The management of COVID-19 appears to be a long-term challenge, even in countries that have managed suppress the epidemic after their initial outbreak. In this paper, we propose model predictive approach for constrained control nonlinear compartmental captures key dynamical properties COVID-19. design uses discrete-time version model, and it is able handle complex, possibly time-dependent constraints, logical relations between variables multiple predefined discrete levels interventions. A...

The aim of the presented research is to elaborate a traffic-responsive optimal signal split algorithm taking uncertainty into account. traffic control objective minimize weighted link queue lengths within an urban network area. problem formulated in centralized rolling-horizon fashion which unknown but bounded demand and influences prediction. An efficient constrained minimax optimization suggested obtain green time combination, minimizes function when worst case appears. As illustrative...

This article proposes an active-learning-based adaptive trajectory tracking control method for autonomous ground vehicles to compensate modeling errors and unmodeled dynamics. The nominal vehicle model is decoupled into lateral longitudinal subsystems, which are augmented with online Gaussian Processes (GPs), using measurement data. estimated mean functions of the GPs used construct a feedback compensator, which, together LPV state controller designed system, gives structure. To assist...

Deep-learning-based nonlinear system identification has shown the ability to produce reliable and highly accurate models in practice. However, these black-box lack physical interpretability, often a considerable part of learning effort is spent on capturing already expected/known behavior due first-principles-based understanding some aspects system. A potential solution integrate prior knowledge directly into model structure, combining strengths physics-based modeling deep-learning-based...

Unmanned Arial Vehicles (UAVs) require the development of some on-board safety equipments before inheriting sky. An collision avoidance system is being built by our team. Due to strict size, weight, power, and costs constraints, visual intruder airplane detection only option. This paper introduces detector algorithm, which designed be operational in clear cloudy situations under regular daylight conditions. To able implement algorithm on-board, we have carefully selected topographic...

Estimating traffic flow states at unmeasured urban locations provides a cost-efficient solution for many ITS applications. In this work, geostatistical framework, kriging is extended in such way that it can both estimate and predict volume speed various unobserved locations, real-time. the paper, different distance metrics are evaluated. Then, new, data-driven one formulated, capturing similarity of measurement sites. with multidimensional scaling distances transformed into hyperspace, where...

Summary The paper presents a novel model order reduction technique for large‐scale linear parameter‐varying (LPV) systems. approach is based on decoupling the original dynamics into smaller dimensional LPV subsystems that can be independently reduced by methods. decomposition starts with construction of modal transformation separates subsystems. Hierarchical clustering applied then to collect dynamically similar larger groups. resulting are reduced. This substantially differs from most...

Abstract In the paper, an analysis method is applied to lateral stabilization problem of vehicle systems. The aim find largest state-space region in which stability can be guaranteed by peak-bounded control input. analysis, nonlinear polynomial sum-of-squares programming applied. A practical computation technique developed calculate maximum controlled invariant set system. calculates sets steering and braking systems at various velocities road conditions. Illustration examples show that,...

This article proposes two control methods for performing a backflip maneuver with miniature quadcopters. First, an existing feedforward approach is improved by finding the optimal sequence of motion primitives via Bayesian optimization, using surrogate Gaussian process (GP) model. To evaluate cost function, flip performed repeatedly in simulation environment. The second method based on closed-loop and it consists main steps: first, novel robust, adaptive controller designed to provide...

One of the missing critical on-board safety equipment Unmanned Arial Vehicles (UAVs) is collision avoidance system. In 2010 we launched a project to research and develop an SAA system for UAS. As will be in small aircraft have minimize weight, volume, power consumption. The acceptable consumption 1-2W mass control maximum 300-500g. Here present concept visual input based See Avoid (SAA) This paper introduces long range detection algorithm implementation aspect many core processor device.

The present paper investigates the real world feasibility of a purely vision based sense and avoid system, required for small unmanned aerial vehicles (UAV) to routinely access national airspace. two distinct functions, sensing avoidance are integrated into common framework. No information is exchanged between aircraft, only passive 2-D available estimate encountering traffic. Based on predicted intruder motion time encounter minimum distance predicted. In case an violates separation onboard...

An unmanned aerial vehicle (UAV) formation in a leader-follower structure, where the UAVs are flying common trajectory determined by route planner hosted on leader is considered. The path description compressed polynomial functions with respect to flight envelope constraints and transmitted followers, model predictive control (MPC) outer loop controller specifies command signals for 7-h locally controlled dynamics nonlinear of aircraft dynamics. Real time feasibility issues associated design...

This paper investigates the real world feasibility of a purely vision based sense and avoid system, required for small unmanned aerial vehicles (UAV) to routinely access national airspace. No information is exchanged between aircraft, only passive 2-D available estimate encountering tra c. The viability system demonstrated on several estimation approaches, using Extended Kalman lter (EKF) Unscented (UKF) implementations. Since it shown that certain type observer movements process remains...

A multi-level reconfiguration framework is proposed for fault tolerant control of overactuated aerial vehicles, where the levels indicate how much authority given to task. On lowest, first level accommodated by modifying only actuator/sensor configuration, so remains hidden from baseline controller. dynamic reallocation scheme applied on this level. The allocation mechanism exploits redundancy available aircraft. In case cannot be managed at process has access Based LPV framework, done...

Abstract A multi-level reconfiguration framework is proposed for fault tolerant control of over-actuated aerial vehicles, where the levels indicate how much authority given to task. On lowest, first level accommodated by modifying only actuator/sensor configuration, so remains hidden from baseline controller. dynamic reallocation scheme applied on this level. The allocation mechanism exploits redundancy available aircraft. When cannot be managed at level, process has access Based LPV...

A dynamic input reconfiguration architecture is proposed for overactuated aerial vehicles to accommodate actuator failures. The method based on the nullspace computed from linear parameter-varying model of plant dynamics. If there no uncertainty in system, then any signal filtered through has effect outputs. This makes it possible reconfigure inputs without influencing nominal control loop and thus performance. Since allocation mechanism independent structure baseline controller, can be...

The paper presents a systematic design procedure for approximate explicit model predictive control constrained nonlinear systems described in linear parameter-varying (LPV) form. method applies Gaussian process (GP) to learn the optimal policy generated by recently developed fast (MPC) algorithm based on an LPV embedding of system. By exploiting advantages GP structure, various active learning methods information theoretic criteria, gradient analysis and simulation data are combined...

The model reduction problem of high dimensional Linear Parameter Varying (LPV) systems is addressed in the paper. Modal representation local computed first for fixed values scheduling parameter. Modes are then matched and a smooth quasi-modal form obtained over entire parameter domain. Classification system modes applied to explore dynamic coherence model. Structured parameter-varying Gramians constructed used balancing eliminating negligible components. Numerical example illustrates...

In this paper, a predictive-control-based approach is proposed for pandemic mitigation with multiple control inputs. Using previous results on the dynamical modeling of symptom-based testing, testing intensity introduced as new manipulable input to system model in addition stringency non-pharmaceutical measures. The objective minimization severity interventions, while main constraints are bounds daily number hospitalized people and total available tests. For design simulation, nonlinear...