Optimization of landscape element layout in environmental art design based on genetic algorithm

Abstract

In traditional environmental art design, it is often challenging to comprehensively account for a multitude of complex factors, and the optimization process tends to be computationally intensive and intricate. Key aspects such as illumination, shadow, and sightlines between landscape elements are frequently overlooked, leading to incoherent layout schemes in practical applications. This study addresses these limitations by employing genetic algorithms (GAs) to optimize the layout of landscape elements, thereby automating the generation of optimal layout schemes. Initially, an objective function is established to evaluate layout schemes by holistically considering landscape aesthetics, functionality, ecological benefits, and feasibility, with each factor’s contribution quantified. The layout scheme is encoded into real-valued representations, where each individual denotes the location, size, and other attributes of landscape elements. An initial population is generated to ensure diversity and adherence to basic constraints. Each generation of individuals is assessed using a fitness function that integrates multiple sub-goals, evaluating the merits of each layout scheme through weighted summation. The tournament selection method is employed to choose individuals with higher fitness in each generation for crossover and mutation operations, thereby generating new layout schemes. Through iterative optimization over multiple generations, the genetic algorithm refines the layout scheme. Experimental results demonstrate that the overall space utilization rate exceeds 0.7, and the weighted scores of the landscape layout schemes range between 8.075 and 8.59. This approach offers a novel solution for environmental art design, significantly enhancing the coherence and practicality of layout schemes.