Bøger af Biao Huang

The presence of non-linearities, such as stiction and deadband, places limits on the performance of control valves. Indeed, in the process industries, stiction is the most common valve problem, and over the last decade numerous different techniques for overcoming it have been proposed.Detection and Diagnosis of Stiction in Control Loops represents a comprehensive presentation of these methods, including their principles, assumptions, strengths and drawbacks. Guidelines and working procedures are provided for the implementation of each method and MATLABbased software can be downloaded from www.ualberta.ca/~bhuang/stiction-book enabling readers to apply the methods to their own data. Methods for the limitation of stiction effects are proposed within the general context of: oscillation detection in control loops, stiction detection diagnosis; , stiction quantification and diagnosis of multiple faults.The state-of-the-art algorithms presented in this book are demonstrated and compared in industrial case studies of diverse origin - chemicals, building, mining, pulp and paper, mineral and metal processing. Industry-based engineers will find the book to be valuable guidance in increasing the performance of their control loops while academic researchers and graduate students interested in control performance and fault detection will discover a wealth of static-friction-related research and useful algorithms.

The series Advances in Industrial Control aims to report and encourage technology transfer in control engineering. New theory, new controllers, actuators, sensors, new industrial processes, computer methods, new applications, new philosophies.

A typical design procedure for model predictive control or control performance monitoring consists of: 1. design of the control law or calculation of performance indices according to the predictor. Both design problems need an explicit model form and both require this three-step design procedure.