In this paper, we develop tools to analyze stability properties of discrete-time switched linear systems driven by switching signals belonging to a given ω- regular language. More precisely, we assume switching signals to be generated by a Büchi automaton where the alphabet corresponds to the modes of the switched system. We define notions of attractivity and uniform stability for this type of systems and also of uniform exponential stability when the considered Büchi automaton is deterministic. We then provide sufficient conditions to check these properties using Lyapunov and automata theoretic techniques. For a subclass of such systems with invertible matrices, we show that these conditions are also necessary. We finally show an example of application in the context of synchronization of oscillators over a communication network.

Digital platforms could facilitate meditation practice by discriminating the participants’ mental state in real time based on neural activities. However, the search for neural correlates of meditative states yields contradictory results in the literature. To identify the neural signature of meditation, we propose a Random Forest classifier to discriminate between a Focused Attention Meditation (FAM) and a problem-solving task, based on two-second samples of EEG data. Two types of classifiers are considered: individual classifiers, trained on EEG data from the considered subject, and general classifiers, trained on inter-individual data. Our results show that the individual classifiers achieve superior performance with an average accuracy of 93% over 14 subjects. The general classifiers display a lower accuracy (74% and 54% depending on whether the data from the tested subject was included in the training set). This study suggests that automatic detection of meditative processes greatly benefits from intra-personal training. The most discriminating EEG features between the two tasks are the Beta mean band amplitude and the Theta-Gamma phase-amplitude coupling, particularly in the occipital and left centro-temporal brain regions. Our findings favor personalized classifiers for FAM.

We consider a set of control tasks sharing a CPU and having stochastic execution requirements. Each task is associated with a deadline: when this constraint is violated the particular execution is dropped. Different choices of the scheduling parameters correspond to a different probability of deadline violation, which can be translated into a different level for the Quality of Control experienced by the feedback loop. For a particular choice of the metric quantifying the global QoC, we show how to find the optimal choice of the scheduling parameters.

In this paper, we propose two adaptive scheduling approaches to support real-time control applications with highly varying computation times. The use of a resource reservation scheduler enables the construction of a dynamic model describing the evolution of the computing delays, which can be incorporated in the system closed loop dynamics. The two approaches differ for the assumptions on the sequence of computation time. In the first approach, we have only an aggregate information (best case and worst case computation time) and design an adaptive scheduler that maintains the delay within the maximum bound compatible with the asymptotic stability of the system. In the second case, we assume a deeper knowledge on the distribution of the computation time and design an adaptive scheduler that ensures second moment stability of the system. The two approaches are evaluated on a case study exposing the different trade-offs between bandwidth and performance.

In this paper, we present a methodology for designing embedded controllers based on the so-called anytime control paradigm. A control law is split into a sequence of subroutine calls, each one fulfilling a control goal and refining the result produced by the previous one. We propose a design methodology to define a feedback controller structured in accordance with this paradigm and show how a switching policy of selecting the controller subroutines can be designed that provides stability guarantees for the closed-loop system. The cornerstone of this construction is a stochastic model describing the probability of executing, in each activation of the controller, the different subroutines. We show how this model can be constructed for realistic real-time task sets and provide an experimental validation of the approach.

In this paper we present a methodology for designing embedded controllers with a variable accuracy. The adopted paradigm is the so called any-time control, which derives from the computing paradigm known as "imprecise computation". The most relevant contributions of the paper are a procedure for designing an incremental control law, whose different pieces cater for increasingly aggressive control requirements, and a modelling technique for the execution platform that allows us to design provably correct switching policies for the controllers. The methodology is validated by both simulations and experimental results.

This paper presents a correct solution to the optimal visual feedback control for a nonholonomic vehicle with limited field-of-view. Previous work on this subject has shown that the search for a shortest path can be limited to simple families of trajectories. We preliminarily provide an extension of the alphabet of optimal control words, to cover some regions of the vehicle plane where the synthesis of turns out to be suboptimal. The main contribution of this paper is an algorithm to translate the optimal synthesis to the image plane, thus enabling a purely image-based optimal control scheme. This allows better performance and increases the robustness of the overall process, avoiding the need of slowly-converging and error-prone parameter estimation algorithms. Simulations and experiments are reported which demonstrate the effectiveness of the proposed technique.

In this paper we address the problem of generating input plans to steer complex dynamical systems in an obstacle-free environment. Plans considered admit a finite description length and are constructed by words on an alphabet of input symbols, which could be e.g. transmitted through a limited capacity channel to a remote system, where they can be decoded in suitable control actions. We show that, by suitable choice of the control encoding, finite plans can be efficiently built for a wide class of dynamical systems, computing arbitrarily close approximations of a desired equilibrium in polynomial time. Moreover, we illustrate by simulations the power of the proposed method, solving the steering problem for an example in the class of underactuated systems, which have attracted wide attention in the recent literature