Key points are not available for this paper at this time.
Roboticists are increasingly dealing with challenging complex problems in system identification for model-based control, and this book lays a foundation of knowledge for the reader to absorb, which can help address the said challenges. That being said, the reader should prepare for a challenge when reading this book. Though it is an excellent text, it is not for the casual reader. A background in Fourier series and Laplace transforms helps as well,though Ljung does a good job with the early chapters, providing an overview of the concepts that will be used throughout the book. This book covers parametric and nonparametric methods, parameter estimation methods in the prediction error framework, frequency domain data and interpretations, various ways to compute estimates, recursive estimation techniques, model validation,and case studies. Each chapter ends with an extensive bibliography for further reading and, often, has an appendix for proofs and derivations. The author presents a thorough and complete picture of system identification as a methodology, set of tools,and practical approach to generating models from data. The author's work has been influential for decades, and anyone interested in buildingmodels from data would be well served to put the effort into mastering thismaterial.
Alex Simpkins (Fri,) studied this question.