Teresa Alvarez

Model-free control based on reinforcement learning for a wastewater treatment problem

This article presents a proposal, based on the model-free learning control (MFLC) approach, for the control of the advanced oxidation process in wastewater plants. This is prompted by the fact that many organic pollutants in industrial wastewaters are resistant to conventional biological treatments, and the fact that advanced oxidation processes, controlled with learning controllers measuring the oxidation-reduction potential (ORP), give a cost-effective solution. The proposed automation strategy denoted MFLC-MSA is based on the integration of reinforcement learning with multiple step actions. This enables the most adequate control strategy to be learned directly from the process response to selected control inputs. Thus, the proposed methodology is satisfactory for oxidation processes of wastewater treatment plants, where the development of an adequate model for control design is usually too costly. The algorithm proposed has been tested in a lab pilot plant, where phenolic wastewater is oxidized to carboxylic acids and carbon dioxide. The obtained experimental results show that the proposed MFLC-MSA strategy can achieve good performance to guarantee on-specification discharge at maximum degradation rate using readily available measurements such as pH and ORP, inferential measurements of oxidation kinetics and peroxide consumption, respectively.

Control of neutralization processes by robust loop shaping

This paper studies the control of a pH neutralization process by using a robust loop shaping approach. The H/sub /spl infin// loop shaping method is applied to calculate an optimal controller. The problem of choosing the desired shape of the open-loop transfer function needed by this method is addressed by considering the available uncertainty information and applying graphical loop shaping ideas. Thus, this methodology considers not only the robustness properties of the shaped plant, but also those of the real plant. The designed controller was tested in real-time on a bench plant. Online results show that the designed control system allows the plant to operate in a range of pH values, despite variations of the plant parameters, obtaining good performance at the desired working points. The methodology presented can be applied to other chemical processes: it is only necessary to consider the possible uncertainty in the nominal model and using available software to design the controller.

Handling infeasibilities in predictive control

During the last years predictive control has received an increasing attention from industry. One of the reasons is that it takes into account the process constraints in a natural way. Nevertheless, there are situations (perturbations, not well defined constraints, etc.) when it is not possible to compute a sequence of future controls such that all the constraints are satisfied, i.e., the problem is not feasible. When this sort of problem appears it is necessary to apply some infeasibility handling procedure that drives the problem to a feasible region. After reviewing briefly some of the different approaches found in the literature, this paper presents a new method for solving the infeasibilities considering a constrained MIMO GPC based controller. The feasibility is recovered applying different techniques or a combination of all of them and the constraints changes are minimised according to a certain criteria. Finally, some computational results are shown.

Predictive control of robotic manipulators

Robot manipulators present restrictions on their performance, such as the maximum torque the motors can apply, limitations in their position, speed, acceleration, etc. This paper studies the application of a multivariable constrained predictive controller for robotic control, that considers these restrictions when calculating the control signal. The model used to calculate predictions is evaluated at every sample-time by linearization of the Denavit-Hartenberg model, and discretization using a bilinear transform. By using simulation, this paper presents the successful application of this technique to a direct-drive manipulator.

Identification and constrained multivariable predictive control of chemical reactors

The basic idea of this paper is to present the identification and constrained multivariable predictive control of an industrial chemical reactor. In the literature, there are numerous papers that discuss and investigate the design and control of chemical reactors. Most of them emphasize steady-state aspects or stability questions. Although, only a few of them present results and conclusions that have application in the design and control of industrial reactors and none of the articles applies predictive controllers or bounded-parameters models. During the last few years predictive control has received an increasing attention from industry because this technique gives understandable and intuitive solutions to industrial problems. Predictive controllers are model based, that is, one needs an explicit model of the system in order to perform the control. So the paper is divided in two parts: the chemical reactor identification and the control strategy.

Robust ∞ Filtering of 2D Roesser Discrete Systems: A Polynomial Approach

The problem of robust 𝐻∞ filtering is investigated for the class of uncertain two-dimensional (2D) discrete systems described by a Roesser state-space model. The main contribution is a systematic procedure for generating conditions for the existence of a 2D discrete filter such that, for all admissible uncertainties, the error system is asymptotically stable, and the 𝐻∞ norm of the transfer function from the noise signal to the estimation error is below a prespecified level. These conditions are expressed as parameter-dependent linear matrix inequalities. Using homogeneous polynomially parameter-dependent filters of arbitrary degree on the uncertain parameters, the proposed method extends previous results in the quadratic framework and the linearly parameter-dependent framework, thus reducing its conservatism. Performance of the proposed method, in comparison with that of existing methods, is illustrated by two examples.

Constrained predictive control of multivariable robotic systems

Robotic systems present physical limitations in the maximum torque that the motors can apply and its variation rate. Also, due to safety regulations, there are strict limitations in the maximum speed and acceleration that any link can operate. To solve these problems online the paper proposes the implementation of a multivariable constrained predictive controller. These limitations are usually very strict, hence sometimes it may not be possible to calculate an adequate sequence of the future controls due to constraint incompatibility, and the optimization problem is infeasible. To solve this difficulty an infeasibility solver can be included in the predictive controller.

Robust H ∞ filtering for uncertain 2-D continuous systems, based on a polynomially parameter-dependent Lyapunov function

The problem of robust H∞ filtering for uncertain two-dimensional (2-D) continuous systems described by the Roesser state-space model is investigated when the parameter uncertainties are polytopic. A sufficient linear matrix inequality (LMI) condition for the existence of a 2-D continuous filter such that, for all admissible uncertainties, the error system is asymptotically stable, and the H∞ norm of the transfer function from the noise signal to the estimation error is below a prespecified level. A sequence of standard LMI conditions that ensure the existence of homogeneous polynomially parameter-dependent (HPPD) matrices of arbitrary degree, that are solutions to the parameter-dependent LMIs is provided in terms of the vertices of the polytope. The proposed method includes results in the quadratic framework and the linearly parameter-dependent framework as special cases. Finally, an example is provided to demonstrate the effectiveness and advantages of the proposed filter design methodology.

How to reduce congestion on TCP/AQM networks with simple adaptive PID controllers

Congestion is a problem in real networks. Users do not want to lose information and data should be delivered as fast and reliably as possible. This is really difficult to achieve. Moreover, networks work in changing environments: number of users, type of traffic, delays in transmission, etc. So this paper presents how to design PID controllers that take network changes into account. Non-linear simulations using ns-2 will show the goodness of the approach when compared with classical PID, drop tail and RED.

Robust stabilization using a sampled-data strategy of uncertain neutral state-delayed systems subject to input limitations

Abstract Stabilization of neutral systems with state delay is considered in the presence of uncertainty and input limitations in magnitude. The proposed solution is based on simultaneously characterizing a set of stabilizing controllers and the associated admissible initial conditions through the use of a free weighting matrix approach. From this mathematical characterization, state feedback gains that ensure a large set of admissible initial conditions are calculated by solving an optimization problem with LMI constraints. Some examples are presented to compare the results with previous approaches in the literature.