The use of the interval Pareto sorting method in simulation-based design of an indoor environment

Abstract

The optimization algorithm and computational fluid dynamics analysis have been widely used to design indoor environment. The optimization algorithm is used to search the optimal design variables, and computational fluid dynamics is used to calculate the values of the design objectives. In the design process, there are usually many design objectives and the target solution is expected to be an interval value rather than an isolated value that is difficult to maintain. In the current article, the genetic algorithm and computational fluid dynamics are used to design an indoor environment. A new sorting method, called the interval Pareto sorting method, is proposed to simultaneously solve the multiple-objective and interval solution problem. The interval Pareto sorting method consists of the Pareto sorting method and hypervolume. The artificial neural network is used as a surrogate model of computational fluid dynamics to reduce the computational cost. Two benchmark cases are used to test the proposed method. The interval solutions are obtained to satisfy the requirements of thermal comfort, air quality, and energy cost.