检索结果-内蒙古大学图书馆

Pursuit-Evasion on a Sphere and When It Can Be Considered Flat

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Milutinović, Dejan Von Moll, Alexander Manyam, Satyanarayana G. Casbeer, David W. Weintraub, Isaac E. Pachter, Meir Electrical and Computer Engineering Dept. University of California Santa Cruz Santa CruzCA United States The Control Science Center Aerospace Systems Directorate Air Force Research Laboratory DaytonOH45435 United States DCS Corporation DaytonOH United States Dept. of Electrical Engineering Air Force Institute of Technology Wright-Patterson AFB OH United States

In classical works on a planar differential pursuit-evasion game with a faster pursuer, the intercept point resulting from the equilibrium strategies lies on the Apollonius circle. This property was exploited for the construction of the equilibrium strategies for two faster pursuers against one evader. Extensions for planar multiple-pursuer single-evader scenarios have been considered. We study a pursuit-evasion game on a sphere and the relation of the equilibrium intercept point to the Apollonius domain on the sphere. The domain is a generalization of the planar Apollonius circle set. We find a condition resulting in the intercept point belonging to the Apollonius domain, which is the characteristic of the planar game solution. Finally, we use this characteristic to discuss pursuit and evasion strategies in the context of two pursuers and a single slower evader on the sphere and illustrate it using numerical *** Codes 58E10, 86A30, 37D40, 91A23, 49L12, 49Q10 © 2024, CC BY-NC-ND.

关键词： Spheres

Using AoI forecasts in communicating and robust distributed model-predictive control

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Hahn, Jannik Schoeffauer, Richard Wunder, Gerhard Stursberg, Olaf Control and System Theory Dept. of Electrical Engineering and Computer Science University of Kassel Germany Heisenberg CIT Group Free University of Berlin Dept. of Mathematics and Computer Science Germany

In order to enhance the performance of cyber-physical systems, this paper proposes the integrated design of distributed controllers for distributed plants and the control of the communication network. Conventional design methods use static interfaces between both entities and therefore rely on worst-case estimations of communication delay, often leading to conservative behavior of the overall system. By contrast, the present approach establishes a robust distributed model-predictive control scheme, in which the local subsystem controllers operate under the assumption of a variable communication schedule that is predicted by a network controller. Using appropriate models for the communication network, the network controller applies a predictive network policy for scheduling the communication among the subsystem controllers across the network. Given the resulting time-varying predictions of the age of information, the paper shows under which conditions the subsystem controllers can robustly stabilize the distributed system. To illustrate the approach, the paper also reports on the application to a vehicle platooning scenario. © 2021, CC BY.

关键词： Networked control systems

Feedforward control in the Presence of Input Nonlinearities: A Learning-based Approach

学校读者我要写书评

暂无评论

arXiv 2022年

作者： van Hulst, Jilles Poot, Maurice Kostić, Dragan Yan, Kai Wa Portegies, Jim Oomen, Tom Control Systems Technology Section Dept. of Mechanical Engineering Eindhoven University of Technology Netherlands Center of Competency ASM Pacific Technology Netherlands CASA Dept. of Mathematics and Computer Science Eindhoven University of Technology Netherlands Delft Center for Systems and Control Delft University of Technology Netherlands

Advanced feedforward control methods enable mechatronic systems to perform varying motion tasks with extreme accuracy and throughput. The aim of this paper is to develop a data-driven feedforward controller that addresses input nonlinearities, which are common in typical applications such as semiconductor back-end equipment. The developed method consists of parametric inverse-model feedforward that is optimized for tracking error reduction by exploiting ideas from iterative learning control. Results on a simulated set-up indicate improved performance over existing identification methods for systems with nonlinearities at the input. © 2022, CC BY-NC-ND.

关键词： Feedforward control

Beyond bandit feedback in online multiclass classification 21

学校读者我要写书评

暂无评论

Beyond bandit feedback in online multiclass classification

Proceedings of the 35th International Conference on Neural Information Processing Systems

作者： Dirk van der Hoeven Federico Fusco Nicolò Cesa-Bianchi Dept. of Computer Science Università degli Studi di Milano Italy Dept. of Computer Control and Management Engineering Sapienza Università di Roma Italy DSRC & Dept. of Computer Science Università degli Studi di Milano Italy

ISBN: (纸本)9781713845393

We study the problem of online multiclass classification in a setting where the learner's feedback is determined by an arbitrary directed graph. While including bandit feedback as a special case, feedback graphs allow a much richer set of applications, including filtering and label efficient classification. We introduce GAP-PLETRON, the first online multiclass algorithm that works with arbitrary feedback graphs. For this new algorithm, we prove surrogate regret bounds that hold, both in expectation and with high probability, for a large class of surrogate losses. Our bounds are of order $B\sqrt{\rho KT}$, where B is the diameter of the prediction space, K is the number of classes, T is the time horizon, and ρ is the domination number (a graph-theoretic parameter affecting the amount of exploration). In the full information case, we show that GAPPLETRON achieves a constant surrogate regret of order B2K. We also prove a general lower bound of order max {B2K, √T} showing that our upper bounds are not significantly improvable. Experiments on synthetic data show that for various feedback graphs our algorithm is competitive against known baselines.

关键词：

Beyond bandit feedback in online multiclass classification

学校读者我要写书评

暂无评论

arXiv 2021年

作者： van der Hoeven, Dirk Fusco, Federico Cesa-Bianchi, Nicolò Dept. of Computer Science Università degli Studi di Milano Italy Dept. of Computer Control and Management Engineering Sapienza Università di Roma Italy DSRC Dept. of Computer Science Università degli Studi di Milano Italy

We study the problem of online multiclass classification in a setting where the learner’s feedback is determined by an arbitrary directed graph. While including bandit feedback as a special case, feedback graphs allow a much richer set of applications, including filtering and label efficient classification. We introduce Gappletron, the first online multiclass algorithm that works with arbitrary feedback graphs. For this new algorithm, we prove surrogate regret bounds that hold, both in expectation and with high probability, for a large class of surrogate losses. Our bounds are of order B√ρKT, where B is the diameter of the prediction space, K is the number of classes, T is the time horizon, and ρ is the domination number (a graph-theoretic parameter affecting the amount of exploration). In the full information case, we show that Gappletron achieves a constant surrogate regret of order B2K. We also prove a general lower bound of order max {B2K, √T} showing that our upper bounds are not significantly improvable. Experiments on synthetic data show that for various feedback graphs our algorithm is competitive against known baselines. Copyright © 2021, The Authors. All rights reserved.

关键词： Directed graphs

Polytopic input constraints in learning-based optimal control using neural networks

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Markolf, Lukas Stursberg, Olaf Control and System Theory Group Dept. of Electrical Engineering and Computer Science University of Kassel Germany

This work considers artificial feed-forward neural networks as parametric approximators in optimal control of discrete-time systems. Two different approaches are introduced to take polytopic input constraints into account. The first approach determines (sub-)optimal inputs by the application of gradient methods. Closed-form expressions for the gradient of general neural networks with respect to their inputs are derived. The approach allows to consider state-dependent input constraints, as well as to ensure the satisfaction of state constraints by exploiting recursive reachable set computations. The second approach makes use of neural networks with softmax output units to map states into parameters, which determine (sub-)optimal inputs by a convex combination of the vertices of the input constraint set. The application of both approaches in model predictive control is discussed, and results obtained for a numerical example are used for illustration. Copyright © 2021, The Authors. All rights reserved.

关键词： Discrete time control systems

Grid-following three-phase inverter with RMS current-limiting property 12

学校读者我要写书评

暂无评论

Grid-following three-phase inverter with RMS current-limitin...

12th Mediterranean Conference on Power Generation, Transmission, Distribution and Energy Conversion, MEDPOWER 2020

作者： Skentzos, Panagiotis Konstantopoulos, George C. Dept. of Automatic Control and Systems Engineering University of Sheffield Mappin Street SheffieldS1 3JD United Kingdom Dept. of Electrical and Computer Engineering University of Patras Rio26504 Greece

ISBN: (纸本)9781839535635

A novel nonlinear controller that inherently limits the RMS value of a three-phase grid-following inverter current is presented in this paper. Using the dynamic model of the inverter in the rotating dq reference frame, a nonlinear control scheme is formulated to ensure accurate real and reactive power regulation with RMS current limitation that avoids the use of saturation units with upper/lower limit adaptation, as commonly used in the existing approaches. In particular, it is analytically proven by using fundamental nonlinear systems theory, that the RMS value of the inverter current will never violate a given maximum value, both during the transient response and at the steady-state. Moreover, a detailed root locus analysis is provided to guarantee the asymptotic stability of the desired equilibrium point for the closed-loop system. Finally, to validate the proposed controller performance, simulation results of a grid-following three-phase inverter equipped with the proposed control approach are presented under both normal and unrealistic values of the real and reactive power setpoints. © 2020 Institution of Engineering and Technology. All rights reserved.

关键词： Asymptotic stability

Smart tourism chatbot system using Multi-domain Tourism Information DST

学校读者我要写书评

暂无评论

Smart tourism chatbot system using Multi-domain Tourism Info...

International Conference on Ubiquitous and Future Networks, ICUFN

作者： Hoon-Chul Kang Ki-Beom Kang Dong-Hyun Kim Myeong-Cheol Jwa Tae-Seung Ko Jeong-Woo Jwa Dept. of Smart Electronic Control Jeju Campus of Korea Polytechnic Jeju South Korea National federation of fisheries cooperatives Jeju South Korea Dept. of Telecomm. Eng. Jeju National University Jeju South Korea School of Computer Science and Eng Korea University of Technology and Education Cheonan-si South Korea

The smart tourism service provides tourists with travel planner services and tour guide services for easy and convenient travel throughout the entire travel process. In this paper, we develop the AI-based chatbot service using a pretrained language model (PLM) and provide tourism information so that tourists can make their travel plans. The proposed chatbot system consists of the DST server, the Neo4J graph DB and MySQL DB servers, and the natural language generation (NLG) server. The dialogue state tracking (DST) server understands the intention of tourists’ questions to overcome the shortcomings of the previous rule-based chatbot system [7]. We define the domains and slots of the tourism information DST model with the 4W1H method and develop the dataset [12] for transfer learning of the SOM DST model [14]. The Neo4J and MySQL web servers search tourism information from the tourism information knowledgebase and the smart tourism information system, respectively. The NLG server provides the searched tourism information to the smart tourism app.

关键词：

FDP-GAMLP: New Financial Distress Prediction Technique based on Genetic Algorithm and Multi-Layer Perceptron 19

学校读者我要写书评

暂无评论

FDP-GAMLP: New Financial Distress Prediction Technique based...

19th IEEE International Multi-Conference on Systems, Signals and Devices, SSD 2022

作者： Khedr, Ahmed M. Bannany, Magdi El Saif, Abdul-Wahid A. University of Sharjah Computer Science Dept. Sharjah27272 United Arab Emirates College of Business Administration Umm Al Quwain University Umm Al Quwain356 United Arab Emirates Ain Shams University Faculty of Business Cairo11566 Egypt King Fahd University of Petroleum & Minerals Control and Instrumentation Engineering Department Dhahran Saudi Arabia

ISBN: (数字)9781665471084

ISBN: (纸本)9781665471084

Developing a reliable financial distress prediction model has been a long-standing research area. Recently, machine learning algorithms have been increasingly popular in the area. In this paper, a hybrid approach of Genetic Algorithm (GA) and Multi-Layer Perceptron (MLP) for Financial Distress Prediction (FDP) (FDP-GAMLP) is proposed. FDP-GAMLP emphasizes on GA-based tuning of the four major hyper-parameters, namely network width, network dept., network optimizer, and dense layer activation function, which can influence on whether the algorithm explodes or converges. An improved GA is utilized to optimize the MLP model's hyper-parameters for better prediction. The prediction performance is evaluated using real data set with samples of companies from countries in Middle East region. The resampling technique using k-fold evaluation metrics is adopted to get unbiased and most accurate results. The simulation results show that FDP-GAMLP outperforms the classical machine learning models in terms of predictive accuracy. © 2022 IEEE.

关键词： Genetic algorithms