Ernesto Granado

A web-based virtual laboratory for teaching automatic control

This paper describes a web‐based virtual laboratory that improves the automatic control teaching of undergraduate courses. This is achieved by optimizing the time the students spend in the real laboratory. The virtual world or three‐dimensional scenes of real laboratory plants gives the students the opportunity to “learning by doing” through web‐based systems at the location they want, and at the time they like. This virtual laboratory gives the students the tools to be more prepared and therefore, use the time more efficiently and effectively when working in the real laboratory plant. The student does not require the installation of special software on his remote computer; only need a common VRML enabled browser. An example of a developed virtual laboratory of interconnected tanks is presented.

Remote Experimentation Using Mobile Technology

This paper shows remote experimentation using mobile technology for didactic purposes. Students using WiFi-enabled mobile devices, such as smartphones, tablets or PDAs can interact with physical laboratory hardware. Through a mobile device interface, in real time the user can verify whats happening on laboratory hardware when they change the system parameters. Additionally, the application provides the students a tool for taking advantage of their spare time at the University. Students can perform the lab practices even when the lab facilities are closed. All user activities are stored in a database for later analysis carry out by the professor. Also, students can store their experiments when is finished in order to make an off-line result analysis.

Design and Implementation of a Predictive Control strategy based on an Industrial Controller

This paper describes the design and implementation of a model-based predictive control strategy based on a commercial, high performance industrial controller, which hardware includes a FPGA device. The implemented algorithm is the result of a combination of conventional processor software and FPGA embedded software in a distributed fashion, to achieve lower computational time compared with conventional processor software only approach. The high computational load required by the predictive control strategy is performed into the FPGA, which acts as an accelerator of the conventional processor. All software is developed using a high-level graphics language, yielding a significant reduction on implementation times. The paper includes a numerical example which verifies the controller performance.

Integrated System Design with PID Controllers via LMIs

Integrated design is based on the idea of including dynamic constraints such as stability or controllability, on the early stages of process design. Hence not only the system parameters are calculated but also a controller. In this work we present a methodology to integrated design that allows to compute the optimal parameters of the system, along with its controller, in this case a PID. The system to design may be non-linear with non-linear constraints. To compute the controller we use LMIs and the controller calculated is robust since some uncertainty may be explicitly considered on the plant model. The methodology is applied to a non-linear hydraulic system where several case studies are performed and evaluated

LMI BASED MPC

Abstract In this work, we present a Model Predictive Controller (MPC) based on Linear Matrix Inequalities (LMIs). As in the standard MPC algorithms, at each (sampling) time, a convex optimization problem is solved to compute the control law. The optimization involves constraints written as LMIs, including those normally associated to MPC problems, such as input and output limits. Even though a state space representation is used, only the measurable output and some statistic properties of the not measurable states are used to determine the controller, hence it is an output feedback control design method. Stability of the closed loop system is demonstrated. The design technic is illustrated with a numerical example.

PID-Control Optimization Using Solis-Wets Algorithm And Laguerre System Identification

In this paper, a design method for PID controllers is proposed, based on the Solis-Wets random search algorithm and closed-loop Laguerre identification. The method is tested considering two sets of benchmark plants, previously considered by many other authors. Results show that this method is able to decrease the regulation cost function value considered, while decreasing the rising and settling time and increasing the frequency of the response when compared with an initial controller. The statistic values associated with the method show that it is not affected in a significantly way by the initial values, system order and time constant assumed in its implementation.

Robust Multivariable PID Design via LMI Approach

Abstract In this work we present the design of a robust multivariable PID controller which guarantees the stability of the closed loop systems subject to polytopic uncertainty. The algorithm is based on an iterative linear matrix inequality (ILMI) approach. Also the result may be used to compute a static output feedback stabilizing controller. The design technique is illustrated with a numerical example.

Synthesis of dynamic feedback controllers for uncertain linear systems

An algorithm for the synthesis of dynamic feedback controllers, using linear programming, is presented. The uncertainty considered is of the parametric type and is present in all the system matrices (A,B,C). The proposed algorithm is based on the Lyapunov asymptotic stability condition, which being linear in both the controller and the Lyapunov candidate, allows a linear programming formulation of the problem. The algorithm maximizes locally the uncertainty which might be effectively stabilized. The controllers so obtained are of full order.

Synthesis of Feedback Controllers for Uncertain Discrete Linear Systems by Linear Programming

Abstract In this note an algorithm for the synthesis of output feedback controllers for uncertain discrete linear systems is presented. The uncertainty considered is of polyhedric type and is present in all the system matrices (A,B,C). The proposed algorithm is based on the Lyapunov asymptotic stability condition, which being convex in the Lyapunov candidate matrix or in the controller matrix, allows a linear programming formulation of the problem. The algorithm maximizes locally the uncertainty which might be effectively stabilized. The controllers obtained are of full order.

An approach to develop a virtual 3D stereoscopic environment for smartphones

This paper presents an approach to develop an immersive virtual three-dimensional stereoscopic environment for control education. Thanks to the technological advances and the actual continuous reduction costs on electronic devices, it is possible to create immersive 3D virtual scenarios applying the side-by-side technique; and using standard gadgets like: smartphones, and head-mounted displays. Due to the smartphone built-in accelerometer, the user can easily and intuitively control the movements around the virtual world. Also, the Wii-remote® is used as an auxiliary input device to interact with the application. The proposed method permits to interact and visualize three-dimensional virtual scenarios without the need of specialized software or browser-plugins installed in the smartphone. By taking advantage of todays smartphone technological features, students feel highly motivated to use this device as an amusing learning tool. Since this application resembles three-dimensional computer games, it attracts students, increases their concentration, and keeps their attention during the learning process.