Real-time Parameter Identification Research Articles

Scan Welding: Thermal Modeling and Control of Material Processing

This article provides a thermal analysis of scan welding, as a redesign of classical joining methods, employing computer technology to ensure the composite morphologic, material and mechanical integrity of the joint. This is obtained by real-time control of the welding temperature field by a proper dynamic heat input distribution on the weld surface. This distribution is implemented in scan welding by a single torch, sweeping the joint surface by a controlled reciprocating motion, and power adjusted by feedback of infrared temperature measurements in-process. An off-line numerical simulation of the thermal field in scan welding is established, as well as a linearized multivariable model with real-time parameter identification. An adaptive thermal control scheme is thus implemented and validated both computationally and experimentally on a robotic Gas-Tungsten Arc Welding setup. The resulting productivity and quality features of scan welding are comparatively analyzed in terms of material structure and properties of the joint.

Structural damage detection using real-time modal parameter identification algorithm

A new approach to conducting structural damage detection is described. The approach employs a real-time modal parameter identification algorithm implemented in a digital-signal-processor-based data acquisition system. Because the modal parameter extraction process is conducted in real time, the algorithm is capable of identifying changes in structural properties attributable to structural damage as soon as they occur. Using the algorithm and a laboratory truss structure, it is demonstrated experimentally that continuous, real-time monitoring of anomalies attributable to structural damage is feasible. This monitoring capability will provide an early warning to an operator so that proper measures can be taken before a catastrophic failure occurs. The results of the damage detection study using the truss are presented along with the description of the real-time modal parameter identification algorithm.

Sensor for determining the water content of oil-in-water emulsion by specific admittance measurement

It is important to know the water content of oil-in-water emulsions for applications such as oil transportation. A new on-line sensor has been designed for this purpose. The real part of the specific admittance is used for the water-content determination. A simplified Van Beck mixing law and an appropriate frequency are applied to obtain the water content without polarization effects. Adaptive compensation is used in order to compensate temperature and salinity changes. On-line real-time parametric identification of conductivity versus temperature deviations observed in the water used in the emulsion is implemented. A microprocessor-based system with an open architecture and modular scheme allows the implementation of the electronic instrumentation and algorithms needed. Other features, such as auto-zero and auto-range, give the instrument a good performance and versatility for a wide range of salinities.

An adaptive power system stabilizer based on the self-optimizing pole shifting control strategy

An adaptive power system stabilizer (APSS) employing a new self-optimizing pole shifting control strategy and its application to a power system are described in this paper. Based on an identified model of the system, the control is computed by an algorithm which shifts the closed-loop poles of the system to some optimal locations inside the unit circle in the z-domain to minimize a given performance criterion. With the self-optimization property, outside intervention in the controller design procedure is minimized, thus simplifying the tuning procedure during commissioning. Also, a new method of calculating the variable forgetting factor in real-time parameter identification is discussed. Studies show that the proposed APSS can provide good damping of the power system over a wide operating range and significantly improve the dynamic performance of the system. >

Modelling, simulation and control in a data-rich environment

This paper describes the use and potential use of mathematical modelling, computer simulation, real-time parameter identification and adaptive feedback control techniques in data-rich clinical environments. We consider a system to be operating in a data-rich environment when measurements of system inouts and outputs are frequently available. Two particular application examples which illustrate the power of these techniques will be examined. The first example that will be described involves the development and testing of a real-time adaptive controller for simulataneous regulation of mean arterial pressure and cardiac output using two drugs. The second application of these techniques involves the real-time control of electrical stimulation for the functional use of paralyzed muscles in neuroprosthetic devices.

Applications of Modern Control Theory to the MMM Industries

The role of the modern control theory in applications to processes in the mining, mineral and metal (MMM) processing industries is elucidated. The processes are in general multivariable, as the product quality and the actions of control are both multiple quantities. The processes include often time-varying parameters, long transportation delays and severe stochastic disturbances.Examples are given in which ideas and methods of the modern control theory have been utilized in practical applications to realize real time parameter identification, state estimation, prediction, feedback control, decoupling and adaptation.It is concluded that the theory is indispensable to guide and support the development of better process control. The key to success is the right process model, in many cases a relatively simple model combining pre-knowledge and real time model fitting.

Unbiased estimator for identification of constant linear discrete systems

An unbiased estimator for real-time parameter identification and state estimation for constant linear discrete systems is presented. Bias correction is accomplished by taking account of the mean value of second-order terms which are generally neglected. The inclusion of the bias correction results in better performance for a quadratic cost criterion.