Introduction to the Special Section

ystem identification refers broadly to the problem of developing empirical mathematical models of systems based on excitation and response measurements. Such models are useful for enhancing physical understanding; analyzing system properties; and performing simulation, prediction, filtering, state estimation, monitoring, and fault diagnosis as well as control. Practical interest in system identification is driven by difficulties associated with deriving models from physical principles. Even when a model structure is available from first principles, it is often necessary to apply identification techniques to estimate parameter values. These scenarios make identification a vital tool in many engineering applications, from vibration suppression to process control. The potential of identification technology transcends engineering, however, and is relevant to problems in fields such as economics, biology, medicine, ecology, and geology. Historically, the origins of system identification can be traced back to Gauss and his least squares method for building models to describe the motion of heavenly bodies. Most of the major advances and the development of systematic procedures did not take place, however, until the advent of digital computers and the widespread availability of instrumentation and measurement hardware in the 20th century. The increasing need in recent decades for systems displaying high performance and reliability has further fostered the need for accurate mathematical models, and thus identification techniques, in many engineering applications. As a consequence, the field that was INTRODUCTION TO THE SPECIAL SECTION