Identification of SCADA systems: case studies

The paper initiates an investigation of the problem of developing quantitative models for (uncertain) supervisory control and data acquisition (SCADA) systems by studying this problem for simple yet illustrative example systems. These studies indicate that interesting and important system identification problems can be addressed by modeling the SCADA system as a suitably parameterized hybrid dynamical system and then adaptively identifying the unknown parameters using measured data together with information regarding the basic structural features of the model. Algorithms are given for identifying both the underlying physical component and the logical decision making component of the SCADA system, and the performance of these algorithms is illustrated through applications in an automated transportation system and an economic system.