kth.sePublications
Change search
Link to record
Permanent link

Direct link
Alternative names
Publications (10 of 971) Show all publications
Sharf, M., Besselink, B. & Johansson, K. H. (2024). Contract composition for dynamical control systems: Definition and verification using linear programming. Automatica, 164, Article ID 111637.
Open this publication in new window or tab >>Contract composition for dynamical control systems: Definition and verification using linear programming
2024 (English)In: Automatica, ISSN 0005-1098, E-ISSN 1873-2836, Vol. 164, article id 111637Article in journal (Refereed) Published
Abstract [en]

Designing large-scale control systems to satisfy complex specifications is hard in practice, as most formal methods are limited to systems of modest size. Contract theory has been proposed as a modular alternative, in which specifications are defined by assumptions on the input to a component and guarantees on its output. However, current contract-based methods for control systems either prescribe guarantees on the state of the system, going against the spirit of contract theory, or are not supported by efficient computational tools. In this paper, we present a contract-based modular framework for discrete-time dynamical control systems. We extend the definition of contracts by allowing the assumption on the input at a time k to depend on outputs up to time k−1, which is essential when considering feedback systems. We also define contract composition for arbitrary interconnection topologies, and prove that this notion supports modular design, analysis and verification. This is done using graph theory methods, and specifically using the notions of topological ordering and backward-reachable nodes. Lastly, we present an algorithm for verifying vertical contracts, which are claims of the form “the conjunction of given component-level contracts implies given contract on the integrated system”. These algorithms are based on linear programming, and scale linearly with the number of components in the interconnected network. A numerical example is provided to demonstrate the scalability of the presented approach, as well as the modularity achieved by using it.

Place, publisher, year, edition, pages
Elsevier BV, 2024
Keywords
Contracts, Formal methods, Graph theory, Interconnection topology, Linear programming, Modular design
National Category
Computer Sciences
Identifiers
urn:nbn:se:kth:diva-345238 (URN)10.1016/j.automatica.2024.111637 (DOI)2-s2.0-85189076874 (Scopus ID)
Note

QC 20240411

Available from: 2024-04-10 Created: 2024-04-10 Last updated: 2024-04-11Bibliographically approved
Liu, C., Johansson, K. H. & Shi, Y. (2024). Distributed empirical risk minimization with differential privacy. Automatica, 162, Article ID 111514.
Open this publication in new window or tab >>Distributed empirical risk minimization with differential privacy
2024 (English)In: Automatica, ISSN 0005-1098, E-ISSN 1873-2836, Vol. 162, article id 111514Article in journal (Refereed) Published
Abstract [en]

This work studies the distributed empirical risk minimization (ERM) problem under differential privacy (DP) constraint. Standard distributed algorithms achieve DP typically by perturbing all local subgradients with noise, leading to significantly degenerated utility. To tackle this issue, we develop a class of private distributed dual averaging (DDA) algorithms, which activates a fraction of nodes to perform optimization. Such subsampling procedure provably amplifies the DP guarantee, thereby achieving an equivalent level of DP with reduced noise. We prove that the proposed algorithms have utility loss comparable to centralized private algorithms for both general and strongly convex problems. When removing the noise, our algorithm attains the optimal O(1/t) convergence for non-smooth stochastic optimization. Finally, experimental results on two benchmark datasets are given to verify the effectiveness of the proposed algorithms.

Place, publisher, year, edition, pages
Elsevier BV, 2024
Keywords
Differential privacy, Distributed optimization, Dual averaging, Empirical risk minimization
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-342825 (URN)10.1016/j.automatica.2024.111514 (DOI)001170781200001 ()2-s2.0-85182977636 (Scopus ID)
Note

QC 20240201

Available from: 2024-01-31 Created: 2024-01-31 Last updated: 2024-03-18Bibliographically approved
Xu, L., Yi, X., Shi, Y., Johansson, K. H., Chai, T. & Yang, T. (2024). Distributed Nonconvex Optimization With Event-Triggered Communication. IEEE Transactions on Automatic Control, 69(4), 2745-2752
Open this publication in new window or tab >>Distributed Nonconvex Optimization With Event-Triggered Communication
Show others...
2024 (English)In: IEEE Transactions on Automatic Control, ISSN 0018-9286, E-ISSN 1558-2523, Vol. 69, no 4, p. 2745-2752Article in journal (Refereed) Published
Abstract [en]

This article considers distributed nonconvex optimization for minimizing the sum of local cost functions by using local information exchange. In order to avoid continuous communication among agents and reduce communication overheads, we develop a distributed algorithm with a dynamic exponentially decaying event-triggered scheme. We show that the proposed algorithm is free of Zeno behavior (i.e., finite number of triggers in any finite time interval) by contradiction and asymptotically converges to a stationary point if the local cost functions are smooth. Moreover, we show that the proposed algorithm exponentially converges to the global optimal point if, in addition, the global cost function satisfies the Polyak-Lojasiewicz condition, which is weaker than the standard strong convexity condition, and the global minimizer is not necessarily unique. The theoretical results are illustrated by a numerical simulation example.

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2024
Keywords
Distributed nonconvex algorithm, event-triggered communication, exponential convergence, Polyak-Lojasiewicz (P-L) condition, Zeno behavior
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-345957 (URN)10.1109/TAC.2023.3339439 (DOI)001194518600062 ()2-s2.0-85179789794 (Scopus ID)
Note

QC 20240430

Available from: 2024-04-30 Created: 2024-04-30 Last updated: 2024-04-30Bibliographically approved
Ruggeri, F., Terra, A., Inam, R. & Johansson, K. H. (2024). Evaluation of Intrinsic Explainable Reinforcement Learning in Remote Electrical Tilt Optimization. In: Proceedings of 8th International Congress on Information and Communication Technology - ICICT 2023: . Paper presented at 8th International Congress on Information and Communication Technology, ICICT 2023, London, United Kingdom of Great Britain and Northern Ireland, Feb 20 2023 - Feb 23 2023 (pp. 835-854). Springer Nature
Open this publication in new window or tab >>Evaluation of Intrinsic Explainable Reinforcement Learning in Remote Electrical Tilt Optimization
2024 (English)In: Proceedings of 8th International Congress on Information and Communication Technology - ICICT 2023, Springer Nature , 2024, p. 835-854Conference paper, Published paper (Refereed)
Abstract [en]

This paper empirically evaluates two intrinsic Explainable Reinforcement Learning (XRL) algorithms on the Remote Electrical Tilt (RET) optimization problem. In RET optimization, where the electrical downtilt of the antennas in a cellular network is controlled to optimize coverage and capacity, explanations are necessary to understand the reasons behind a specific adjustment. First, we formulate the RET problem in the reinforcement learning (RL) framework and describe how we apply Decomposed Reward Deep Q Network (drDQN) and Linear ModelU-Tree (LMUT), which are two state-of-the-art XRL algorithms. Then, we train and test such agents in a realistic simulated network. Our results highlight both advantages and disadvantages of the algorithms. DrDQN provides intuitive contrastive local explanations for the agent’s decisions to adjust the downtilt of an antenna, while achieving the same performance as the original DQN algorithm. LMUT reaches high performance while employing a fully transparent linear model capable of generating both local and global explanations. On the other hand, drDQN adds a constraint on the reward design that might be problematic for the specification of the objective, whereas LMUT could generate misleading global feature importance and needs additional developments to provide more user-interpretable local explanations.

Place, publisher, year, edition, pages
Springer Nature, 2024
Keywords
Artificial intelligence, Cellular networks, Explainable reinforcement learning, Reinforcement learning, Remote electrical tilt optimization
National Category
Computer Sciences
Identifiers
urn:nbn:se:kth:diva-339274 (URN)10.1007/978-981-99-3236-8_67 (DOI)2-s2.0-85174736754 (Scopus ID)
Conference
8th International Congress on Information and Communication Technology, ICICT 2023, London, United Kingdom of Great Britain and Northern Ireland, Feb 20 2023 - Feb 23 2023
Note

Part of ISBN 9789819932351

QC 20231106

Available from: 2023-11-06 Created: 2023-11-06 Last updated: 2023-11-06Bibliographically approved
Rikos, A. I., Hadjicostis, C. N. & Johansson, K. H. (2024). Finite time quantized average consensus with transmission stopping guarantees and no quantization error. Automatica, 163, Article ID 111522.
Open this publication in new window or tab >>Finite time quantized average consensus with transmission stopping guarantees and no quantization error
2024 (English)In: Automatica, ISSN 0005-1098, E-ISSN 1873-2836, Vol. 163, article id 111522Article in journal (Refereed) Published
Abstract [en]

Networked control systems, which are composed of spatially distributed sensors and actuators that communicate through wireless networks, are emerging as a fundamental infrastructure technology in 5G and IoT technologies. In order to increase flexibility and reduce deployment and maintenance costs, their operation needs to guarantee (i) efficient communication between nodes and (ii) preservation of available energy. Motivated by these requirements, we present and analyze a novel distributed average consensus algorithm, which (i) operates exclusively on quantized values (in order to guarantee efficient communication and data storage), (ii) relies on event-driven updates (in order to reduce energy consumption, communication bandwidth, network congestion, and/or processor usage), and (iii) allows each node to cease transmissions once the exact average of the initial quantized values has been reached (in order to preserve its stored energy). We characterize the properties of the proposed algorithm and show that its execution, on any time-invariant and strongly connected digraph, allows all nodes to reach in finite time a common consensus value that is equal to the exact average (represented as the ratio of two quantized values). Then, we present upper bounds on (i) the number of transmissions and computations each node has to perform during the execution of the algorithm, and (ii) the memory and energy requirements of each node in order for the algorithm to be executed. Finally, we provide examples that demonstrate the operation, performance, and potential advantages of our proposed algorithm.

Place, publisher, year, edition, pages
Elsevier BV, 2024
Keywords
Digraphs, Event-triggered distributed algorithms, Multi-agent systems, Quantization, Quantized average consensus
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-343663 (URN)10.1016/j.automatica.2024.111522 (DOI)2-s2.0-85184838548 (Scopus ID)
Note

QC 20240222

Available from: 2024-02-22 Created: 2024-02-22 Last updated: 2024-02-22Bibliographically approved
Xing, Y. & Johansson, K. H. (2024). Transient behavior of gossip opinion dynamics with community structure. Automatica, 164, Article ID 111627.
Open this publication in new window or tab >>Transient behavior of gossip opinion dynamics with community structure
2024 (English)In: Automatica, ISSN 0005-1098, E-ISSN 1873-2836, Vol. 164, article id 111627Article in journal (Refereed) Published
Abstract [en]

We study transient behavior of gossip opinion dynamics, in which agents randomly interact pairwise over a weighted graph with two communities. Edges within a community have identical weights different from edge weights between communities. We first derive an upper bound for the second moment of agent opinions. Using this result, we obtain upper bounds for probability that a large proportion of agents have opinions close to average opinions. The results imply a phase transition of transient behavior of the process: When edge weights within communities are larger than those between communities and those between regular and stubborn agents, most agents in the same community hold opinions close to the average opinion of that community with large probability, at an early stage of the process. However, if the difference between intra- and inter-community weights is small, most of the agents instead hold opinions close to everyone's average opinion at the early stage. In contrast, when the influence of stubborn agents is large, agent opinions settle quickly to steady state. We then conduct numerical experiments to validate the theoretical results. Different from traditional asymptotic analysis in most opinion dynamics literature, the paper characterizes the influence of stubborn agents and community structure on the initial phase of the opinion evolution.

Place, publisher, year, edition, pages
Elsevier BV, 2024
Keywords
Community structure, Gossip model, Opinion dynamics, Phase transition, Transient behavior
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-345729 (URN)10.1016/j.automatica.2024.111627 (DOI)2-s2.0-85189518224 (Scopus ID)
Note

QC 20240418

Available from: 2024-04-18 Created: 2024-04-18 Last updated: 2024-04-18Bibliographically approved
Deplano, D., Bastianello, N., Franceschelli, M. & Johansson, K. H. (2023). A Unified Approach to Solve the Dynamic Consensus on the Average, Maximum, and Median Values with Linear Convergence. In: 2023 62nd IEEE Conference on Decision and Control, CDC 2023: . Paper presented at 62nd IEEE Conference on Decision and Control, CDC 2023, Singapore, Singapore, Dec 13 2023 - Dec 15 2023 (pp. 6442-6448). Institute of Electrical and Electronics Engineers (IEEE)
Open this publication in new window or tab >>A Unified Approach to Solve the Dynamic Consensus on the Average, Maximum, and Median Values with Linear Convergence
2023 (English)In: 2023 62nd IEEE Conference on Decision and Control, CDC 2023, Institute of Electrical and Electronics Engineers (IEEE) , 2023, p. 6442-6448Conference paper, Published paper (Refereed)
Abstract [en]

This manuscript proposes novel distributed algorithms for solving the dynamic consensus problem in discrete-time multi-agent systems on three different objective functions: the average, the maximum, and the median. In this problem, each agent has access to an external time-varying scalar signal and aims to estimate and track a function of all the signals by exploiting only local communications with other agents. By recasting the problem as an online distributed optimization problem, the proposed algorithms are derived based on the distributed implementation of the alternating direction method of multipliers (ADMM) and are thus amenable to a unified analysis technique. A major contribution is that of proving linear convergence of these ADMM-based algorithms for the specific dynamic consensus problems of interest, for which current results could only guarantee sub-linear convergence. In particular, the tracking error is shown to converge within a bound, whereas the steady-state error is zero. Numerical simulations corroborate the theoretical findings, empirically show the robustness of the proposed algorithms to re-initialization errors, and compare their performance with that of state-of-the-art algorithms.

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2023
Series
Proceedings of the IEEE Conference on Decision and Control, ISSN 0743-1546
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-343705 (URN)10.1109/CDC49753.2023.10383290 (DOI)001166433805046 ()2-s2.0-85184823289 (Scopus ID)
Conference
62nd IEEE Conference on Decision and Control, CDC 2023, Singapore, Singapore, Dec 13 2023 - Dec 15 2023
Note

QC 20240223

 Part of ISBN 979-8-3503-0124-3

Available from: 2024-02-22 Created: 2024-02-22 Last updated: 2024-03-26Bibliographically approved
Yang, Z., Fonseca, J., Zhu, S., Chen, C., Guan, X. & Johansson, K. H. (2023). Adaptive Estimation for Environmental Monitoring Using an Autonomous Underwater Vehicle. In: 2023 62nd IEEE Conference on Decision and Control, CDC 2023l, CDC 2023: . Paper presented at 62nd IEEE Conference on Decision and Control, CDC 2023, Singapore, Singapore, Dec 13 2023 - Dec 15 2023 (pp. 2521-2528). Institute of Electrical and Electronics Engineers (IEEE)
Open this publication in new window or tab >>Adaptive Estimation for Environmental Monitoring Using an Autonomous Underwater Vehicle
Show others...
2023 (English)In: 2023 62nd IEEE Conference on Decision and Control, CDC 2023l, CDC 2023, Institute of Electrical and Electronics Engineers (IEEE) , 2023, p. 2521-2528Conference paper, Published paper (Refereed)
Abstract [en]

This paper considers the problem of monitoring and adaptively estimating an environmental field, such as temperature or salinity, using an autonomous underwater vehicle (AUV). The AUV moves in the field and persistently measures environmental scalars and its position in its local coordinate frame. The environmental scalars are approximately linearly distributed over the region of interest, and an adaptive estimator is designed to estimate the gradient. By orthogonal decomposition of the velocity of the AUV, a linear time-varying system is equivalently constructed, and the sufficient conditions on the motion of the AUV are established, under which the global exponential stability of the estimation error system is rigorously proved. Furthermore, an estimate of the exponential convergence rate is given, and a reference trajectory that maximizes the estimate of the convergence rate is obtained for the AUV to track. Numerical examples verify the stability and efficiency of the system.

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2023
Series
Proceedings of the IEEE Conference on Decision and Control, ISSN 0743-1546
Keywords
Adaptive Estimation, Autonomous Underwater Vehicle, Environmental Monitoring, Exponential Stability
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-343722 (URN)10.1109/CDC49753.2023.10383773 (DOI)2-s2.0-85184828866 (Scopus ID)
Conference
62nd IEEE Conference on Decision and Control, CDC 2023, Singapore, Singapore, Dec 13 2023 - Dec 15 2023
Note

Part of proceedings ISBN 979-835030124-3

QC 20240222

Available from: 2024-02-22 Created: 2024-02-22 Last updated: 2024-03-26Bibliographically approved
Fonseca, J., Rocha, A., Aguiar, M. & Johansson, K. H. (2023). Adaptive Sampling of Algal Blooms Using an Autonomous Underwater Vehicle and Satellite Imagery. In: 2023 IEEE Conference on Control Technology and Applications, CCTA 2023: . Paper presented at 2023 IEEE Conference on Control Technology and Applications, CCTA 2023, Bridgetown, Barbados, Aug 16 2023 - Aug 18 2023 (pp. 638-644). Institute of Electrical and Electronics Engineers (IEEE)
Open this publication in new window or tab >>Adaptive Sampling of Algal Blooms Using an Autonomous Underwater Vehicle and Satellite Imagery
2023 (English)In: 2023 IEEE Conference on Control Technology and Applications, CCTA 2023, Institute of Electrical and Electronics Engineers (IEEE) , 2023, p. 638-644Conference paper, Published paper (Refereed)
Abstract [en]

This paper proposes a method that uses satellite data to improve adaptive sampling missions. We find and track algal bloom fronts using an autonomous underwater vehicle (AUV) equipped with a sensor that measures the concentration of chlorophyll a. Chlorophyll a concentration indicates the presence of algal blooms. The proposed method learns the kernel parameters of a Gaussian process model using satellite images of chlorophyll a from previous days. The AUV estimates the chlorophyll a concentration online using locally collected data. The algal bloom front estimate is fed to the motion control algorithm. The performance of this method is evaluated through simulations using a real dataset of an algal bloom front in the Baltic. We consider a real-world scenario with sensor and localization noise and with a detailed AUV model.

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2023
National Category
Oceanography, Hydrology and Water Resources
Identifiers
urn:nbn:se:kth:diva-338992 (URN)10.1109/CCTA54093.2023.10252251 (DOI)2-s2.0-85173889475 (Scopus ID)
Conference
2023 IEEE Conference on Control Technology and Applications, CCTA 2023, Bridgetown, Barbados, Aug 16 2023 - Aug 18 2023
Note

Part of ISBN 9798350335446

QC 20231123

Available from: 2023-11-01 Created: 2023-11-01 Last updated: 2023-11-23Bibliographically approved
Xing, Y. & Johansson, K. H. (2023). Almost Exact Recovery in Gossip Opinion Dynamics Over Stochastic Block Models. In: 2023 62nd IEEE Conference on Decision and Control, CDC 2023: . Paper presented at 62nd IEEE Conference on Decision and Control, CDC 2023, Singapore, Singapore, Dec 13 2023 - Dec 15 2023 (pp. 2421-2426). Institute of Electrical and Electronics Engineers (IEEE)
Open this publication in new window or tab >>Almost Exact Recovery in Gossip Opinion Dynamics Over Stochastic Block Models
2023 (English)In: 2023 62nd IEEE Conference on Decision and Control, CDC 2023, Institute of Electrical and Electronics Engineers (IEEE) , 2023, p. 2421-2426Conference paper, Published paper (Refereed)
Abstract [en]

We study community detection based on state observations from gossip opinion dynamics over stochastic block models (SBM). It is assumed that a network is generated from a two-community SBM where each agent has a community label and each edge exists with probability depending on its endpoints' labels. A gossip process then evolves over the sampled network. We propose two algorithms to detect the communities out of a single trajectory of the process. It is shown that, when the influence of stubborn agents is small and the link probability within communities is large, an algorithm based on clustering transient agent states can achieve almost exact recovery of the communities. That is, the algorithm can recover all but a vanishing part of community labels with high probability. In contrast, when the influence of stubborn agents is large, another algorithm based on clustering time average of agent states can achieve almost exact recovery. Numerical experiments are given for illustration of the two algorithms and the theoretical results of the paper.

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2023
National Category
Control Engineering
Identifiers
urn:nbn:se:kth:diva-343716 (URN)10.1109/CDC49753.2023.10383465 (DOI)001166433802008 ()2-s2.0-85184803048 (Scopus ID)
Conference
62nd IEEE Conference on Decision and Control, CDC 2023, Singapore, Singapore, Dec 13 2023 - Dec 15 2023
Note

Part of ISBN 9798350301243

QC 20240222

Available from: 2024-02-22 Created: 2024-02-22 Last updated: 2024-04-05Bibliographically approved
Organisations
Identifiers
ORCID iD: ORCID iD iconorcid.org/0000-0001-9940-5929

Search in DiVA

Show all publications