Article
Computer Science, Artificial Intelligence
Ethan N. Evans, Andrew P. Kendall, Evangelos A. Theodorou
Summary: Correlated with the trend of increasing degrees of freedom in robotic systems is a rising interest in spatio-temporal systems described by partial differential equations among the robotics and control communities. These systems exhibit dramatic under-actuation, high dimensionality, bifurcations, and multimodal instabilities, posing significant challenges for control and actuation design. A novel sampling-based stochastic optimization framework in Hilbert spaces is proposed for semi-linear SPDEs in robotics and applied physics, demonstrating efficacy through simulated experiments on various systems including a soft robotic manipulator with infinite degrees of freedom.
Article
Engineering, Industrial
Andre Gustavo Carlon, Henrique Machado Kroetz, Andre Jacomel Torii, Rafael Holdorf Lopez, Leandro Fleck Fadel Miguel
Summary: This paper proposes a stochastic gradient based method for Risk Optimization problems, approximating failure probabilities using the Chernoff bound and solving the problem with a Stochastic Gradient Descent algorithm. The approach efficiently avoids direct computation of failure probabilities and their gradients.
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2022)
Article
Automation & Control Systems
Margarita Rebolledo, Daan Zeeuwe, Thomas Bartz-Beielstein, A. E. Eiben
Summary: Evolutionary robotics focuses on optimizing autonomous robots for specific tasks. This paper studies the impact of energy consumption on robot evolution by adding a battery module to the robot simulator framework. The results show that considering energy consumption improves the robots' task performance and speeds up the evolution process.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2022)
Article
Thermodynamics
Tobi Michael Alabi, Lin Lu, Zaiyue Yang
Summary: A novel modelling strategy for a realistic and highly flexible zero-carbon multi-energy system (ZCMES) is proposed, incorporating energy storage aging influence and integrated demand response (IDR). The approach involves multi-objective optimization and Markowitz portfolio risk theory to evaluate and mitigate uncertainties during decision-making. The simulation results demonstrate significant reductions in investment cost, operation cost, and overall expenditure, providing insights into optimal planning for ZCMES.
Article
Engineering, Electrical & Electronic
Amrit Singh Bedi, Alec Koppel, Ketan Rajawat, Panchajanya Sanyal
Summary: This work addresses optimization problems with nonlinear objective functions of expected values, introducing a memory-efficient stochastic algorithm COLK for compositional stochastic programs. The tradeoff between complexity of function parameterization and convergence accuracy is provided for both convex and non-convex objectives under constant step-sizes. Experimental results demonstrate COLK's consistent convergence and reliability in risk-sensitive supervised learning tasks, showing a favorable tradeoff between model complexity, convergence, and statistical accuracy for heavy-tailed data distributions.
IEEE TRANSACTIONS ON SIGNAL PROCESSING
(2021)
Article
Automation & Control Systems
Hong Cai, Yasamin Mostofi
Summary: This paper addresses the problem of a robot navigating from a start position to a destination, sensing sites along the way, and transmitting collected data to a remote station. The goal is to minimize the robot's energy costs by co-optimizing its path, data transmission, and sensing decisions. The authors propose a specially-designed MDP and utilize MCTS to efficiently solve the joint optimization problem.
IEEE TRANSACTIONS ON CONTROL OF NETWORK SYSTEMS
(2022)
Article
Computer Science, Software Engineering
S. Garatti, M. C. Campi
Summary: This paper introduces a scenario optimization method based on empirical knowledge and explores the relationship between the risk of not achieving performance or violating constraints and the complexity of the solution. The outcome reveals that the joint probability distribution of risk and complexity exhibits concentration, indicating that complexity can be used to accurately assess risk.
MATHEMATICAL PROGRAMMING
(2022)
Article
Computer Science, Artificial Intelligence
Yifan Shi, Zhiwen Yu, C. L. Philip Chen, Huanqiang Zeng
Summary: This article proposes a novel consensus clustering method, CC-CMO, which improves clustering results by optimizing the co-association matrix and considering information from both label space and feature space. The method achieves optimization on global structure and local affinity, and extensive experiments demonstrate its superior performance compared to existing methods.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Management
Nam Ho-Nguyen, Fatma Kilinc-Karzan
Summary: This paper examines the impact of prediction methods on optimization performance and identifies conditions that ensure good optimization performance. The study analyzes existing methods and conducts computational experiments to demonstrate the detrimental effect of the lack of Fisher consistency in prediction methods on performance.
MANAGEMENT SCIENCE
(2022)
Article
Thermodynamics
Yixin Liu, Haoqi Shi, Li Guo, Tao Xu, Bo Zhao, Chengshan Wang
Summary: This paper proposes a risk-aware energy scheduling and stochastic optimization method to enhance the long-term operational reliability of independent microgrids. The method considers future extreme scenarios and maximizes the reliable power supply probability through optimizing energy scheduling strategies and reserve requirements.
Article
Engineering, Aerospace
Jincheng Hu, Hongwei Yang, Shuang Li, Yingjie Zhao
Summary: To address the issue of time-consuming training in reinforcement learning caused by sparse reward functions, an efficient dense reward framework is proposed for robust low-thrust trajectory optimization. The framework includes dense reward functions designed separately for deterministic and uncertain scenarios, and incorporates exponential acceleration terms to improve training efficiency in rendezvous missions. Empirical forbidden zones and exponential terms are also introduced to design a dense reward function for the deterministic scenario. The method is demonstrated to be effective and efficient in Earth-Mars and Earth-Venus missions.
ADVANCES IN SPACE RESEARCH
(2023)
Article
Robotics
Sangsin Park
Summary: The researcher presented and derived a closed-form solution for an optimal ZMP pattern, meeting ZMP boundary conditions and an additional constraint. The solution allows for generating a walking pattern for every step period and connecting seamlessly with previous patterns. Real-time adjustments to the walking pattern are demonstrated based on varying step lengths.
JOURNAL OF FIELD ROBOTICS
(2022)
Article
Mathematics, Applied
Florian Beiser, Brendan Keith, Simon Urbainczyk, Barbara Wohlmuth
Summary: We introduce adaptive sampling methods for stochastic programs with deterministic constraints. First, we propose and analyze a variant of the stochastic projected gradient method, where the sample size used to approximate the reduced gradient is determined on-the-fly and updated adaptively. Second, we propose an SQP-type method based on similar adaptive sampling principles. The benefits of these methods are demonstrated in numerical experiments and a simplified engineering design application.
IMA JOURNAL OF NUMERICAL ANALYSIS
(2023)
Article
Economics
Noureddine Kouaissah
Summary: In this paper, a framework is proposed for robustifying reward-risk-based portfolio optimization with weak type second-order stochastic dominance constraints, significantly improving upon conventional versions. By utilizing stable sub-Gaussian and Student's t distributions, a popular robust optimization technique in conventional statistical estimation methods is extended, and a new notion of weak second-order stochastic dominance is considered. The effects of distributional assumptions on optimal portfolios are studied, and estimation errors are directly addressed in the portfolio optimization process. Empirical analyses demonstrate that the robustified formulations improve performance measures for out-of-sample portfolios.
QUARTERLY REVIEW OF ECONOMICS AND FINANCE
(2023)
Article
Computer Science, Hardware & Architecture
Mario Di Mauro, Giovanni Galatro, Fabio Postiglione, Marco Tambasco
Summary: Service provisioning mechanisms in 5G infrastructures utilize the concept of network service chain, coupled with NFV paradigm. The IP Multimedia Subsystem (IMS) is a key infrastructure of 5G networks, characterized by performance and availability aspects. Through a designed testbed, softIMS architecture is assessed in terms of performance and availability using stochastic analysis, optimizing resource allocation and minimizing deployment costs.
IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING
(2022)