Back to Statistics: Tail-Aware Control Performance Assessment

Due to the popularity of artificial intelligence today, scientists and practitioners tend to ignore the achievements of statistical methods. This book aims to re-establish statistical methods, especially for control practitioners, for whom the task of assessing control performance is important. The author introduces the elements of statistical theory, including basic statistical concepts and the description of distributions. Extending the most common observations toward the tails and extreme statistics, he demonstrates the robust statistics approach that deals with the tails, followed by the description of methods that can visualize and bring forward statistical properties as long as the derivative issues. By addressing statistical issues of sustainability, extreme statistics, non-Gaussianity, L-moments, tail index and index ratio diagrams, the author aims to change traditional concepts and broaden the scope of available methods. Simulation case studies and real industrial cases are also presented. The book will be of interest to site control engineers and scholars assessing control systems and process performance.