Abstract
The paper presents the structural modeling of the wind's influence over the heat flow of a greenhouse, starting from recorded data of an experimental greenhouse. A structural model is much more versatile than a synthetic one, however greenhouses are complicated and highly non-linear, and therefore such a model should be built by aggregating several elementary sub-models with physical significance. The paper is aiming to identify a sub-model for the windy dry nights. The modeling method consists in the identification of the sub-models with low order systems with nonlinear coefficients and their optimization using genetic algorithms. The knowledge of the wind's influence over the thermal behavior of buildings and vehicles help us to improve the designing of such products. Two case studies are presented: a passive greenhouse provided with a heat pump/wind generator renewable energy source and a railway coach. The simplicity of such models encourages implementations at micro-controllers /DSP levels.