پیش بینی معکوس و بهینه سازی شرایط کنترل جریان برای فضاهای محدود با استفاده از یک الگوریتم ژنتیک مبتنی بر دینامیک سیالات محاسباتی

Optimizing an indoor flow pattern according to specific design goals requires systematic evaluation and prediction of the influences of critical flow control conditions such as flow inlet temperature and velocity. In order to identify the best flow control conditions, conventional approach simulates a large number of flow scenarios with different boundary conditions. This paper proposes a method that combines the genetic algorithm (GA) with computational fluid dynamics (CFD) technique, which can efficiently predict and optimize the flow inlet conditions with various objective functions. A coupled simulation platform based on GenOpt (GA program) and Fluent (CFD program) was developed, in which the GA was improved to reduce the required CFD simulations. A mixing convection case in a confined space was used to evaluate the performance of the developed program. The study shows that the method can predict accurately the inlet boundary conditions, with given controlling variable values in the space, with fewer CFD cases. The results reveal that the accuracy of inverse prediction is influenced by the error of CFD simulation that need be controlled within 15%. The study further used the Predicted Mean Vote (PMV) as the cost function to optimize the inlet boundary conditions (e.g., supply velocity, temperature, and angle) of the mixing convection case as well as two more realistic aircraft cabin cases. It presents interesting optimal correlations among those controlling parameters.

With rapid developments in fluid dynamics, numerical science and computer technologies, computational fluid dynamics (CFD) has become an efficient tool for indoor environment study and system design. Optimizing an indoor flow pattern according to specific design goals requires systematic evaluation and prediction of the influences of critical flow control conditions such as flow inlet temperature, velocity and angle. In order to identify the best flow control conditions, conventional approach simulates a large number of flow scenarios with different boundary conditions. Previous studies reveal advanced search and optimization algorithms such as genetic algorithm (GA) can effectively reduce the total number of iterations to reach an (or a group of) optimal solution(s) [1]. GA is an optimization algorithm that simulates natural evolution in the search of optimal solutions [2]. It has been applied to a variety of engineering design, parameter identification and system optimization. Efforts of coupling GA with CFD, however, were mostly on the optimization of exterior geometries of various objects. For instance, Obayashi and Takanashi [3] combined GA with CFD to optimize the target pressure distributions of an airfoil. Other examples include the shape design of cars [4], melt blowing slot die [5], and transition piece [6]. Malkawi et al. [7] used the CFD-based GA method to search the optimal room size and ventilation system which can satisfy both thermal and ventilation requirements. The study was for a relatively simple case and did not evaluate the influence of CFD prediction error on optimal design. Kato and Lee [8] applied a similar approach to optimizing a hybrid air-conditioning system with natural ventilation. The study aimed at minimizing the energy consumption of the mechanical systems. The study developed a two-step method to reduce the computing effort. Optimal results were acquired first with a coarse CFD mesh, which were then refined with a fine mesh for detailed analyses. It should be noted that CFD results with coarse meshes may produce incorrect flow field which can lead to wrong optimal results. Zhou and Haghighat [9] and [10] employed the artificial neural network (ANN) to reduce the modeling time; however, training the ANN still requires a great number of CFD cases for individual projects. This study develops a general simulation tool by integrating GA and CFD programs. The tool can inversely predict control conditions based on limited experiment data of indoor flow field. It can also be used to optimize various control conditions of indoor flows under different objective functions (PMV in this paper). The tested control conditions include supply velocity (vector) and temperature of flow inlet, etc. Locally optimal solutions and multiple solutions may exist for multi-variable optimization problems, where the GA presents the special strength in capturing the global optimal results and multiple solutions.

This study developed a CFD-based GA simulation platform that can be used to predict and optimize the inlet flow control in various confined spaces. The case studies demonstrate the capability and accuracy of the developed algorithm. The platform can inversely predict inlet conditions based on simulated room flow conditions with high accuracy. The study also presented an efficient way to inversely optimize the supply air conditions for thermal comfort in confined spaces.