate \(\mu _{j,i}\) of the reduced ten states

$$\begin{aligned} \lambda _{10,9}(t)&= \lambda _{2,1}^{1}(t)+\lambda _{3,1}^{1}(t)+\lambda _{4,1}^{1}(t) +\lambda _{5,1}^{1}(t)\\&\quad +\,\lambda _{4,3}^{1}(t)+\lambda _{5,3}^{1}(t)+\lambda _{3,2}^{2}(t)\\&\quad +\,\lambda _{4,2}^{2}(t)+\lambda _{5,2}^{2}(t);\\ \lambda _{10,8}(t)&= 15\lambda _{5,4}^{3}(t);\\ \lambda _{10,7}(t)&= \lambda _{2,1}^{1}(t)+\lambda _{3,1}^{1}(t)+\lambda _{5,1}^{1}(t)\\&\quad +\,\lambda _{5,4}^{1}(t)+\lambda _{2,1}^{2}(t) +\lambda _{3,1}^{2}(t)\!+\!\lambda _{4,1}^{2}(t)\!+\!\lambda _{5,1}^{2}(t);\\ \lambda _{10,6}(t)&= \lambda _{2,1}^{1}(t)+\lambda _{3,1}^{1}(t)+\lambda _{4,1}^{1}(t)\\&\quad +\,\lambda _{5,1}^{1}(t)+\lambda _{4,2}^{1}(t)+\lambda _{5,2}^{1}(t)+\lambda _{3,2}^{2}(t)\\&\quad +\,\lambda _{4,2}^{2}(t)+\lambda _{5,2}^{2}(t);\\ \lambda _{10,5}(t)&= \lambda _{5,3}^{1}(t)+\lambda _{3,1}^{2}(t)+\lambda _{4,1}^{2}(t)\\&\quad +\,\lambda _{5,1}^{2}(t)+15\lambda _{5,3}^{3}(t);\\ \lambda _{10,4}(t)&= \lambda _{5,2}^{1}(t)+\lambda _{3,1}^{2}(t) +\lambda _{4,1}^{2}(t)+\lambda _{5,1}^{2}(t);\\ \lambda _{10,3}(t)&= \lambda _{4,1}^{1}(t)+\lambda _{5,1}^{1}(t);\\ \lambda _{10,2} (t)&= 15\lambda _{5,2}^{3} (t);\\ \lambda _{10,1}(t)&= \lambda _{5,1}^{1}(t)+15\lambda _{5,1}^{3}(t);\\ \lambda _{9,8}(t)&= 2\lambda _{{5},{4}}^{3} (t);\lambda _{9,7}(t)=\lambda _{5,4}^{2}(t);\\ \lambda _{9,6}(t)&= \lambda _{3,2}^{1}(t)+\lambda _{5,3}^{2}(t);\\ \lambda _{9,5}(t)&= \lambda _{2,1}^{2}(t)+\lambda _{5,3}^{3}(t);\\ \lambda _{9,3}(t)&= \lambda _{3,1}^{1}(t)+\lambda _{5,2}^{2}(t);\\ \lambda _{9,2}(t)&= 2\lambda _{5,2}^{3}(t);\\ \lambda _{9,1}(t)&= \lambda _{5,1}^{2}(t)+2\lambda _{5,1}^{3}(t);\\ \lambda _{8,7}(t)&= \lambda _{2,1}^{1}(t)+\lambda _{3,1}^{1}(t)+\lambda _{5,1}^{1}(t)+\lambda _{5,4}^{1}(t)\\&\quad +\,\lambda _{2,1}^{2}(t)+\lambda _{3,1}^{2}(t)+\lambda _{4,1}^{2}(t)\!+\!\lambda _{5,1}^{2}(t)\!+\!\lambda _{5,4}^{2}(t);\\ \lambda _{8,6}(t)&= \lambda _{2,1}^1 (t)+\lambda _{3,1}^1 (t)+\lambda _{4,1}^1 (t)+\lambda _{5,1}^1 (t)+\lambda _{3,2}^1(t)\\&\quad +\,\lambda _{4,2}^1 (t)+\lambda _{5,2}^1 (t)+\lambda _{3,2}^2 (t)+\lambda _{4,2}^2 (t)+\lambda _{5,2}^2(t)\\ \lambda _{8,5}(t)&= \lambda _{5,3}^{1}(t)+\lambda _{3,1}^{2}(t)\!+\!\lambda _{4,1}^{2}(t)\!+\!\lambda _{5,1}^{2}(t)+17\lambda _{4,3}^{3}(t);\\ \lambda _{8,4}(t)&= \lambda _{5,2}^{1}(t)+\lambda _{3,1}^{2}(t)+\lambda _{4,1}^{2}(t)+\lambda _{5,1}^{2}(t);\\ \lambda _{8,3}(t)&= \lambda _{3,1}^{1}(t)+\lambda _{4,1}^{1}(t)+\lambda _{5,1}^{1}(t)+\lambda _{5,2}^{2}(t);\\ \lambda _{8,2}(t)&= 17\lambda _{4,2}^{3} (t);\\ \lambda _{8,1}(t)&= \lambda _{5,1}^{1}(t)+\lambda _{5,1}^{2}(t)+17\lambda _{4,1}^{3}(t);\\ \lambda _{7,6}(t)&= 2\lambda _{4,3}^{2}(t);\\ \lambda _{7,5}(t)&= 2\lambda _{4,3}^{1}(t)+2\lambda _{4,3}^{3}(t)+2\lambda _{5,3}^{3}(t);\\ \lambda _{7,4}(t)&= 2\lambda _{4,2}^1 (t);\lambda _{7,3}(t)=2*\lambda _{4,2}^{2}(t);\\ \lambda _{7,2}(t)&= 2\lambda _{4,2}^{3}(t)+2\lambda _{5,2}^{3}(t);\\ \lambda _{7,1}(t)&= 2\lambda _{4,1}^{1}(t)+2\lambda _{4,1}^{2}(t)+2\lambda _{4,1}^{3}(t)+2\lambda _{5,1}^{3}(t);\\ \lambda _{6,5}(t)&= 2\lambda _{4,3}^{3}(t)+2\lambda _{5,3}^{3}(t);\\ \lambda _{6,4}(t)&= 2\lambda _{2,1}^{2}(t); \end{aligned}$$
$$\begin{aligned} \lambda _{6,3}(t)&= 2\lambda _{2,1}^{1}(t)+2\lambda _{3,2}^{2}(t);\\ \lambda _{6,2}(t)&= 2\lambda _{4,2}^{3}(t)+2\lambda _{5,2}^{3}(t);\\ \lambda _{6,1}(t)&= 3\lambda _{3,1}^{2}(t)\!+\!\lambda _{4,1}^{2}(t)\!+\!\lambda _{5,1}^{2}(t)+2\lambda _{4,1}^{3}(t)+2\lambda _{5,1}^{3}(t);\\ \lambda _{5,4}(t)&= 3\lambda _{3,2}^{1}(t)+\lambda _{4,2}^{1}(t)+\lambda _{5,2}^{1}(t)+\lambda _{2,1}^{2}(t)+\lambda _{3,1}^{2}(t)\\&\quad +\,\lambda _{4,1}^{2}(t)+\lambda _{5,1}^{2}(t);\\ \lambda _{5,3}(t)&= \lambda _{2,1}^{1}(t)+\lambda _{3,1}^{1}(t)+\lambda _{4,1}^{1}(t)+\lambda _{5,1}^{1}(t)\\&\quad +\,\lambda _{3,2}^{2}(t)+\lambda _{4,2}^{2}(t)+\lambda _{5,2}^{2}(t);\\ \lambda _{5,2} (t)&= 22\lambda _{3,2}^{3} (t)+\lambda _{4,2}^{3} (t)+\lambda _{5,2}^{3} (t);\\ \lambda _{5,1}(t)&= 3\lambda _{3,1}^{1}(t)+\lambda _{4,1}^{1}(t)+\lambda _{5,1}^{1}(t)\\&\quad +\,17\lambda _{3,1}^{3}(t)+\lambda _{4,1}^{3}(t)+\lambda _{5,1}^{3}(t);\\ \lambda _{4,2}(t)&= \lambda _{3,2}^{2}(t)+\lambda _{4,2}^{2}(t)+\lambda _{5,2}^{2}(t);\\ \lambda _{4,1}(t)&= 2\lambda _{{2},{1}}^{1} (t)+\lambda _{{3},{1}}^{3} (t)\\&\quad +\,\lambda _{{4},{1}}^{3} (t)+\lambda _{{5},{1}}^{3} (t);\\ \lambda _{3,2}(t)&= \lambda _{3,2}^{3}(t)+\lambda _{4,2}^{3}(t)+\lambda _{5,2}^{3}(t);\\ \lambda _{3,1}(t)&= 3\lambda _{2,1}^{2}(t)+\lambda _{3,1}^{3}(t)+\lambda _{4,1}^{3}(t)+\lambda _{5,1}^{3}(t);\\ \lambda _{2,1}(t)&= 2\lambda _{2,1}^{1}(t)+\lambda _{3,1}^{1}(t)+\lambda _{4,1}^{1}(t)+\lambda _{5,1}^{1}(t)+\lambda _{2,1}^{2}(t)\\&\quad +\,\lambda _{3,1}^{2}(t)+\lambda _{4,1}^{2}(t)+\lambda _{5,1}^{2}(t)+24\lambda _{2,1}^{3}(t);\\ \mu _{9,10}&= \mu _{1,2}^{1}+\mu _{1,3}^{1}+\mu _{1,4}^{1}+\mu _{1,5}^{1}+\mu _{3,4}^{1}+\mu _{3,5}^{1}\\&\quad +\,\mu _{2,4}^{2}+\mu _{2,5}^{2};\\ \mu _{8,10}&= 15\mu _{4,5}^{3};\\ \mu _{7,10}&= \mu _{1,2}^{1}+\mu _{1,3}^{1}+\mu _{1,4}^{1}+\mu _{1,5}^{1}+\mu _{4,5}^{1}\\&\quad +\,\mu _{3,4}^{1}+\mu _{3,5}^{1}+\mu _{2,4}^{2}+\mu _{2,5}^{2};\\ \mu _{6,10}&= \mu _{1,2}^{1}+\mu _{1,3}^{1}+\mu _{1,4}^{1}+\mu _{1,5}^{1}+\mu _{2,4}^{1}+\mu _{2,5}^{1}\\&\quad +\,\mu _{2,4}^{2}+\mu _{2,5}^{2};\\ \mu _{5,10}&= \mu _{3,5}^{1}+\mu _{1,3}^{2}+\mu _{1,4}^{2}+\mu _{1,5}^{2}+15\mu _{3,5}^{3};\\ \mu _{4,10}&= \mu _{2,5}^{1}+\mu _{1,3}^{2}+\mu _{1,4}^{2}+\mu _{1,5}^{2};\\ \mu _{3,10}&= \mu _{1,4}^1 +\mu _{1,5}^1 ;\mu _{2,10}=15\mu _{2,5}^{3};\\ \mu _{1,10}&= 15\mu _{1,5}^{3};\\ \mu _{8,9}&= 2\mu _{4,5}^3 ;\mu _{7,9}=\mu _{4,5}^{2};\\ \mu _{6,9}&= \mu _{2,3}^{1}+\mu _{3,5}^{2};\\ \mu _{5,9}&= 2\mu _{3,5}^{3};\\ \mu _{3,9}&= \mu _{1,3}^{1}+\mu _{2,5}^{2};\\ \mu _{2,9}&= 2\mu _{2,5}^{3};\\ \mu _{1,9}&= \mu _{1,5}^{2}+2\mu _{1,5}^{3};\\ \mu _{7,8}&= \mu _{1,2}^{1}+\mu _{1,3}^{1}+\mu _{1,4}^{1}+\mu _{1,5}^{1}+\mu _{4,5}^{1}\\&\quad +\,\mu _{1,2}^{2}+\mu _{1,3}^{2}+\mu _{1,4}^{2}+\mu _{1,5}^{2}+\mu _{4,5}^{2};\\ \mu _{6,8}&= \mu _{1,2}^{1}+\mu _{1,3}^{1}+\mu _{1,4}^{1}+\mu _{1,5}^{1}+\mu _{2,3}^{1}\\&\quad +\,\mu _{2,4}^{1}+\mu _{2,5}^{1}+\mu _{2,4}^{2}+\mu _{2,5}^{2}+\mu _{3,5}^{2};\\ \mu _{5,8}&= \mu _{3,5}^{1}+\mu _{1,3}^{2}+\mu _{1,4}^{2}+\mu _{1,5}^{2}+17\mu _{3,4}^{3};\\ \mu _{4,8}&= \mu _{2,5}^{1}+\mu _{1,3}^{2}+\mu _{1,4}^{2}+\mu _{1,5}^{2}; \end{aligned}$$
$$\begin{aligned} \mu _{3,8}&= \mu _{1,3}^{1}+\mu _{1,4}^{1}+\mu _{1,5}^{1}+\mu _{2,5}^{2};\\ \mu _{2,8}&= 17\mu _{2,4}^{3};\\ \mu _{1,8}&= \mu _{1,5}^{2}+17\mu _{1,4}^{3};\\ \mu _{7,8}&= \mu _{1,2}^{1}+\mu _{1,3}^{1}+\mu _{1,4}^{1}+\mu _{1,5}^{1}+\mu _{4,5}^{1}\\&\quad +\,\mu _{1,2}^{2}+\mu _{1,3}^{2}+\mu _{1,4}^{2}+\mu _{1,5}^{2}+\mu _{4,5}^{2};\\ \mu _{6,8}&= \mu _{1,2}^{1}+\mu _{1,3}^{1}+\mu _{1,4}^{1}+\mu _{1,5}^{1}+\mu _{2,3}^{1}+\mu _{2,4}^{1}\\&\quad +\,\mu _{2,5}^{1}+\mu _{2,4}^{2}+\mu _{2,5}^{2}+\mu _{3,5}^{2};\\ \mu _{5,8}&= \mu _{3,5}^{1}+\mu _{1,3}^{2}+\mu _{1,4}^{2}+\mu _{1,5}^{2}+17\mu _{3,4}^{3};\\ \mu _{4,8}&= \mu _{2,5}^{1}+\mu _{1,3}^{2}+\mu _{1,4}^{2}+\mu _{1,5}^{2};\\ \mu _{3,8}&= \mu _{1,3}^{1}+\mu _{1,4}^{1}+\mu _{1,5}^{1}+\mu _{2,5}^{2};\\ \mu _{2,8}&= 17\mu _{2,4}^{3};\\ \mu _{1,8}&= \mu _{1,5}^{2}+17\mu _{1,4}^{3};\\ \mu _{7,8}&= \mu _{1,2}^{1}+\mu _{1,3}^{1}+\mu _{1,4}^{1}+\mu _{1,5}^{1}+\mu _{4,5}^{1}+\mu _{1,2}^{2}\\&\quad +\,\mu _{1,3}^{2}+\mu _{1,4}^{2}+\mu _{1,5}^{2}+\mu _{4,5}^{2};\\ \mu _{6,8}&= \mu _{1,2}^{1}+\mu _{1,3}^{1}+\mu _{1,4}^{1}+\mu _{1,5}^{1}+\mu _{2,3}^{1}+\mu _{2,4}^{1}\\&\quad +\,\mu _{2,5}^{1}+\mu _{2,4}^{2}+\mu _{2,5}^{2}+\mu _{3,5}^{2};\\ \mu _{5,8}&= \mu _{3,5}^{1}+\mu _{1,3}^{2}+\mu _{1,4}^{2}+\mu _{1,5}^{2}+17\mu _{3,4}^{3};\\ \mu _{4,8}&= \mu _{2,5}^{1}+\mu _{1,3}^{2}+\mu _{1,4}^{2}+\mu _{1,5}^{2};\\ \mu _{3,8}&= \mu _{1,3}^{1}+\mu _{1,4}^{1}+\mu _{1,5}^{1}+\mu _{2,5}^{2};\\ \mu _{2,8}&= 17\mu _{2,4}^{3};\\ \mu _{1,8}&= \mu _{1,5}^{2}+17\mu _{1,4}^{3};\\ \mu _{6,7}&= 2\mu _{3,4}^{2};\\ \mu _{5,7}&= 2\mu _{3,4}^{1}+2\mu _{3,4}^{3}+2\mu _{3,5}^{3};\\ \mu _{4,7}&= 2\mu _{2,4}^{1};\\ \mu _{3,7}&= 2\mu _{2,4}^{2};\\ \mu _{2,7}&= 2\mu _{2,4}^{3}+2\mu _{2,5}^{3};\\ \mu _{1,7}&= \mu _{1,4}^{1}+2\mu _{1,4}^{2}+2\mu _{1,4}^{3}+2\mu _{1,5}^{3};\\ \mu _{5,6}&= 2\mu _{3,4}^{3}+2\mu _{3,5}^{3};\\ \mu _{4,6}&= 2\mu _{1,2}^{2};\\ \mu _{3,6}&= 2\mu _{1,2}^{1}+2\mu _{2,3}^{2};\\ \mu _{2,6}&= 2\mu _{2,4}^{3}+2\mu _{2,5}^{3};\\ \mu _{1,6}&= 2\mu _{1,3}^{2}+2\mu _{1,4}^{3}+2\mu _{1,5}^{3};\\ \mu _{4,5}&= 3\mu _{2,3}^{1}+\mu _{2,4}^{1}+\mu _{2,5}^{1}+\mu _{1,2}^{2}\\&\quad +\,\mu _{1,3}^{2}+\mu _{1,4}^{2}+\mu _{1,5}^{2};\\ \mu _{3,5}&= \mu _{1,2}^{1}+\mu _{1,3}^{1}+\mu _{1,4}^{1}+\mu _{1,5}^{1}\\&\quad +\,\mu _{2,3}^{2}+\mu _{2,4}^{2}+\mu _{2,5}^{2};\\ \mu _{2,5}&= 22\mu _{2,3}^{3}+\mu _{2,4}^{3}+\mu _{2,5}^{3}; \end{aligned}$$
$$\begin{aligned} \mu _{1,5}&= \mu _{1,3}^{2}+\mu _{1,4}^{2}+\mu _{1,5}^{2}+22\mu _{1,3}^{3}+\mu _{1,4}^{3}+\mu _{1,5}^{3};\\ \mu _{2,4}&= \mu _{2,3}^{3}+\mu _{2,4}^{3}+\mu _{2,5}^{3};\\ \mu _{1,4}&= 3\mu _{1,2}^{1}+\mu _{1,3}^{3}+\mu _{1,4}^{3}+\mu _{1,5}^{3};\\ \mu _{2,3}&= \mu _{2,3}^{3}+\mu _{2,4}^{3}+\mu _{2,5}^{3};\\ \mu _{1,3}&= 3\mu _{1,2}^{2}+\mu _{1,3}^{3}+\mu _{1,4}^{3}+\mu _{1,5}^{3};\\ \mu _{1,2}&= \mu _{1,2}^{1}+2\mu _{1,2}^{2}+2\mu _{1,3}^{2}+\mu _{1,5}^{2}+24\mu _{1,2}^{3}+\mu _{1,4}^{2}; \end{aligned}$$

Appendix 2: NHCTMM Chapman–Kolmogorov equations of the reduced ten states

$$\begin{aligned} dP_{10} (t)/dt&= -\alpha _{10}(t)P_{10} (t)+\mu _{9,10} P_9 (t)+\mu _{8,10} P_8 (t)\\&\quad +\,\mu _{7,10} P_7 (t)+\mu _{6,10} P_6 (t)+\mu _{5,10} P_5 (t)\\&\quad +\,\mu _{4,10} P_4 (t) +\,\mu _{3,10} P_3 (t)+\mu _{2,10} P_2 (t)\\&\quad +\,\mu _{1,10} P_1 (t)\\ dP_9 (t)/dt&= \lambda _{10,9}(t)P_{10} (t)-\alpha _9(t)P_9 (t)\\&\quad +\,\mu _{8,9} P_8 (t)+\,\mu _{7,9} P_7 (t)+\mu _{6,9} P_6 (t)\\&\quad +\,\mu _{5,9} P_5 (t) +\mu _{4,9} P_4 (t)+\,\mu _{3,9} P_3 (t)\\&\quad +\,\mu _{2,9} P_2 (t)+\mu _{1,9} P_1 (t) \\ dP_8 (t)/dt&= \lambda _{10,8}(t)P_{10} (t)+\lambda _{9,8}(t)P_9 (t)-\alpha _8(t)P_8 (t)\\&\quad +\,\mu _{7,8} P_7 (t)+\mu _{6,8} P_6 (t)+\mu _{5,8} P_5 (t) \\&\quad +\,\mu _{4,8} P_4 (t)+\,\mu _{3,8} P_3 (t)+\mu _{2,8} P_2 (t)\\&\quad +\,\mu _{1,8} P_1 (t) \\ dP_7 (t)/dt&= \lambda _{10,7}(t)P_{10} (t)+\lambda _{9,7}(t)P_9 (t)+\lambda _{8,7}(t)P_8 (t)\\&\quad -\,\alpha _7(t)P_7 (t)+\mu _{6,7} P_6 (t)\\&\quad +\,\mu _{5,7} P_5 (t)+\mu _{4,7} P_4 (t) \\&\quad +\,\mu _{3,7} P_3 (t)+\mu _{2,7} P_2 (t)+\mu _{1,7} P_1 (t) \\ dP_6 (t)/dt&= \lambda _{10,6}(t)P_{10} (t)+\lambda _{9,6}(t)P_9 (t)+\lambda _{8,6}(t)P_8 (t)\\&\quad +\,\lambda _{7,6}(t)P_7 (t)-\alpha _6(t)P_6 (t)+\mu _{5,6} P_5 (t) \\&\quad +\,\mu _{4,6} P_4 (t)+\mu _{3,6} P_3 (t)+\mu _{2,6} P_2 (t)\\&\quad +\,\mu _{1,6} P_1 (t) \\ dP_5 (t)/dt&= \lambda _{10,5}(t)P_{10} (t)+\lambda _{9,5}(t)P_9 (t)+\lambda _{8,5}(t)P_8 (t)\\&\quad +\,\lambda _{7,5}(t)P_7 (t)+\lambda _{6,5}(t)P_6 (t)-\alpha _5(t)P_5 (t) \\&\quad +\,\mu _{4,5} P_4 (t)+\mu _{3,5} P_3 (t)+\mu _{2,5} P_2 (t)\\&\quad +\,\mu _{1,5} P_1 (t) \\ dP_4 (t)/dt&= \lambda _{10,4}(t)P_{10} (t)+\lambda _{9,4}(t)P_9 (t)+\lambda _{8,4}(t)P_8 (t)\\&\quad +\,\lambda _{7,4}(t)P_7 (t)+\lambda _{6,4}(t)P_6 (t)+\lambda _{5,4}(t)P_5 (t) \\&\quad -\,\alpha _4(t)P_4 (t)+\mu _{3,4} P_3 (t)+\mu _{2,4} P_2 (t)\\&\quad +\,\mu _{1,4} P_1 (t) \\ dP_3 (t)/dt&= \lambda _{10,3}(t)P_{10} (t)+\lambda _{9,3}(t)P_9 (t)+\lambda _{8,3}(t)P_8 (t)\\&\quad +\,\lambda _{7,3}(t)P_7 (t)+\lambda _{6,3}(t)P_6 (t)+\lambda _{5,3}(t)P_5 (t) \\&\quad +\,\lambda _{4,3}(t)P_4 (t)-\alpha _3(t)P_3 (t)+\mu _{2,3} P_2 (t)\\&\quad +\,\mu _{1,3} P_1 (t) \\ dP_2 (t)/dt&= \lambda _{10,2}(t)P_{10} (t)+\lambda _{9,2}(t)P_9 (t)+\lambda _{8,2}(t)P_8 (t)\\&\quad +\,\lambda _{7,2}(t)P_7 (t)+\lambda _{6,2}(t)P_6 (t)+\lambda _{5,2}(t)P_5 (t) \\&\quad +\,\lambda _{4,2}(t)P_4 (t)+\lambda _{3,2}(t)P_3 (t)-\alpha _2(t)P_2 (t)\\&\quad +\,\mu _{1,2} P_1 (t) \\ dP_1 (t)/dt&= \lambda _{10,1}(t)P_{10} (t)+\lambda _{9,1}(t)P_9 (t)+\lambda _{8,1}(t)P_8 (t)\\&\quad +\,\lambda _{7,1}(t)P_7 (t)+\lambda _{6,1}(t)P_6 (t)+\lambda _{5,1}(t)P_5 (t) \\&\quad +\,\lambda _{4,1}(t)P_4 (t)+\lambda _{3,1}(t)P_3 (t)+\lambda _{2,1}(t)P_2 (t)\\&\quad -\,\alpha _1(t)P_1 (t) \end{aligned}$$

where:

$$\begin{aligned} \alpha _{10}(t)&= \lambda _{{2},{1}}^{1} (t)+\lambda _{{4},{3}}^{1} (t)+\lambda _{{3},{2}}^{2} (t)+\,15\lambda _{{5},{4}}^{3} (t)+\lambda _{{2},{1}}^{1} (t)\\&\quad +\,\lambda _{{5},{4}}^{1} (t)\!+\!\lambda _{{2},{1}}^{2} (t)\!+\!\lambda _{{2},{1}}^{1} (t)+\lambda {{3},{2}}^{2} (t)+\lambda _{{3},{1}}^{1} (t) \\&\quad +\,\lambda _{{4},{1}}^{1} (t)+\,\lambda _{{5},{1}}^{1} (t)+\lambda _{{5},{3}}^{1}(t)+\lambda _{{4},{2}}^{2} (t)+\lambda _{{5},{2}}^{2} (t)\\&\quad +\,\lambda _{{3},{1}}^{1} (t)+\lambda _{{5},{1}}^{1}(t)+\lambda _{{3},{1}}^{2} (t)+\lambda _{{4},{1}}^{2} (t)+\,\lambda _{{5},{1}}^{2} (t) \\&\quad +\,\lambda _{{3},{1}}^{1}(t)+\lambda _{{4},{1}}^{1} (t)+\,\lambda _{{5},{1}}^{1} (t)+\lambda _{{4},{2}}^{1} (t)+\lambda _{{5},{2}}^{1} (t)\\&\quad +\,\lambda _{{4},{2}}^{2} (t)+\lambda _{{5},{2}}^{2} (t)\lambda _{{5},{3}}^{1} (t)+\,\lambda _{{3},{1}}^{2} (t)\\&\quad +\,\lambda _{{4},1}^2 (t)+\lambda _{5,1}^{2} (t)+15\lambda _{5,{3}}^3 (t)\!+\!\lambda _{5,{2}}^1 (t)\!+\!\lambda _{3,1}^2 (t)\\&\quad +\,\lambda _{4,1}^{2} (t)+\lambda _{5,1}^2 (t)+\lambda _{4,1}^1 (t)+\,\lambda _{5,1}^{1} (t)\\&\quad +\,15\lambda _{5,{2}}^3 (t)+\lambda _{5,1}^{1} (t)+15\lambda _{5,1}^3 (t) \\ \alpha _9(t)&= \mu _{1,2}^{1} +\mu _{1,3}^{1} +\mu _{1,4}^{1} +\mu _{1,5}^{1} +\mu _{3,4}^{1} +\mu _{{2},5}^2 \\&\quad +\,2\lambda _{5,4}^3 (t)+\lambda _{5,4}^2 (t)+\,\mu _{3,5}^{1} +\mu _{{2},4}^2 \\&\quad +\,\lambda _{3,2}^{1} (t)+\lambda _{{5},3}^2 (t)+\lambda _{2,1}^2 (t)+2\lambda _{5,3}^3 (t) \\&\quad +\,\lambda _{3,1}^1 (t)+\lambda _{5,2}^2 (t)+{2}\lambda _{{5},2}^3 (t)+2\lambda _{5,{1}}^3 (t)\!+\!\lambda _{5,{1}}^{2} (t) \\ \alpha _8(t)&= 15\mu _{4,5}^3 +2\mu _{4,5}^3 +\lambda _{2,1}^1 (t)+\lambda _{{5},{4}}^1 (t)+\lambda _{{2},{1}}^2 (t)\\&\quad +\,\lambda _{3,1}^{1} (t)+\lambda _{5,{1}}^1 (t)+\lambda _{3,{1}}^2 (t)+\lambda _{4,1}^{2} (t)+\lambda _{5,{1}}^2 (t)\\&\quad +\,\lambda _{5,4}^2 (t)+\,\lambda _{3,{1}}^{1} (t)+\lambda _{4,1}^{1} (t)+\lambda _{5,1}^1 (t)+\lambda _{4,{2}}^1 (t)\\&\quad +\,\lambda _{5,2}^{1} (t)+\lambda _{4,2}^2 (t)+\lambda _{5,2}^{2} (t)+17\lambda _{{4},3}^3 (t)\!+\!\lambda _{{5},3}^1 (t)\\&\quad +\,\lambda _{3,1}^2 (t)+\lambda _{{3},{1}}^2 (t) +\lambda _{{4},{1}}^2 (t)+\,\lambda _{{5},{1}}^2 (t)+\lambda _{5,{2}}^{1} (t)\\&\quad +\,\lambda _{{4},1}^{2} (t)+\lambda _{{5},1}^{2} (t)+\lambda _{{5},2}^2 (t)+\,\lambda _{{3}{1}}^1 (t)+\lambda _{{4},1}^1 (t)\\&\quad +\,\lambda _{5,1}^1 (t)\!+\!17\lambda _{4,2}^3 (t) \!+\!\lambda _{5,1}^1 (t) \!+\!17\lambda _{{4},1}^{3} (t)\!+\!\lambda _{{5},1}^{2} (t) \\ \alpha _7(t)&= \mu _{{1},{2}}^{1} +\mu _{{1},{3}}^{1} +\mu _{{1},{4}}^1 +\mu _{{1},{5}}^1 +\mu _{{4},{5}}^{1} \\&\quad +\,\mu _{{1},{2}}^{2} +\mu _{{1},{3}}^{2} +\mu _{{1},{4}}^2 +\mu _{{1},{5}}^{2} +\mu _{{4},{5}}^2 +\mu _{{1},{2}}^{1} \\&\quad +\,\mu _{{1},{3}}^{1} +\mu _{{1},{4}}^{1} +\mu _{{1},{5}}^1 +\mu _{4,{5}}^1 +\mu _{{1},2}^{2} \\&\quad +\,\mu _{{1},{3}}^2 +\mu _{{1},{4}}^{2} +\mu _{{1},{5}}^{2} +\mu _{{4},{5}}^{2} +{2}\lambda _{{4},{3}}^2 (t)\\&\quad +\,2\lambda _{{4},{3}}^{1} (t)+{2}\lambda _{{4},{3}}^{3} (t)+{2}\lambda _{5,{3}}^{3} (t)\\&\quad +\,{2}\lambda _{{4},{2}}^{1} (t)+2\lambda _{{4},{2}}^{3} (t)\\&\quad +\,2\lambda _{{5},{2}}^{3} (t)+{2}\lambda _{{4},2}^2 (t)+{2}\lambda _{{4},{1}}^1 (t)\\&\quad +\,2\lambda _{{4},1}^{3} (t)+{2}\lambda _{5,1}^{3} (t)+{2}\lambda _{4,{1}}^{1} (t) \\ \alpha _6(t)&= \mu _{{1},{2}}^{1} +\mu _{{1},{3}}^{1} +\mu _{{1},{4}}^1 +\mu _{{1},{5}}^1 \\&\quad +\,\mu _{{2},{4}}^{1} +\mu _{{2},{5}}^{1} +\mu _{{2},{4}}^{2} +\mu _{{2},{5}}^2 +\mu _{{2},{3}}^{1} \\&\quad +\,\mu _{{3},{5}}^2 +\mu _{{1},{2}}^{1} +\mu _{{1},{3}}^{1} +\mu _{{1},{4}}^{1} +\mu _{{1},{5}}^1 +\,\mu _{{2},{3}}^1 \\&\quad +\,\mu _{{2},{4}}^{1}+\,\mu _{{2},{5}}^{1} +\mu _{{2},{4}}^{2} +\mu _{{2},{5}}^{2} +\mu _{{3},{5}}^{2} +{2}\mu _{{3},{4}}^2\\&\quad +\,2\lambda _{{4},{3}}^{3} (t)+{2}\lambda _{{5},{3}}^{3} (t)+{2}\lambda _{{2},{1}}^{2} (t)+{2}\lambda _{{2},{1}}^{1} (t)\\&\quad +\,2\lambda _{{3},{2}}^{2} (t)+{2}\lambda _{{4},{2}}^{3} (t) \\&\quad +\,{2}\lambda _{{5},2}^{3} (t)+{2}\lambda _{{3},{1}}^{2} (t)+2\lambda _{{4},1}^{3} (t)+{2}\lambda _{5,1}^{3} (t) \end{aligned}$$
$$\begin{aligned} \alpha _5(t)&= \mu _{3,5}^{1} +\mu _{{1},{3}}^2 +\mu _{{1},{4}}^2 +\mu _{{1},{5}}^2 +15\mu _{3,5}^3 \\&\quad +\,2\mu _{3,{5}}^3 +\mu _{3,5}^1 +\mu _{1,3}^2 +\mu _{1,4}^2 +\mu _{1,{5}}^2 +17\mu _{3,4}^3 \\&\quad +\,2\mu _{3,4}^{1} +2\mu _{3,{4}}^3 \\&\quad +\,2\mu _{3,{5}}^3 +2\mu _{3,4}^3 +2\mu _{3,5}^3 +3\lambda _{3,2}^1 (t)+\lambda _{{2},{1}}^{2} (t)\\&\quad +\,\lambda _{3,{1}}^2 (t)+\lambda _{4,1}^{2} (t)+\lambda _{5,1}^2 (t)+\lambda _{4,2}^1 (t)\\&\quad +\,\lambda _{5,2}^1 (t)+\lambda _{2,1}^1 (t)\\&\quad +\,\lambda _{3,2}^2 (t)+\lambda _{3,1}^1 (t)+\lambda _{4,1}^1 (t)+\lambda _{5,{1}}^1 (t)+\lambda _{4,2}^2 (t)\\&\quad +\,\lambda _{5,{2}}^{2} (t)+22\lambda _{3,{2}}^{3} (t)+\lambda _{4,2}^{3} (t)+\lambda _{5,2}^3 (t)\\&\quad +\,\lambda _{{4},2}^1 (t)+\lambda _{5,{2}}^{1} (t)\\&\quad +\,3\lambda _{{3},1}^{1} (t)+\lambda _{{4},{1}}^{1} (t)+\lambda _{{5},{1}}^{1} (t)\\&\quad +\,22\lambda _{{3},{1}}^{3} (t)+\lambda _{{4,1}}^{3} (t)+\lambda _{{5},1}^{3} (t)+\lambda _{{3},{1}}^{2} (t)\\&\quad +\,\lambda _{{4},{1}}^{2} (t)+\lambda _{{5},{1}}^{2} (t) \end{aligned}$$
$$\begin{aligned} \alpha _4(t)&= \mu _{{2},5}^{1} +\mu _{{1},{3}}^2 +\mu _{{1},{4}}^2 +\mu _{{1},{5}}^2\\&\quad +\,\mu _{{2},5}^{1} +\mu _{{1},{3}}^{2} +\mu _{1,4}^2 +\mu _{1,{5}}^2 +{2}\mu _{{2},4}^{1}\\&\quad +\,2\mu _{{1},{2}}^{2} +{3}\mu _{{2},{3}}^{1} +\mu _{{2},{4}}^{1} +\mu _{{2},{5}}^{1}\\&\quad +\,\mu _{{1},{2}}^{2}+\mu _{{1},{3}}^{2} \\&\quad +\,\mu _{{1},{4}}^{2} +\mu _{1,{5}}^2 +\lambda _{3,2}^{3} (t)+\lambda _{{4},{2}}^{3} (t)\\&\quad +\,\lambda _{{5},{2}}^{3} (t)+2\lambda _{{2},1}^{1} (t)+\lambda _{{3},1}^{3} (t)+\lambda _{4,{1}}^{3} (t)\\&\quad +\,\lambda _{5,{1}}^{3} (t) \\ \alpha _3(t)&= \mu _{{1},{4}}^{1} +\mu _{{1},{5}}^{1} +\mu _{{1},{3}}^{1} +\mu _{{2},{5}}^2 +\mu _{{1},{3}}^{1}\\&\quad +\,\mu _{{1},{4}}^{1} +\mu _{1,{5}}^{1} +\mu _{{2},{5}}^2 +{2}\mu _{{2},4}^{2} +2\mu _{{1},{2}}^{1}\\&\quad +\,{2}\mu _{{2},{3}}^{2} +\mu _{{1},{2}}^{1} +\mu _{{1},{3}}^{1} +\mu _{{1},{4}}^{1} +\mu _{{1},{5}}^{1} \\&\quad +\,\mu _{{2},{3}}^{2} +\mu _{{2},{4}}^2 +\mu _{{2},{5}}^2 +\lambda _{3,2}^{3} (t)+\lambda _{{4},{2}}^{3} (t)\\&\quad +\,\lambda _{{5},{2}}^{3} (t){+3}\lambda _{{2},1}^{2} (t)+\lambda _{{3},1}^{3} (t)+\lambda _{4,{1}}^{3} (t)+\lambda _{5,{1}}^{3} (t) \\ \alpha _2(t)&= 15\mu _{{2},{5}}^{3} +2\mu _{{2},{5}}^{3} +17\mu _{{2},{4}}^{3} +{2}\mu _{{2},{4}}^{3} +{2}\mu _{{2},{5}}^{3} \!+\!2\mu _{{2},{4}}^{3}\\&\quad +\,{2}\mu _{{2},{5}}^{3} {+22}\mu _{{2},{3}}^{3} +\mu _{{2},4}^{3} +\mu _{{2},{5}}^{3} +\mu _{{2},{3}}^{3} +\mu _{{2},{4}}^{3} \\&\quad +\,\mu _{{2},{5}}^{3} +\mu _{{2},{3}}^{3} +\mu _{{2},{4}}^{3} +\mu _{{2},{5}}^{3} +2\lambda _{{2},{1}}^{1} (t)+\lambda _{{2},{1}}^{2} (t)\\&\quad +\,{24}\lambda _{{2},{1}}^{3} (t)+\lambda _{{3},1}^{1} (t)+\lambda _{{4},1}^{1} (t)+\lambda _{{5},{1}}^{1} (t)+\lambda _{{3},{1}}^{2} (t)\\&\quad +\,\lambda _{{4},1}^{2} (t)+\,\lambda _{{5},{1}}^{2} (t) \\ \alpha _1(t)&= 15\mu _{{1},{5}}^{3} +\mu _{{1},{5}}^{2} +{2}\mu _{{1},{5}}^{3} +\mu _{{1},{5}}^{2} \\&\quad +\,17\mu _{{1},{4}}^{3} +\mu _{{1},{4}}^{1} +{2}\mu _{{1},{4}}^{2} +2\mu _{{1},{4}}^{3} +{2}\mu _{{1},{5}}^{3} +{2}\mu _{{1},{3}}^{2}\\&\quad +\,{2}\mu _{{1},{4}}^{3} +2\mu _{{1},{5}}^{3} +\mu _{{1},{3}}^{2} \\&\quad +\,\mu _{{1},{4}}^{2} +\mu _{{1},{5}}^{2} +22\mu _{{1},{3}}^{3} +\mu _{{1},{4}}^{3} \\&\quad +\,\mu _{{1},{5}}^{3} +{3}\mu _{{1},{2}}^{1} +\mu _{{1},{3}}^{3} +\mu _{{1},{4}}^{3} +\mu _{{1},{5}}^{3} +{3}\mu _{{1},{2}}^{2} \\&\quad +\,\mu _{{1},{3}}^{3}+\mu _{{1},4}^{3} +\mu _{{1},{5}}^{3} +\mu _{{1},{2}}^{1} \\&\quad +\,\mu _{{1},{2}}^{2} +\,\mu _{{1},{3}}^{2} +\mu _{{1},{5}}^{2} {+24}\mu _{{1},{2}}^{3} +\mu _{{1},{4}}^{2} \\ \end{aligned}$$

Appendix 3: NHCTMM Chapman–Kolmogorov equations of each component

$$\begin{aligned}&Component 1\\&\quad \times \left\{ {\begin{array}{lll} dP_5 (t)/dt&{}=&{}-(\lambda _{5,4}^1 (t)+\lambda _{5,3}^1 (t)+\lambda _{5,2}^1 (t)+\lambda _{5,1}^1 (t))P_5 (t)+\mu _{4,5}^1 P_4 (t)+\,\mu _{3,5}^1 P_3 (t) +\mu _{2,5}^1 P_2 (t)+\mu _{1,5}^1 P_1 (t) \\ dP_4 (t)/dt&{}=&{}\lambda _{5,4}^{1} (t)P_5 (t)-(\mu _{{4},{5}}^1 +\lambda _{{4},3}^1 (t) +\lambda _{{4},2}^1 (t)+\lambda _{{4},1}^1 (t))P_4 (t)+\,\mu _{{3},{4}}^{1} P_3 (t)+\mu _{2,4}^{1} P_2 (t)+\mu _{1,4}^{1}P_1 (t) \\ dP_3 (t)/dt&{}=&{}\lambda _{5,3}^{1} (t)P_5 (t)+\lambda _{4,3}^{1} (t)P_4 (t)-\,(\mu _{{3},{5}}^1+\mu _{{3},{4}}^1+\lambda _{{3},2}^1(t) +\lambda _{{3},1}^1 (t))P_3 (t)+\mu _{2,3}^{1} P_2 (t)+\mu _{1,3}^{1} P_1 (t) \\ dP_2 (t)/dt&{}=&{}\lambda _{5,2}^{1} (t)P_5 (t)+\lambda _{4,2}^{1} (t)P_4 (t) +\lambda _{3,2}^{1} (t)P_3 (t)-\,(\mu _{{2},{5}}^1+\mu _{{2},{4}}^1 +\mu _{{2},{3}}^1 +\lambda _{{2},1}^1 (t))P_2 (t) +\mu _{1,2}^{1} P_1 (t) \\ dP_1 (t)/dt&{}=&{}\lambda _{5,1}^{1} (t)P_5 (t)+\lambda _{4,1}^{1} (t)P_4 (t) +\lambda _{3,1}^{1} (t)P_3 (t)+\lambda _{2,1}^{1} (t)P_2 (t) -\,(\mu _{{1},{5}}^1+\mu _{{1},{4}}^1 +\mu _{{1},{3}}^1 +\mu _{{1},{2}}^1)P_1 (t) \\ \end{array}} \right. \end{aligned}$$
$$\begin{aligned}&Component 2\\&\quad \times \left\{ {\begin{array}{lll} dP_5 (t)/dt&{}=&{}-(\lambda _{5,4}^{2} (t)+\lambda _{5,3}^{2} (t) +\lambda _{5,2}^{2} (t)+\lambda _{5,1}^{2} (t))P_5 (t) +\,\mu _{4,5}^{2} P_4(t)+\mu _{3,5}^{2} P_3 (t) +\mu _{2,5}^{2} P_2 (t)+\mu _{1,5}^{2} P_1(t) \\ dP_4 (t)/dt&{}=&{}\lambda _{5,4}^{2} (t)P_5 (t)-(\mu _{{4},{5}}^{2} +\lambda _{{4},3}^{2} (t)+\lambda _{{4},2}^{2} (t) +\lambda _{{4},1}^{2} (t))P_4 (t) +\,\mu _{{3},{4}}^{2} P_3 (t) +\mu _{2,4}^{2}P_2 (t)+\mu _{1,4}^{2} P_1 (t) \\ dP_3 (t)/dt&{}=&{}\lambda _{5,3}^{2} (t)P_5 (t)+\lambda _{4,3}^{2} (t)P_4 (t) -\,(\mu _{{3},{5}}^{2}+\mu _{{3},{4}}^{2} +\lambda _{{3},2}^{2}(t)+\lambda _{{3},1}^{2} (t))P_3 (t) +\mu _{2,3}^{2} P_2 (t)+\mu _{1,3}^{2} P_1 (t) \\ dP_2 (t)/dt&{}=&{}\lambda _{5,2}^{2} (t)P_5 (t)+\lambda _{4,2}^2 (t)P_4(t) +\lambda _{3,2}^2 (t)P_3 (t) -\,(\mu _{{2},{5}}^2 +\mu _{{2},{4}}^2+\mu _{{2},{3}}^2 +\lambda _{{2},1}^2 (t))P_2 (t)+\mu _{1,2}^2 P_1(t) \\ dP_1 (t)/dt&{}=&{}\lambda _{5,1}^2 (t)P_5 (t)+\lambda _{4,1}^2 (t)P_4(t) +\lambda _{3,1}^2 (t)P_3 (t)+\lambda _{2,1}^2 (t)P_2 (t) -\,(\mu _{{1},{5}}^2+\mu _{{1},{4}}^2 +\mu _{{1},{3}}^2 +\mu _{{1},{2}}^2)P_1 (t) \end{array}} \right. \end{aligned}$$
$$\begin{aligned}&Component 3\\&\quad \times \left\{ {\begin{array}{lll} dP_5 (t)/dt&{}=&{}-(\lambda _{5,4}^3 (t)+\lambda _{5,3}^3 (t) +\lambda _{5,2}^3 (t)+\lambda _{5,1}^3 (t))P_5 (t) +\,\mu _{4,5}^3 P_4 (t)+\mu _{3,5}^3 P_3 (t) +\mu _{2,5}^3 P_2 (t)+\mu _{1,5}^3 P_1 (t) \\ dP_4 (t)/dt&{}=&{}\lambda _{5,4}^3 (t)P_5 (t)-(\mu _{{4},{5}}^3 +\lambda _{{4},3}^3 (t)+\lambda _{{4},2}^3 (t) +\lambda _{{4},1}^3 (t))P_4(t)+\,\mu _{{3},{4}}^3 P_3 (t) +\mu _{2,4}^3 P_2 (t)+\mu _{1,4}^3 P_1(t)\\ dP_3 (t)/dt&{}=&{}\lambda _{5,3}^3 (t)P_5 (t)+\lambda _{4,3}^3 (t)P_4 (t) -\,(\mu _{{3},{5}}^3+\mu _{{3},{4}}^3 +\lambda _{{3},2}^3(t)+\lambda _{{3},1}^3 (t))P_3 (t) +\mu _{2,3}^3 P_2 (t)+\mu _{1,3}^3P_1 (t) \\ dP_2 (t)/dt&{}=&{}\lambda _{5,2}^3 (t)P_5 (t)+\lambda _{4,2}^3 (t)P_4(t) +\lambda _{3,2}^3 (t)P_3 (t)-\,(\mu _{{2},{5}}^{3} +\mu _{{2},{4}}^{3} +\mu _{{2},{3}}^{3} +\lambda _{{2},1}^{3} (t))P_2(t)+\mu _{1,2}^{3} P_1 (t) \\ dP_1 (t)/dt&{}=&{}\lambda _{5,1}^{3} (t)P_5 (t)+\lambda _{4,1}^{3} (t)P_4(t) +\lambda _{3,1}^{3} (t)P_3 (t)+\lambda _{2,1}^{3} (t)P_2 (t) -\,(\mu _{{1},{5}}^{3}+\mu _{{1},{4}}^{3} +\mu _{{1},{3}}^{3} +\mu _{{1},{2}}^{3} )P_1 (t) \\ \end{array}} \right. \end{aligned}$$

Appendix 4: NHCTMRM Chapman–Kolmogorov equations of each component

$$\begin{aligned}&Component 1\\&\quad \times \left\{ {\begin{array}{lll} dV_5 (t)/dt&{}=&{}-(\lambda _{5,4}^1 (t)+\lambda _{5,3}^1 (t) +\lambda _{5,2}^1 (t)+\lambda _{5,1}^1 (t))V_5 (t) +\lambda _{5,4}^1 (t)V_4(t)+\lambda _{5,3}^1 (t)V_3 (t) +\,\lambda _{5,2}^1 (t)V_2 (t)\\ &{}&{} +\,\,\lambda _{5,1}^1 (t)V_1 (t)\\ dV_4 (t)/dt&{}=&{}54\mu _{4,5}^{1} +\mu _{4,5}^{1} V_5 (t) -(\mu _{{4},{5}}^1 +\lambda _{{4},3}^1 (t) +\lambda _{{4},2}^1 (t)+\lambda _{{4},1}^1 (t))V_4 (t) +\lambda _{4,3}^{1} (t)V_3 (t) +\,\lambda _{4,2}^{1} (t)V_2 (t)\\ &{}&{}+\,\lambda _{4,1}^{1} (t)V_1 (t)\\ dV_3 (t)/dt&{}=&{}72\mu _{3,5}^{1} +54\mu _{3,4}^{1} +\mu _{3,5}^{1} V_5(t)+\mu _{3,4}^{1} V_4 (t) -(\mu _{{3},{5}}^1 +\mu _{{3},{4}}^1 +\lambda _{{3},2}^1 (t)+\lambda _{{3},1}^1 (t))V_3 (t) +\,\lambda _{3,2}^{1} (t)V_2 (t)\\ &{}&{} +\,\,\lambda _{3,1}^{1} (t)V_1 (t)\\ dV_2 (t)/dt&{}=&{}90\mu _{2,5}^{1} +72\mu _{2,4}^{1} +54\mu _{2,3}^{1}+\mu _{2,5}^{1} V_5 (t) +\mu _{2,4}^{1} V_4 (t)+\mu _{2,3}^{1} V_3(t) -\,(\mu _{{2},{5}}^1 +\mu _{{2},{4}}^1 +\mu _{{2},{3}}^1\\ &{}&{} +\,\,\lambda _{{2},1}^1 (t))V_2 (t) +\lambda _{2,1}^{1}(t)V_1 (t)\\ dV_1 (t)/dt&{}=&{}180\mu _{1,5}^{1} +90\mu _{1,4}^{1} +72\mu _{1,3}^{1}+54\mu _{1,2}^{1} +\mu _{1,5}^{1} V_5 (t)+\mu _{1,4}^{1} V_4 (t) +\mu _{1,3}^{1} V_3 (t) +\,\mu _{1,2}^{1} V_2 (t)\\ &{}&{} -\,\,(\mu _{{1},{5}}^1 +\mu _{{1},{4}}^1 +\mu _{{1},{3}}^1 +\mu _{{1},{2}}^1 )V_1 (t) \end{array}} \right. \end{aligned}$$
$$\begin{aligned}&Component 2\\&\quad \times \left\{ {\begin{array}{lll} dV_5 (t)/dt&{}=&{}-(\lambda _{5,4}^{2} (t)+\lambda _{5,3}^{2}(t) +\lambda _{5,2}^{2} (t)+\lambda _{5,1}^{2} (t))V_5 (t) +\lambda _{5,4}^{2} (t)V_4 (t)+\lambda _{5,3}^{2} (t)V_3 (t) +\,\lambda _{5,2}^{2} (t)V_2 (t)\\ &{}&{}+\,\,\lambda _{5,1}^{2} (t)V_1 (t)\\ dV_4 (t)/dt&{}=&{}72\mu _{4,5}^{2} +\mu _{4,5}^{2} V_5 (t) -(\mu _{{4},{5}}^{2} +\lambda _{{4},3}^{2} (t) +\lambda _{{4},2}^{2}(t)+\lambda _{{4},1}^{2} (t))V_4 (t) +\lambda _{4,3}^{2} (t)V_3 (t)+\,\lambda _{4,2}^{2} (t)V_2 (t)\\ &{}&{}+\,\,\lambda _{4,1}^{2} (t)V_1(t)\\ dV_3 (t)/dt&{}=&{}96\mu _{3,5}^{2} +72\mu _{3,4}^{2} +\mu _{3,5}^{2} V_5(t)+\mu _{3,4}^{2} V_4 (t) -(\mu _{{3},{5}}^{2} +\mu _{{3},{4}}^{2} +\lambda _{{3},2}^{2} (t)+\lambda _{{3},1}^{2} (t))V_3 (t) +\,\lambda _{3,2}^{2} (t)V_2 (t)\\ &{}&{}+\,\,\lambda _{3,1}^{2} (t)V_1 (t)\\ dV_2 (t)/dt&{}=&{}120\mu _{2,5}^{2} +96\mu _{2,4}^{2} +72\mu _{2,3}^{2}+\mu _{2,5}^{2} V_5 (t) +\mu _{2,4}^{2} V_4 (t)+\mu _{2,3}^{2} V_3(t) -\,(\mu _{{2},{5}}^{2} +\mu _{{2},{4}}^{2} +\mu _{{2},{3}}^{2}\\ &{}&{} +\,\,\lambda _{{2},1}^{2} (t))V_2 (t) +\lambda _{2,1}^{2} V_1 (t)\\ dV_1 (t)/dt&{}=&{}240\mu _{1,5}^{2} +120\mu _{1,4}^{2} +96\mu _{1,3}^{2}+72\mu _{1,2}^{2} +\mu _{1,5}^{2} V_5 (t)+\mu _{1,4}^{2} V_4 (t) +\mu _{1,3}^{2} V_3 (t) +\,\mu _{1,2}^{2} V_2 (t)\\ &{}&{}-\,\,(\mu _{{1},{5}}^{2} +\mu _{{1},{4}}^{2} +\mu _{{1},{3}}^{2} +\mu _{{1},{2}}^{2} )V_1 (t) \end{array}} \right. \end{aligned}$$
$$\begin{aligned}&Component 3\\&\quad \times \left\{ {\begin{array}{lll} dV_5 (t)/dt&{}=&{}-(\lambda _{5,4}^{3} (t)+\lambda _{5,3}^{3}(t) +\lambda _{5,2}^{3} (t)+\lambda _{5,1}^{3} (t))V_5 (t) +\lambda _{5,4}^{3} (t)V_4 (t)+\lambda _{5,3}^{3} (t)V_3 (t) +\,\lambda _{5,2}^{3} (t)V_2 (t)\\ &{}&{}+\,\,\lambda _{5,1}^{3} (t)V_1 (t)\\ dV_4 (t)/dt&{}=&{}120\mu _{4,5}^{3} +\mu _{4,5}^{3} V_5 (t) -(\mu _{{4},{5}}^{3} +\lambda _{{4},3}^{3} (t) +\lambda _{{4},2}^{3}(t)+\lambda _{{4},1}^{3} (t))V_4 (t) +\lambda _{4,3}^{3} (t)V_3 (t)+\,\lambda _{4,2}^{3} (t)V_2 (t)\\ &{}&{}+\,\,\lambda _{4,1}^{3} (t)V_1(t)\\ dV_3 (t)/dt&{}=&{}160\mu _{3,5}^{3} +120\mu _{3,4}^{3} +\mu _{3,5}^{3}V_5 (t)+\mu _{3,4}^{3} V_4 (t) -(\mu _{{3},{5}}^{3} +\mu _{{3},{4}}^{3} +\lambda _{{3},2}^{3} (t)+\lambda _{{3},1}^{3}(t))V_3 (t) +\,\lambda _{3,2}^{3} V_2 (t)\\ &{}&{}+\,\,\lambda _{3,1}^{3} (t)V_1 (t)\\ dV_2 (t)/dt&{}=&{}200\mu _{2,5}^{3} +160\mu _{2,4}^{3} +120\mu _{2,3}^{3}+\mu _{2,5}^{3} V_5 (t) +\mu _{2,4}^{3} V_4 (t)+\mu _{2,3}^{3} V_3(t) -\,(\mu _{{2},{5}}^{3} +\mu _{{2},{4}}^{3} +\mu _{{2},{3}}^{3}\\ &{}&{} +\,\,\lambda _{{2},1}^{3} (t))V_2 (t) +\lambda _{2,1}^{3} V_1 (t)\\ dV_1 (t)/dt&{}=&{}240\mu _{1,5}^{3} +200\mu _{1,4}^{3} +160\mu _{1,3}^{3} +120\mu _{1,2}^{3} +\mu _{1,5}^{3} V_5 (t)+\mu _{1,4}^{3}V_4 (t) +\mu _{1,3}^{3} V_3 (t) +\,\mu _{1,2}^{3} V_2 (t)\\ &{}&{}-\,\,(\mu _{{1},{5}}^{3} +\mu _{{1},{4}}^{3} +\mu _{{1},{3}}^{3} +\mu _{{1},{2}}^{3} )V_1 (t) \end{array}} \right. \end{aligned}$$

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Huang, CH., Wang, CH. Optimization of preventive maintenance for a multi-state degraded system by monitoring component performance. J Intell Manuf 27, 1151–1170 (2016). https://doi.org/10.1007/s10845-014-0940-5

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10845-014-0940-5

Keywords

Navigation