Research on the Geographical Rationality of FAI Allocation in Guangzhou Districts Based on Differential Evolution

The allocation of Fixed Asset Investment (FAI) across districts in Guangzhou significantly influences the coordinated development of the regional economy. Particularly under the constraints of geographical adjacency, the rationality of FAI allocation has garnered considerable attention. In this study, we propose a regional layout optimization method based on soft constraints of geographical adjacency to explore the rationality of FAI allocation. This paper constructs a model integrating economic indicators such as GDP, resident population, total retail sales of consumer goods, and FAI. The weights of these indicators are calculated using Gradient Boosting Decision Tree (GBDT) and Principal Component Analysis (PCA). The rationality of FAI allocation in 2022 is analyzed by combining Multilayer Perceptron (MLP) and Differential Evolution (DE). Furthermore, Kepler Optimization Algorithm (KOA) is introduced to model and optimize the decline in FAI in some districts in 2023, thereby enhancing the analytical accuracy. The results indicate that FAI allocation across Guangzhou districts exhibits a certain level of rationality under geographical adjacency constraints. However, districts such as Panyu and Nansha still require optimized investment strategies. This study provides quantitative support for policymakers, contributing to enhanced regional economic resilience and coordinated development.