Intelligent planning and design of landscaping based on BPNN

As an important part of ecological civilization construction, urban landscape greening directly affects urban ecological quality and landscape aesthetics through its planning and design standards. This study develops a genetic algorithm-back propagation neural network (GA-BPNN) intelligent quantitative evaluation technique for landscape greening planning and design, which combines GA and BPNN to enhance the scientific nature and objectivity of design evaluation. The study creates a hierarchical index system and determines index weights using the Analytic Hierarchy Process (AHP), with a focus on fundamental factors such as ecological benefits, landscape functions, artistic value, economic rationality, and social implications. Questionnaire surveys are used to collect 800 valid samples from four distinct gardens. Experimental results show that the mean squared error (MSE) and root mean squared error (RMSE) converge steadily during model training, and high prediction accuracy is achieved on the test set. The mean absolute error, mean square error, and root mean square error for the four gardens are 2.12, 5.01, and 2.24, respectively. The R 2 values of all gardens exceed 0.95, with an average of 0.963. In the case application, scenario simulation optimization was applied to the landscape with the lowest overall score. The overall score improved from 68.95 to 79.83, reflecting an increase of 15.8%. Specifically, the aesthetic value increased by 28.7%, and the landscape functionality improved by 19.7%. The results demonstrate that GA-BPNN can effectively capture complex nonlinear relationships and has strong generalization ability. The study’s conclusions are crucial for advancing the shift in landscape design from experience-based to data-driven intelligent decision-making, as well as for enhancing the quality of the urban ecological environment and landscape.