Filetype
Many of the classical approaches to controller synthesis do not scale well for large and complex systems. This is mainly due to computational complexity and the lack of distributed structure in the resulting controllers. It is important that limitations on the information given and processed by sensors and actuators can be incorporated into the design procedure. However, such constraints may great
We consider a real-time system of multiple tasks, each task having a plant to control. The overall quadratic control cost is to be optimized. We exploit the periodicity of the task response time, which corresponds to a periodic delay pattern in the feedback control loop. Perturbed periods are used as a tool to find a finite hyperperiod. We present an analytical procedure to design a periodic linea
This paper presents a straightforward method to convert non-uniformly sampled data to uniform samples in order to be processed by some off-line techniques such as system identification. In such scenarios, we deal with a finite-length measurement sequence. We assume periodic extension of the signals, which results in a simple system of linear equations. Furthermore, we take into account a number of
The relay autotuner provides a simple way of finding PID controller parameters. Even though relay autotuning is much investigated in the literature, the practical aspects are not that well-documented. In this paper an asymmetric relay autotuner with features such as a startup procedure and adaptive relay amplitudes is proposed. Parameter choices and handling of noise, disturbances, start in non-st
Low rank approximation is an important tool in many applications. Given an observed matrix with elements corrupted by Gaussian noise it is possible to find the best approximating matrix of a given rank through singular value decomposition. However, due to the non-convexity of the formulation it is not possible to incorporate any additional knowledge of the sought matrix without resorting to heuris
Many popular first order algorithms for convex optimization, such as forward-backward splitting, Douglas-Rachford splitting, and the alternating direction method of multipliers (ADMM), can be formulated as averaged iteration of a nonexpansive mapping. In this paper we propose a line search for averaged iteration that preserves the theoretical convergence guarantee, while often accelerating practic
This paper presents a convex-optimization-based technique to obtain parameters for a PID feedback controller, used to control the infusion rate of the anesthetic drug propofol. The controller design is based on a set of identified patient models, relating propofol infusion to an EEG-based conciousness index. The main contribution lies in the method automatically taking inter-patient variability in
Numerous auto-scaling strategies have been proposed in the past few years for improving various Quality of Service (QoS) indicators of cloud applications, for example, response time and throughput, by adapting the amount of resources assigned to the application to meet the workload demand. However, the evaluation of a proposed auto-scaler is usually achieved through experiments under specific cond
Nonlinear Model Predictive Control (NMPC) is a control strategy based on repeatedly solving an optimal control problem. In this paper we present a new MPC framework for the JModelica.org platform, developed specifically for use in NMPC schemes. The new framework utilizes the fact that the optimal control problem to be solved does not change between solutions, thus decreasing the computation time n
This paper presents a hierarchical approach to feedback-based trajectory generation for improved vehicle autonomy. Hierarchical vehicle-control structures have been used before—for example, in electronic stability control systems, where a low-level control loop tracks high-level references. Here, the control structure includes a nonlinear vehicle model already at the high level to generate optimiz
As automatic sensing and Information and Communication Technology (ICT) get cheaper, building monitoring data is easier to obtain. The abundance of data leads to new opportunities in the context of energy efficiency in buildings. This paper describes ongoing developments and first results of data-driven grey-box modelling for buildings. A Python toolbox is developed based on a Modelica library wit
Dynamic optimization problems involving differential-algebraic equation (DAE) systems are traditionally solved while retaining the semi-explicit or implicit form of the DAE. We instead consider symbolically transforming the DAE into an ordinary differential equation (ODE) before solving the optimization problem using a collocation method. We present a method for achieving this, which handles DAE-c
The solution of generic dynamic optimization problems described by Modelica, and its extension Optimica, code using direct collocation methods is discussed. We start by providing a description of dynamic optimization problems in general and how to solve them by means of direct collocation. Next, an existing implementation of a collocation algorithm in JModelica.org, using CasADi and IPOPT, is pres
Dynamic optimization problems involving parametric sensitivities, such as optimal experimental design, are typically solved using shooting-based methods, while leveraging numerical integrators with sensitivity computation capabilities. In this paper we present how simultaneous discretization can be employed to solve these problems, by augmenting the dynamic optimization problems with forward sensi
The position and orientation estimation problem for mobile robots is approached by fusing measurements from inertial sensors, wheel encoders, and a camera. The sensor fusion approach is based on the standard extended Kalman filter, which is modified to handle measurements from the camera with unknown prior delay. A real-time implementation is done on a four-wheeled omni-directional mobile robot, u
Managing the resources of a virtualized data-center is a key issue in cloud computing. Existing research mostly assumes that applications are either allocated the required resources or fail. Combined with the fact that most cloud applications have dynamic resource requirements, this imposes a fundamental limitation to cloud computing: To guarantee on-demand resource allocations, the data-center ne
Traditionally, control loops are closed using periodic sensing and actuation. When communication resources are scarce, however, much may be gained from transmitting only when something important has happened in the loop. This paper presents a simple model of the interference between event-based control loops caused by sharing a common medium, based on the coupled dynamics of a Markov chain represe
With larger penetrations of wind power, the uncertainty increases in power systems operations. The wind power forecast errors must be accounted for by adapting existing operating tools or designing new ones. A switch from the deterministic framework used today to a probabilistic one has been advocated. This two-part paper presents a framework for risk-based operations of power systems. This framew
