Identification of Distributed Parameter Systems Based on Sensor Networks and Multivariable Estimation Techniques

The domain of system identification may be developed today using the powerful tool represented by the intelligent sensor networks, placed in real distributed parameter systems. The sensor networks, as a “distributed sensor”, allow the usage of multivariable estimation techniques, in different ways: classical linear methods of modelling or methods based on artificial intelligence for complex non-linear systems. The paper presents a short survey of some results obtained in the study of using of sensor networks with multivariable estimation techniques to estimate heat transfer in space. Key-Words: Wireless sensor networks, system identification, distributed parameter systems, neural networks, multivariable estimation techniques, auto-regression, heat distribution.