We provide a novel robust stability analysis for moving horizon estimation (MHE) using Lyapunov function. In addition, we introduce linear matrix inequalities (LMIs) to verify the necessary incremental input/output-to-state ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\boldsymbol{\delta }$</tex-math></inline-formula> -IOSS) detectability condition. consider an MHE formulation with time-discounted...

This work presents a fully data-driven, black-box pipeline to obtain an optimal control policy for multi-loop building problem based on historical and weather data, thus without the need complex physics-based modelling. We demonstrate method joint of room temperature bidirectional EV charging maximize occupant thermal comfort energy savings while leaving enough in battery next trip. modelled with recurrent neural network piece-wise linear function. Using these models as simulation...

While distributed algorithms provide advantages for the control of complex large-scale systems by requiring a lower local computational load and less memory, it is challenging task to design high-performance policies. Learning-based offer promising opportunities address this challenge, but generally cannot guarantee safety in terms state input constraint satisfaction. A recently proposed framework centralized linear ensures matching learning-based online with initial model predictive law...

This paper presents a novel model order reduction technique tailored for nonlinear power systems with large share of inverter-based energy resources. Such exhibit an increased level dynamic stiffness compared to traditional systems, posing challenges time-domain simulations and control design. Our approach involves rotation the coordinate system linearized using transformation matrix derived from real Jordan canonical form, leading mode decoupling. The fast modes are then truncated in...

From both an educational and research point of view, experiments on hardware are a key aspect robotics control. In the last decade, many open-source software frameworks for wheeled robots have been presented, mainly in form unicycles car-like robots, with goal making accessible to wider audience support control systems development. Unicycles usually small inexpensive, therefore facilitate larger fleet, but they not suited high-speed motion. Car-like more agile, expensive, thus requiring...

In this paper, we study state estimation for general nonlinear systems with unknown parameters and persistent process measurement noise. particular, are interested in stability properties of the estimate absence persistency excitation (PE). With a simple academic example, show that existing moving horizon (MHE) approaches as well classical adaptive observers can result diverging estimates PE, even if noise is small. We propose novel MHE formulation involving regularization based on constant...

Distributed model predictive control methods for uncertain systems often suffer from considerable conservatism and can tolerate only small uncertainties due to the use of robust formulations that are amenable distributed design optimization methods. In this work, we propose a stochastic (DSMPC) scheme dynamically coupled linear discrete-time subject unbounded additive disturbances potentially correlated in time. An indirect feedback formulation ensures recursive feasibility DSMPC problem,...

We present an output feedback stochastic model predictive control (SMPC) approach for linear systems subject to Gaussian disturbances and measurement noise probabilistic constraints on system states inputs. The presented combines a Kalman filter state estimation with indirect SMPC, which is initialized predicted nominal state, while of the current estimate enters through objective SMPC problem. For this combination, we establish recursive feasibility problem due chosen initialization,...

This paper introduces a novel optimization-based approach for parametric nonlinear system identification. Building upon the prediction error method framework, traditionally used linear identification, we extend its capabilities to systems. The predictions are computed using moving horizon state estimator with constant arrival cost. Eventually, both parameters and cost estimated by minimizing sum of squared errors. Since induced estimator, can be viewed as tuning based on predictive...

This paper presents an open-source miniature car-like robot with low-cost sensing and a pipeline for optimization-based system identification, state estimation, control. The overall robotics platform comes at cost of less than $700 thus significantly simplifies the verification advanced algorithms in realistic setting. We present modified bicycle model Pacejka tire forces to dynamics considered all-wheel drive vehicle prevent singularities low velocities. Furthermore, we provide...

This paper presents a robust moving horizon estimation (MHE) approach with provable error bounds for solving the simultaneous localization and mapping (SLAM) problem. We derive sufficient conditions to guarantee stability in ego-state estimates bounded errors landmark position estimates, even under limited visibility which directly affects overall system detectability. is achieved by decoupling MHE updates positions, enabling individual only when required detectability are met. The decoupled...

To control a dynamical system it is essential to obtain an accurate estimate of the current state based on uncertain sensor measurements and existing knowledge. An optimization-based moving horizon estimation (MHE) approach uses model system, further allows for integration physical constraints states uncertainties, trajectory estimates. In this work, we address problem in case constrained linear systems with parametric uncertainty. The proposed makes use differentiable convex optimization...

From both an educational and research point of view, experiments on hardware are a key aspect robotics control. In the last decade, many open-source software frameworks for wheeled robots have been presented, mainly in form unicycles car-like robots, with goal making accessible to wider audience support control systems development. Unicycles usually small inexpensive, therefore facilitate larger fleet, but they not suited high-speed motion. Car-like more agile, expensive, thus requiring...

This paper presents a novel model order reduction technique tailored for power systems with large share of inverter-based energy resources. Such exhibit an increased level dynamic stiffness compared to traditional systems, posing challenges time-domain simulations and control design. Our approach involves rotation the coordinate system linearized using transformation matrix derived from real Jordan canonical form, leading mode decoupling. The fast modes are then truncated in rotated obtain...

We consider the design of functional estimators, i.e., approaches to compute an estimate a nonlinear function state general dynamical system subject process noise based on noisy output measurements. To this end, we introduce novel detectability notion in form incremental input/output-to-output stability ($\delta$-IOOS). show that $\delta$-IOOS is necessary condition for existence estimator satisfying input-to-output type property. Additionally, prove detectable if and only it admits...

We provide a novel robust stability analysis for moving horizon estimation (MHE) using Lyapunov function. Additionally, we introduce linear matrix inequalities (LMIs) to verify the necessary incremental input/output-to-state ($δ$-IOSS) detectability condition. consider an MHE formulation with time-discounted quadratic objective nonlinear systems admitting exponential $δ$-IOSS show that suitable parameterization of objective, function serves as $M$-step MHE. Provided is chosen large enough,...

We present an output feedback stochastic model predictive control (SMPC) approach for linear systems subject to Gaussian disturbances and measurement noise probabilistic constraints on system states inputs. The presented combines a Kalman filter state estimation with indirect SMPC, which is initialized predicted nominal state, while of the current estimate enters through objective SMPC problem. For this combination, we establish recursive feasibility problem due chosen initialization,...