# Parallel Solution of Robust Nonlinear Model Predictive Control Problems in Batch Crystallization

^{1}

^{2}

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Problem Formulations

#### 2.1. NMPC Formulation

#### 2.2. MHE Formulation

#### 2.3. Robust NMPC Formulation

#### 2.4. Efficient Optimization via the Simultaneous Approach

## 3. Efficient Parallel Schur Complement Method for Stochastic Programs

## 4. Performance of Robust NMPC on Batch Crystalization

#### 4.1. Case Study: Multidimensional Unseeded Batch Crystallization Model

#### 4.2. Numerical Results

## 5. Conclusions

## Acknowledgments

## Author Contributions

## Conflicts of Interest

## References

- Jiang, Z.P.; Wang, Y. Input-To-State stability for discrete-time nonlinear systems. Automatica
**2001**, 37, 857–869. [Google Scholar] [CrossRef] - Magni, L.; Scattolini, R. Robustness and robust design of MPC for nonlinear discrete-time systems. In Assessment and Future Directions of Nonlinear Model Predictive Control; Springer: Berlin Heidelberg, Germmany, 2007; pp. 239–254. [Google Scholar]
- Scokaert, P.; Mayne, D. Min-Max feedback model predictive control for constrained linear systems. IEEE Trans. Autom. Control
**1998**, 43, 1136–1142. [Google Scholar] [CrossRef] - Huang, R.; Patwardhan, S.C.; Biegler, L.T. Multi-Scenario-Based robust nonlinear model predictive control with first principle models. Comput. Aided Chem. Eng.
**2009**, 27, 1293–1298. [Google Scholar] - Nagy, Z.K.; Braatz, R.D. Robust nonlinear model predictive control of batch processes. AIChE J.
**2003**, 49, 1776–1786. [Google Scholar] [CrossRef] - Magni, L.; De Nicolao, G.; Scattolini, R.; Allgöwer, F. Robust model predictive control for nonlinear discrete-time systems. Int. J. Robust Nonlinear Control
**2003**, 13, 229–246. [Google Scholar] [CrossRef] - Lucia, S.; Finkler, T.; Engell, S. Multi-Stage nonlinear model predictive control applied to a semi-batch polymerization reactor under uncertainty. J. Process Control
**2013**, 23, 1306–1319. [Google Scholar] [CrossRef] - Telen, D.; Houska, B.; Logist, F.; Van Derlinden, E.; Diehl, M.; Van Impe, J. Optimal experiment design under process noise using Riccati differential equations. J. Process Control
**2013**, 23, 613–629. [Google Scholar] [CrossRef] [Green Version] - Streif, S.; Kögel, M.; Bäthge, T.; Findeisen, R. Robust Nonlinear Model Predictive Control with Constraint Satisfaction: A relaxation-based Approach. In Proceedings of the 19th IFAC World Congress, Cape Town, South Africa, 24–29 August 2014; pp. 11073–11079.
- Zavala, V.M.; Laird, C.D.; Biegler, L.T. Interior-Point decomposition approaches for parallel solution of large-scale nonlinear parameter estimation problems. Chem. Eng. Sci.
**2008**, 63, 4834–4845. [Google Scholar] [CrossRef] - Kang, J.; Cao, Y.; Word, D.P.; Laird, C. An interior-point method for efficient solution of block-structured NLP problems using an implicit Schur-complement decomposition. Comput. Chem. Eng.
**2014**, 71, 563–573. [Google Scholar] [CrossRef] - Lubin, M.; Petra, C.; Anitescu, M. The parallel solution of dense saddle-point linear systems arising in stochastic programming. Optim. Methods Softw.
**2012**, 27, 845–864. [Google Scholar] [CrossRef] - Cao, Y.; Laird, C.; Zavala, V. Clustering-Based Preconditioning for Stochastic Programs. Comput. Optim. Appl.
**2015**, 64, 379–406. [Google Scholar] [CrossRef] - Gay, D.M.; Kernighan, B. AMPL: A Modeling Language for Mathematical Programming, 2nd ed.; Cengage Learning: Boston, MA, USA, 2002; Volume 2. [Google Scholar]
- Watson, J.P.; Woodruff, D.L.; Hart, W.E. PySP: Modeling and solving stochastic programs in Python. Math. Program. Comput.
**2012**, 4, 109–149. [Google Scholar] [CrossRef] - Huchette, J.; Lubin, M.; Petra, C. Parallel algebraic modeling for stochastic optimization. In Proceedings of the 1st First Workshop for High Performance Technical Computing in Dynamic Languages, New Orleans, Louisiana, 16–21 November 2014; pp. 29–35.
- Shapiro, A.; Dentcheva, D.; Ruszczynski, A. Lectures on Stochastic Programming: Modeling and Theory; SIAM-Society for Industrial and Applied Mathematics: Philadelphia, PA, USA, 2014; Volume 16. [Google Scholar]
- Cuthrell, J.E.; Biegler, L.T. On the optimization of differential-algebraic process systems. AIChE J.
**1987**, 33, 1257–1270. [Google Scholar] [CrossRef] - Forsgren, A.; Gill, P.E.; Wright, M.H. Interior methods for nonlinear optimization. SIAM Rev.
**2002**, 44, 525–597. [Google Scholar] [CrossRef] - Mesbah, A.; Nagy, Z.; Huesman, A.; Kramer, H.; Van den Hof, P. Real-time control of industrial batch crystallization processes using a population balance modeling framework. IEEE Trans. Control Syst. Technol.
**2012**, 20, 1188–1201. [Google Scholar] [CrossRef] - Acevedo, D.; Nagy, Z.K. Systematic classification of unseeded batch crystallization systems for achievable shape and size analysis. J. Cryst. Growth
**2014**, 394, 97–105. [Google Scholar] [CrossRef] - Cao, Y.; Acevedo, D.; Nagy, Z.K.; Laird, C.D.; School of Chemical Engineering, Purdue University, West Lafayette, IN, USA. Unpublished work. 2015.
- Gunawan, R.; Ma, D.L.; Fujiwara, M.; Braatz, R.D. Identification of kinetic parameters in multidimensional crystallization processes. Int. J. Modern Phys. B
**2002**, 16, 367–374. [Google Scholar] [CrossRef] - Majumder, A.; Nagy, Z.K. Prediction and control of crystal shape distribution in the presence of crystal growth modifiers. Chem. Eng. Sci.
**2013**, 101, 593–602. [Google Scholar] [CrossRef]

**Figure 1.**Optimal temperature profile for nominal NMPC (nonlinear model predictive control) and robust NMPC.

**Table 1.**The robust performance (value of $cost$) of different control strategies evaluated using 100 test scenarios and exact information.

Control Strategies | Nominal | Average | Standard Deviation | Worst-Case |
---|---|---|---|---|

Ideal | $2\times {10}^{-4}$ | 30 | 66 | 499 |

Open-loop | $2\times {10}^{-4}$ | 167 | 223 | 1339 |

Nominal NMPC | 0.2 | 93 | 159 | 955 |

Exact Min–max NMPC | 32 | 78 | 113 | 677 |

Exact Min–expected NMPC | 12 | 99 | 169 | 1076 |

**Table 2.**The robust performance of the robust NMPC using different numbers of scenarios evaluated using 100 test scenarios.

Type | S | Nominal | Average | Standard Deviation | Worst-Case |
---|---|---|---|---|---|

Min–max | 25 | 15 | 99 | 170 | 1062 |

50 | 13 | 102 | 178 | 1129 | |

75 | 13 | 95 | 156 | 946 | |

100 | 25 | 80 | 120 | 767 | |

Min–expected | 25 | 21 | 89 | 138 | 902 |

50 | 11 | 100 | 172 | 1085 | |

75 | 12 | 99 | 169 | 1064 | |

100 | 12 | 99 | 169 | 1074 |

**Table 3.**The solution time of solving a robust optimization problem with 100 optimization scenarios.

# Processors | Full Factorization | Schur Complement Method | ||
---|---|---|---|---|

Time(s) | Time(s) | Speedup | ||

Building Model | 1 | 44.3 | 64.2 | - |

2 | - | 34.8 | 1.8 | |

5 | - | 14.9 | 4.3 | |

10 | - | 8.6 | 7.5 | |

20 | - | 6.3 | 10.2 | |

25 | - | 4.7 | 13.7 | |

Solving NLP | 1 | 406 | 426.9 | - |

2 | - | 216.3 | 2.0 | |

5 | - | 90.8 | 4.7 | |

10 | - | 51.0 | 8.4 | |

20 | - | 35.8 | 11.9 | |

25 | - | 30.0 | 14.2 |

**Table 4.**Robust performance of min–max NMPC with different numbers of optimization scenarios from Bayesian inference evaluated using 100 simulations.

Type | S | Nominal | Average | Standard Deviation | Worst-Case |
---|---|---|---|---|---|

Min–max | 12 | 18 | 74 | 120 | 744 |

25 | 13 | 61 | 96 | 584 | |

50 | 11 | 71 | 114 | 655 | |

Min–expected | 12 | 17 | 81 | 141 | 943 |

25 | 12 | 84 | 145 | 949 | |

50 | 11 | 84 | 145 | 934 |

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Cao, Y.; Kang, J.; Nagy, Z.K.; Laird, C.D.
Parallel Solution of Robust Nonlinear Model Predictive Control Problems in Batch Crystallization. *Processes* **2016**, *4*, 20.
https://doi.org/10.3390/pr4030020

**AMA Style**

Cao Y, Kang J, Nagy ZK, Laird CD.
Parallel Solution of Robust Nonlinear Model Predictive Control Problems in Batch Crystallization. *Processes*. 2016; 4(3):20.
https://doi.org/10.3390/pr4030020

**Chicago/Turabian Style**

Cao, Yankai, Jia Kang, Zoltan K. Nagy, and Carl D. Laird.
2016. "Parallel Solution of Robust Nonlinear Model Predictive Control Problems in Batch Crystallization" *Processes* 4, no. 3: 20.
https://doi.org/10.3390/pr4030020