This report deals with the detection and diagnosis of faults (FDD) when they develop in different parts of a wastewater treatment plant, situated in Manresa, Spain. When a fault occurs estimates of parameters in a non...
详细信息
This report deals with the detection and diagnosis of faults (FDD) when they develop in different parts of a wastewater treatment plant, situated in Manresa, Spain. When a fault occurs estimates of parameters in a non-linear mathematical model of the plant change. Some methods for detection and tracking the values of different parameters are proposed, where only large parameter changes are considered in this work. The methods have a common structure with three basic elements: a basic identification algorithm, a detector of the parameter changes and an adaptive procedure for improving the tracking capability of the identifier. A set of simulation experiments are presented in order to compare them.
The development and evaluation of an expert advisor for system identification (EASI) is described. The principles behind a suitable knowledge representation paradigm for a system which includes both intelligent tutori...
详细信息
The development and evaluation of an expert advisor for system identification (EASI) is described. The principles behind a suitable knowledge representation paradigm for a system which includes both intelligent tutoring and software package front-end concepts are outlined. The system has been constructed using an expert system ‘text-animation’ shell with linkage into Prolog-2 to provide enhancements. These extensions have included an online dictionary and bibliographic material. The identification domain aspects covered include experimental design, model structure determination, estimation algorithms, with applications to SISO and MIMO linear systems, and SISO nonlinear dynamics. Different levels of user-model are provided, and the advisor acts as a front-end to a number of in-house identification packages.
Measurements are inevitably bound to experimental errors, whose probability distributions are generally unknown. When models of metabolic-endocrine systems or of drug kinetics are identified starting from noisy data, ...
详细信息
Measurements are inevitably bound to experimental errors, whose probability distributions are generally unknown. When models of metabolic-endocrine systems or of drug kinetics are identified starting from noisy data, the question arises of how the parameters estimates are affected by the assumptions made. Aim of this work is to investigate whether and to what extent unreliable assumptions concerning the noise distribution influence the estimation process and to compare the efficiency of two estimation methods in the case of ‘ad hoc’ simulated data. Al though these are generated from two-compartment systems which can model some aspects of a β-blocking agent kinetics, both linear and nonlinear, it is believed that the findings of this study are of more general interest. The results show that the minimizing algorithms do not significantly differ as for their capability of fitting a bi-exponential curve to the simulated data. Evidence is given of the ambiguities encountered when hypotheses concerning the system structure are tested against the algorithms results. Most of the findings of this work can be expected from estimation theory; however they allow a better insight into the limits of applying estimation methods to biomedical problems, because the systems structure has been drawn from drug kinetics and a large variety of additive noise characteristics has been considered.
This paper presents a new estimation algorithm, which is essentially a modification of the weighted recursive least squares algorithm (W-RLS), for systems with bounded disturbances. Assuming the knowledge of a disturb...
详细信息
This paper presents a new estimation algorithm, which is essentially a modification of the weighted recursive least squares algorithm (W-RLS), for systems with bounded disturbances. Assuming the knowledge of a disturbance upper bound, the algorithm's derivation is performed the minimization of a cost function weighted by two factors : one is fixed by the user and exponentially weights the arrival information, the other is time-varying and data-dependent. The purpose of the latter is to ensure the algorithm stability faced with the bounded disturbances and to enhance the convergence rate, as much as permitted by the available data. The final algorithm has a very compact and simple form. It completely depends upon the significative information and it stops when the arrival data are redundant. After the algorithm derivation, the paper presents the convergence properties, analyses the algorithm behavior faced with time-varying systems and presents some simulations.
The bearings-only estimation of target velocity and position has been reformulated and treated as a combined estimation and identification problem which has been tackled through partitioning. Certain advantages of the...
详细信息
The bearings-only estimation of target velocity and position has been reformulated and treated as a combined estimation and identification problem which has been tackled through partitioning. Certain advantages of the partition estimation algorithm have been exploited to improve convergence. Here a different numerical realization which reduces the amount of computation is presented. Simulation examples are given.
暂无评论