H. Lopez

Specification, design and Modula-2 implementation of a low cost industrial control system

Abstract The purpose of this paper is to show the application of modern software engineering techniques to the construction of a small/medium size package for real time control. This package, denoted by SCM (acronym of “Sistema de Control Modular”), has been written in Modula-2 and presents all advantages offered by this high-level structured programming language. It has been implemented on a PC-AT compatible computer.

GA-P Based Search of Structures and Parameters of Dynamical Process Models

The most effective approaches for evolutionary identifying dynamical processes depend on iterative trial-error searches in a hierarchical fashion: a new structure is proposed first; then, its set of parameters is numerically determined, and the process is repeated until a model accurate enough is found.

Different Control Strategy Results of an Hydro-Power Plant, using Ada

In the present paper, we show how we have implemented, using Ada language programming, different control strategies based on pole assignment and optimal criterions, for an hydro-power plant. We have solved the servodesign problem either by a feedforward control or by handling errors. An adaptive control procedure based in an I/O minimal structure is proposed, with the robustness of having a twenty four hours continuously performance with the object of being applicable industrially without the switch off from the adaptive loop. This control strategies can be used in the case we are presenting, replacing some techniques classier already proved.

A User-Customizable Supervision Level for Improving the Applicability of Explicit Self-Tuning Controllers

Abstract This paper is devoted to the analysis, development and evaluation of protection mechanisms, which have to be incorporated in self-tuning controllers so that they can operate correctly in industrial applications (section 1). The analysis begins with the definition of the problems which arise from the industrial utilization of this kind of controllers and the extension of the general loop structure of a self-tuning controller in order to incorporate the mentioned protection mechanisms (section 2). Next, a detailed study of the protection functions -supervisions-, is accomplished. The first one is the supervision of the start-up phase (section 3). Subsequently, the supervision of the parameter estimator is analized (section 4), carrying out a comparative study of the methods available at present, in order to preserve the estimator integrity. In the synthesis step supervision (section 5), the fiability of the new estimated model is analized, determining if it is adequate to synthesize a new controller. This leads to cautious controllers. It is also verified if the control law leads to an instable loop when it is combined with the estimated model (section 6). To achieve this, the behaviour of the loop’s signals, u(t) and y(t), is supervised. Next (section 7), the complete configuration selected for the supervision level is summarized. In order to show the effectiveness of this supervisor, several illustrative examples of its operation are presented. Finally (section 8), the conclusions obtained from this work are resumed.

Temperature monitoring system in the mould of a slab continuous casting line

In this article a study is introduced that was carried out for the implementation of a temperature monitoring system in the mould of a slab continuous casting line in ACERALIAs LD3 steel factory in Aviles (Asturias). To achieve this, instrumentation has been proposed consisting on precision thermocouples placed along the vertical mid-line of both the broad face and the narrow face of the mould. Signals are converted and sent through an industrial bus to the acquisition station. Here, the data from the process computer (the conditions under which the casting develops) is also stored. The ultimate objective is the retrieval of actual data on temperatures at specific locations of the mould. These data can be used for the adjustment of models of mould operative behaviour.

Control of coating thickness in a tinplate line

Abstract In a tinplate line the coating thickness can be controlled by means of the line speed and current applied. A control strategy has been developed that calculates the maximum possible speed and the total necessary current, distributing it into the available rectifiers. The equations used for the calculations are based on a mathematical model obtained using Genetic Programming.

A Dynamic Path Planner for Autonomous Vehicles

This paper proposes an architecture and a method for the design of a dynamic path planner which utilizes fuzzy logic for navigation of mobile robots in uncertain environments. The method consists of identifying a fuzzy mapping from examples generated by an expert, a model-based path planner. This planner maps the ultrasonic sensor readings and odometer measures to the robot control commands. The correct mapping is found by identifying a set of fuzzy rules from desired input-output data pairs. The effectiveness of the proposed method is verified by a series of simulations.

Implementation and Design of State Space Self Tuning Regulator for Speed Control Using Ada

Abstract In the present paper, we show how we have implemented, using Ada language programming, a motion control system of electric machines, for instance like an electric engine and others industrial systems more complex like an hydro-power plant. We combine an half optimal regulator with different methods of estimating the state variables, as much as static form as dynamic one. We have solved the servodesign problem or by a feedforward control or by handling errors. An adaptive control procedure based in an I/O minimal structure is proposed for deterministic and stochastic cases, with the robustness of having a twenty four hours continuously performance with the object of being applicable industrially without the switch off the adaptive loop. This control strategies can be used in the cases we present replacing some techniques classier already proved.

Development of Robust Adaptive Estimators for On-line Industrial Utilization Analysis of Available Techniques and Presentation of New Solutions

Abstract In this paper, the problems originated by the real-time estimation of time- varying system parameters are analyzed so that the requirements to be verified by an online industrial estimator of this kind can be defined. Next, a critical analysis of the more usual adaptive estimation techniques is accomplished, emphasizing three basic aspects: the operating security of the method, its capability of adaptation, its use facility and its computational cost. The recursive estimators are studied so that they can verify the two contradictory objectives of an accurate tracking of varying parameters and avoiding the blow-up in absence of excitation. for this extensive study, the methods have been grouped in the following way: – Scalar exponential forgetting (Astrom, Fortescue, Well stead, Irving) – Vectorial exponential forgetting (Saelid, Tuffs) – Direct modification of the covariance matrix (Goodwing) – Utilization of finite-data windows (Goodwing, Young) – Switching on/off the estimator (Isermann, Radke, Lachmann) – Directional exponential forgetting (Kulhavy) – Variable directional exponential forgetting (Hagglund, Kulhavy, Bertin) Finally, based on the previous analysis, new methods, which verify all previous requirements, are developed. Basically, the new methods are based on vectorial exponential forgetting factors or direct modification of the covariance matrix, but incorporating an explicit control-loop for the elements of the covariance matrix. When the excitation is high, the new estimators allow the covariance matrix to take reduced values performing the minimum-quadratic estimator properly, presenting its typical high speed of convergence. However, if the excitation is reduced, these new estimators control the covariance matrix. Then the minimum-quadratic estimators perform as one of gradient type, with lower speed of convergence but avoiding the blow-up problem.

GA-P BASED MODELING OF NONLINEAR DYNAMIC SYSTEMS

Abstract Model construction is usually guided by a trial-error process, where each iteration is divided into two steps: (i) physical modeling and (ii) identification. Genetic programming has been applied to automate this process in different ways. One of the most complete approaches is the described in project SMOG, where a set of model structures is evolved, being the set of parameters of each model optimized by means of classical methods. In this paper, a GA-P algorithm (a hybrid between genetic algorithms and genetic programming) is applied to the task permitting the evolution in parallel of model structures and parameters.