This article gives a survey of the issue of the thermal dynamics identification of the sensor network based systems, such as buildings. Given problems: weaknesses of the modern sensor network based intelligence systems, that mostly limits their potential and increasing calculation costs are overviewed. Article describes a novel approach for intelligence control systems design based on sensor networks, using methods of fuzzy modeling. The key future of the given approach is in saving of high efficiency characteristics of modeling in case of sensor network nodes amount decreasing.  As an example of the approach usage there are presented results of experimental test space thermal dynamics simulation model analyzing. The advantages of the developed solution to other existing commercial solutions are shown. The methodology was tested and its efficiency is shown.

Keywords: intellectual control systems, sensor network, data analyzing, fuzzy modeling, fuzzy classification, membership function.