USE OF A NEURAL NETWORK APPROACH TO OPTIMIZING CONCRETE COMPOSITION WITH RESPECT TO STRENGTH

摘要

A neural network method is described for optimizing the composition of any composite material, which makes it possible to reduce expenditure on time and material provisions for performing research work. An optimum prescription is proposed for ash-containing concrete, using the method developed, providing maximum strength.Currently in accordance with existing nature protection legislation all enterprises of the fuel and energy complex pay for contamination of the environment, and for formation and disposal of solid wastes. These payments are very significant, and with an increase in the proportion of combustible carbon they may increase even more [1 ]. In view of this, a considerable increase is proposed in the volume of ash and slag materials used in enterprises of the building industry. Ash and slag materials are a promising raw material source, whose use has important ecological and economic value, particularly in the production of concrete and ferroconcrete materials, objects and structures for different forms of building. For example, addition of ash to concrete makes it possible to save considerably very scarce granite gravel and sand.