Sustainability Enhancement in Healthcare Facility Design: Structural and Functional Optimization Based on GCN

The increasing age of the population and the acceleration of urbanisation are giving rise to a growing emphasis on the layout and functional optimisation of medical facilities. To address this challenge, this study proposes a model for optimising the structure and function of medical facilities based on a Graph Convolutional Network (GCN). This combines existing deep learning and graph convolution methods to enhance the capabilities of traditional optimisation models by integrating the advantages of graph convolutional networks in modelling spatial and functional relationships. In this model, a graph representation of medical facilities is first constructed, with nodes representing facilities and edges representing spatial and functional connections between facilities. These relationships are then learned through a graph convolutional network, with the objective of capturing the distribution and demand characteristics of medical resources. Secondly, in order to enhance the model's adaptability to diverse scenarios, an adaptive graph convolution mechanism is introduced on the basis of the traditional GCN. This improves the model's generalisation ability in different medical environments by dynamically adjusting the graph structure and weights. Ultimately, the model optimises the allocation of medical resources through end-to-end training, thereby maximising the coverage and efficiency of medical services. The experimental results demonstrate that the proposed model enhances the efficiency and quality of medical resource allocation in comparison to existing optimisation methods.