Artificial intelligence is increasingly integrated into architectural design, notably in layout planning, where it aids in identifying innovative designs from numerous options. This paper focuses on kindergarten architecture, tackling inefficiencies and high error rates in conventional spatial layout design. A Layout Generation and Screening framework based on RHINO + GH is proposed to facilitate non-programmer architects to manipulate layouts through an intuitive interface. Additionally, an expert evaluation scale is developed to assess the generated outcomes. The framework's viability and efficacy are confirmed through its implementation in a case study.