The gap between the models used in control synthesis and those obtained from identification experiments is considered by investigating connection uncertain data. model validation problem addressed is: given experimental data a with both additive noise norm-bounded perturbations, it possible that could produce observed input-output data? This studied for standard H/sub infinity // mu framework models. A necessary condition such to describe an datum obtained. For large class of robust...

This paper reports the final results of predictive building control project OptiControl-II that encompassed seven months model (MPC) a fully occupied Swiss office building. First, this provides comprehensive literature review experimental MPC studies. Second, we describe chosen setup and modeling, main results, as well simulation-based comparisons to industry-standard using EnergyPlus simulation software. Third, costs benefits for cases similar investigated are analyzed. In experiments,...

The increasing availability of Internet connectivity and networked actuation sensing components has supported the growth in control systems operated over public networks. Controllers plants no longer need to be physically colocated as measurements signals can sent digitally. Supervisory monitor geographically widespread components. However, such are now exposed risk remote interference. A feedback structure that allows an attacker take plant while remaining hidden from supervisory system(s)...

Because physics-based building models are difficult to obtain as each is individual, there an increasing interest in generating suitable for MPC directly from measurement data. Machine learning methods have been widely applied this problem and validated mostly simulation; are, however, few studies on a direct comparison of different or validation real buildings be found the literature. Methods that indeed application often lead computationally complex non-convex optimization problems. Here...

In this paper, we present an impulse response identification scheme that incorporates the internal positivity side-information of system. The realization theory positive systems establishes specific criteria for existence a given transfer function. These function are translated to set suitable conditions on shape and structure responses systems. Utilizing these conditions, estimation problem is formulated as constrained optimization in reproducing kernel Hilbert space equipped with stable...

This paper considers control design for systems with input magnitude saturation. Four examples, 2 SISO and MIMO, are used to illustrate the properties of several existing schemes. A new method based on a modification conventional antiwindup compensation is introduced. It assumed that reader familiar problem integral windup saturating plants schemes dealing it.

Model predictive control (MPC) is a promising alternative in building with the potential to improve energy efficiency and comfort enable demand response capabilities. Creating an accurate model that simple enough allow resulting MPC problem be tractable challenging but crucial task development. In this paper we introduce Building Resistance-Capacitance Modeling (BRCM) Matlab Toolbox facilitates physical modeling of buildings for MPC. The provides means fast generation (bi-)linear...

Recently, various algorithms for data-driven simulation and control have been proposed based on the Willems' fundamental lemma. However, when collected data are noisy, these methods lead to ill-conditioned model structures. In this work, we present a maximum likelihood framework obtain an optimal model, signal matrix in presence of output noise. Data compression noise level estimation schemes also apply algorithm efficiently large datasets unknown scenarios. Two approaches system...

We propose a performance-based autotuning method for cascade control systems, where the parameters of linear axis drive motion controller from two loops are tuned jointly. Using Bayesian optimization as all simultaneously, is guaranteed to converge asymptotically global optimum cost. The data-efficiency and performance studied numerically several training configurations compared those achieved with classical tuning methods exhaustive evaluation On real system, tracking robustness against...

This article presents an automated, model-free, data-driven method for the safe tuning of PID cascade controller gains based on Bayesian optimization. The optimization objective is composed performance metrics and modeled using Gaussian processes. safety requirement imposed via a barrier-like term in objective, which introduced to account operational changes system. We further introduce constraint that captures stability requirements from system data. Numerical evaluation shows proposed...

Model predictive control (MPC) strategies can be applied to the coordination of energy hubs reduce their consumption. Despite effectiveness these techniques, potential for savings are potentially underutilized due fact that demands often assumed fixed quantities rather than controlled dynamic variables. The joint optimization and buildings' management systems result in higher savings. This paper investigates how different MPC perform on buildings hubs. We first discuss two approaches;...

The control of cooperative formations vehicles can be based on parallel estimation, where each vehicle determines its action from a locally maintained estimate the entire observable formation state. Vehicles may communicate with one another allowing local estimates to incorporate information other estimators in formation. This paper studies dynamics that arise this situation and provides complete analysis stability for class decentralized problems. In absence communication,...

Spacecraft formations in deep space can be specified terms of the relative spacecraft positions and absolute orientations.Accurate position measurements are not available a formation control design approach, based on information, is presented.This approach exploits redundancy inherent specification to develop family equivalent topologies for optimal control.Switching between provides allows combinations measured communicated estimates used implementation control.Redundancy may also exist...

This paper addresses the setup of an aggregation office buildings to provide services for electricity grid. In order determine energy shifting potential different a given time day, standardized assessment procedure based on Model Predictive Control and predefined price signals is proposed. The aim aggregator with index describing additional used desired change in power consumption each building hour enabling him choose good portfolio providing grid service cost-effectively. We first show...

Subspace identification techniques have gained widespread acceptance as a method of obtaining low-order model from data. These are based on using the singular-value decomposition means estimating underlying system order and extracting basis for extended observability space. In presence noise rank determination becomes difficult low estimates lose structure required exact realizability. Furthermore corrupts singular values in manner that is inconsistent with physical processes. problems...

The newly released KUKA LBR iiwa 14 R820 robot stands for intelligent industrial work assistant (iiwa) and is, like its predecessor IV, equipped with torque sensors in each joint, can be controlled through a real-time interface. Although the dynamic model of is not published by manufacturer, knowledge indispensable simulation control based on system model. This paper presents identification minimal set base parameters, as well consistent physical parameters rigid-link robot, including...

We provide convex necessary and sufficient conditions for the robust stability of linear positively dominated systems. In particular, we show that structured singular value is always equal to its upper bound nonnegative matrices use this result derive Linear Matrix Inequality (LMI) involve only system's static gain. how approach can be applied test Foschini-Miljanic algorithm power control in wireless networks presence uncertain interference.

Model predictive control (MPC) techniques are considered for industrial centrifugal compression systems with nonlinear dynamics, to address process and antisurge reaching the desired pressure ratio surge distance. We consider a contractive MPC formulation that ensures asymptotic stability of closed-loop system by imposing decrease quadratic Lyapunov function via an additional constraint. discuss recursive feasibility estimate region attraction numerical methods. also alternative...

The cooperative energy management of aggregated buildings has recently received a great deal interest due to substantial potential savings. These gains are mainly obtained in two ways: 1) exploiting the load shifting capabilities and 2) utilizing expensive but energy-efficient equipment that is commonly shared by building community (e.g., heat pumps, batteries, photovoltaics). Several deterministic stochastic control schemes strive realize these savings have been proposed literature. A...

The increasing adoption of electric vehicles (EVs) and the related need for efficient battery charging leads to additional challenges power network energy providers. One main issues regards intrinsic uncertainty affecting EV process, which calls appropriate strategies ensure reliable solutions. From perspective a parking lot, vehicle process should be managed in order guarantee recharge at competitive price. In this paper, problem pricing under is addressed. Specifically, we propose new...

This paper describes results in applying robust control techniques to achieve vibration suppression of an active precision truss structure. The structure incorporates piezoelectric members which serve as both structural and actuation elements. problem considered is multiple-input, multiple-output with non-collocated actuators sensors. Several characterizations uncertainty are studied the resulting controllers compared experimentally. One characterization uses a novel approach involving...