Article
Automation & Control Systems
Guanru Pan, Ruchuan Ou, Timm Faulwasser
Summary: Data-driven stochastic predictive control scheme is proposed for LTI systems subject to unbounded additive process disturbances. A data-driven surrogate optimal control problem is constructed using a stochastic extension of the fundamental lemma and leveraging polynomial chaos expansions. Sufficient conditions for recursive feasibility and stability of the proposed scheme are provided, along with an online selection strategy of the initial condition. Numerical examples illustrate the efficacy and closed-loop properties of the proposed scheme for process disturbances governed by different distributions.
INTERNATIONAL JOURNAL OF ROBUST AND NONLINEAR CONTROL
(2023)
Article
Automation & Control Systems
Petr Listov, Johannes Schwarz, Colin N. Jones
Summary: This paper discusses the importance of model-based methods in autonomous driving and advanced driving assistance, and addresses the problem of safe trajectory planning under parametric model uncertainties in automotive applications. The proposed approach utilizes generalized polynomial chaos expansions and distributionally robust inequalities to efficiently propagate nonlinear uncertainties and approximate chance constraints. By incorporating an ancillary feedback controller inspired by tube-based model predictive control, the approach reduces variance and ensures constraints satisfaction with high probability.
OPTIMAL CONTROL APPLICATIONS & METHODS
(2023)
Article
Engineering, Aerospace
Rui Cao, Yanbin Liu, Yuping Lu
Summary: This paper proposes a multi-constrained robust trajectory optimization method to reduce the design burden of Aerospace Vehicles control systems. The method combines the characteristics of ASVs, sets control performance as an optimization objective, considers parameter uncertainty, and uses polynomial chaos expansion algorithm.
CHINESE JOURNAL OF AERONAUTICS
(2022)
Article
Engineering, Mechanical
Xujia Zhu, Marco Broccardo, Bruno Sudret
Summary: The fragility model plays a key role in the performance-based earthquake engineering (PBEE) framework. The computation of such models is a challenge, and there is still a research gap in this domain. This study proposes a new method using stochastic polynomial chaos expansions to estimate the conditional distribution and fragility functions, and compares it with state-of-the-art methods in two case studies.
PROBABILISTIC ENGINEERING MECHANICS
(2023)
Article
Energy & Fuels
Andrea Pozzi, Davide M. Raimondo
Summary: This paper proposes the use of stochastic MPC for optimal charging of a Li-ion battery pack to account for parameter uncertainties. The results highlight the advantages of stochastic MPC over deterministic MPC in different scenarios.
JOURNAL OF ENERGY STORAGE
(2022)
Article
Engineering, Multidisciplinary
Xiaoshu Zeng, Roger Ghanem
Summary: This work addresses the accurate stochastic approximations in high-dimensional parametric space using uncertainty quantification tools. A novel approach combining basis adaptation and projection pursuit regression is proposed to simultaneously learn the optimal low-dimensional spaces and polynomial chaos expansions from given data. The method demonstrates the ability to discover low-dimensional manifolds and learn surrogate models with high accuracy with limited data.
COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING
(2023)
Article
Automation & Control Systems
Weixin Han, Zhenhua Wang, Yi Shen, Bin Xu
Summary: This article proposes a two-step interval estimation method for linear systems with time-invariant probabilistic uncertainty, utilizing PCE and zonotopic technique. By approximating error dynamics via PCE and analyzing intervals of the expanded system with zonotopic technique, the interval estimation is achieved by combining nominal observer state and estimated error interval. Experimental and simulation examples in a case study demonstrate the effectiveness of the proposed method.
IEEE TRANSACTIONS ON AUTOMATIC CONTROL
(2021)
Article
Automation & Control Systems
Arman Oshnoei, Morteza Kheradmandi, Rahmat Khezri, Amin Mahmoudi
Summary: This article presents a robust model predictive control for gate-controlled series capacitor (GCSC) to assist in load frequency control. The control model involves a two-layer MPC to minimize frequency response errors between nominal and actual systems, optimize weighting coefficients, and consider system constraints and GCSC control. Case studies demonstrate the method's effectiveness in handling uncertainties and the impact of communication delay on control scheme performance.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Article
Engineering, Geological
Qiangqiang Sun, Daniel Dias
Summary: This study investigates the feasibility of using surrogate models for global sensitivity analysis in seismic soil-tunnel systems, validating the accuracy and efficiency of the method. Parametric sensitivity analysis reveals that soil shear wave velocity and modulus reduction factor are key variables influencing seismic deformations in tunnels.
SOIL DYNAMICS AND EARTHQUAKE ENGINEERING
(2021)
Article
Mathematics, Applied
Maxime Breden
Summary: Generalized polynomial chaos (gPC) expansions are a powerful tool for efficiently approximating random invariant sets associated with differential equations with random coefficients. This work introduces a new framework for conducting validated continuation in parameter-dependent systems, allowing for rigorous computation of isolated branches of solutions. The proposed methodology is applied to compute random invariant periodic orbits in the Lorenz system and steady states of the Swift-Hohenberg equation.
SIAM JOURNAL ON APPLIED DYNAMICAL SYSTEMS
(2023)
Article
Computer Science, Interdisciplinary Applications
Mohammad Jamous, Reza Marsooli, Mahmoud Ayyad
Summary: Predicting coastal erosion requires an accurate morphodynamic model, such as XBeach. Sensitivity analysis is conducted using a computationally cost-effective approach based on the Non-Intrusive Polynomial Chaos Expansion method. Applied to Hurricane Sandy-induced coastal erosion in New Jersey, the results demonstrate the spatial variation of model sensitivity and the reduction of parameter interaction with increasing boundary conditions, leading to a reduction in uncertainty of model output.
ENVIRONMENTAL MODELLING & SOFTWARE
(2023)
Article
Thermodynamics
Sufia Khatoon, Jyoti Phirani, Supreet Singh Bahga
Summary: We propose a fast Bayesian inference framework for solving inverse heat conduction problems. The framework combines polynomial chaos expansions and dimensionality reduction based on Karhunen-Loeve expansion to generate efficient surrogate models. We demonstrate the potential of this approach using three model problems for heat flux estimation.
APPLIED THERMAL ENGINEERING
(2023)
Article
Mathematics, Applied
Alberto Olivares, Ernesto Staffetti
Summary: This paper utilizes a spectral approach to formulate and solve robust optimal control problems for compartmental epidemic models, allowing uncertainty propagation and variability representation through a polynomial expansion of stochastic state variables. It provides increased predictability, estimation of probability distributions, and global sensitivity analysis for optimal control strategies of epidemic models, as demonstrated in a COVID-19 transmission model.
COMMUNICATIONS IN NONLINEAR SCIENCE AND NUMERICAL SIMULATION
(2022)
Article
Computer Science, Interdisciplinary Applications
Laura Lyman, Gianluca Iaccarino
Summary: Polynomial chaos methods can be used to approximate solutions to PDEs with stochastic inputs. Challenges include difficulty in reusing existing solutions from previous systems and lack of flexibility in choosing important variables in the system. The bluff-and-fix iterative strategy proposed in Lyman and Iaccarino (2020) addresses these challenges effectively.
JOURNAL OF COMPUTATIONAL SCIENCE
(2021)
Article
Computer Science, Interdisciplinary Applications
Laura Lyman, Gianluca Iaccarino
Summary: Polynomial chaos methods are used to approximate the solution to PDEs with stochastic inputs by expressing the solution as a truncated infinite polynomial expansion. A promising iterative strategy (bluff-and-fix) is introduced to address challenges in reusing existing solutions and choosing variables to estimate accurately, showing that bluff-and-fix can be more effective than monolithic methods. This extended version of the study demonstrates how bluff-and-fix successfully allows for choice in accurately approximating variables in estimating statistical properties of the system solution.
JOURNAL OF COMPUTATIONAL SCIENCE
(2021)
