Use of artificial intelligence for the optimal design of solar systems

The objective of this work is to use artificial intelligence methods for the optimal design of solar energy systems. The lifecycle savings of the system is used as the optimisation parameter. The variable parameters in this optimisation are the collector area, slope and mass flow rate and the volume of the storage tank. An artificial neural network is trained, using the results of a small number of simulations carried out with TRNSYS program, to learn the correlation of the above variable parameters on the auxiliary energy required by the system from which the lifecycle savings can be estimated. Subsequently, a genetic algorithm is employed to estimate the optimum size of the variable parameters, which maximises lifecycle savings. As an example, the optimisation of a large hot water system is presented. The optimum solution obtained from the present methodology is achieved very quickly as compared to the time required to obtain the same solution by the traditional trial and error method, which would require thousands of runs of TRNSYS to cover all possible combinations considered by the genetic algorithm