Identification of Distributed Parameter Systems, Based on Sensor Networks and Artificial Intelligence

The paper presents a short survey of three topics: modern sensor networks, distributed parameter systems and estimation techniques, especially using artificial intelligence tools, to be involved in the new domain of identification of distributed parameter systems, based on sensor networks and artificial intelligence. As smart and small devices the modern sensors are capable to be implemented in large distributed parameter systems. Sensor networks, with hundred and thousands of ad-hoc tiny sensor nodes spread across a geographical, are acting as a distributed sensor in a distributed parameter system. Sensor network topics, sensor network architectures and sensor network applications are presented. Examples of distributed parameter systems with large application in practice as the process of heat conduction, applications related to electricity domain, motion of fluids, the processes of cooling and drying, phenomenon of diffusion and other applications are presented. The identification techniques are useful for applications ranging from control systems, fault detection and diagnosis, signal processing to time-series analysis. Methods to estimate linear back box models and models of artificial intelligence, as fuzzy logic and neural network are presented. The artificial intelligence tools may be used for identification of nonlinear complex systems as the distributed parameter systems are. A case study of malicious nod detection based on a neural autoregression method in the process of plane heat propagation is developed.