Structural neighborhood modeling of the industrial ventilation system.
Both in the chemical and metallurgy production there is a problem of exceeding the allowable concentration of harmful substances in the premises, shops and in the environment. If we consider cement production, we deal with dust, associated with a non-optimal operation of the dust-free ventilation system in the clinker burning department. The optimally organized ventilation system in any type of production ensures the microclimate of the production premises, corresponding to the sanitary norms and rules, which contributes to the increase of the staff’s efficiency. The optimal operating modes of the industrial ventilation system associated with the technological process allow solving energy saving issues in the ventilation section and the maximum productivity. In this paper, the questions of the neighborhood modeling of the ventilation system in the premises of the cement production shop are considered. Any neighborhood model is a system of equations on an oriented graph such that the equations of the model correspond to the vertices of the graph and the entering edges specify the sets of variables participating in the corresponding equation. Therefore, any neighborhood model is structurally identified at least at the level of occurrences of variables in the equations. Further, postulating the type of equations (linear, bilinear, etc), we usually deal with parametric identification. Thus we eliminate the difficult problem of structural identification due to the introduction of a large number of parameters, and this is the advantage of the neighborhood-oriented approach. However, for reliable identification of the parameters we need a large number of experimental data and, in addition, we must be prepared for the difficulties created by multicollinearity. Therefore, it will be useful to take into account any available information related to the structure of the equations. In fact, for the ventilation-filtration model some simple physical considerations make it possible to significantly reduce the number of parameters.The proposed measures allow for purification of fresh air, remove excess heat, moisture, dust, harmful gases and vapors entering the air of workspaces and the atmosphere. The refinement of the structure of the model is based on physical considerations and leads to a piecewise trilinear dependencies with a significantly reduced number of coefficients subject to further parametric identification. A system for minimizing energy costs and reducing dust emission in the clinker burning shop is proposed, which allows increasing the environmental safety of production. This system can be applied to chemical and metallurgical production, where harmful substances are released into the air, with a slight adjustment for a specific production process. In the case of cement production workshop, the described ventilationfiltration model in the simplest case leads to a linearly constrained quadratic optimization problem.
Keywords: neighborhood structure, ventilation system, structural identification, parametric identification, complex systems.