Sustainable Management of Decorative Patterns of Shanghai Shikumen Architectural Houses Based on ROI Optimization

The decorative patterns of Shikumen architectural dwellings in Shanghai are of great investment value, and their cultural benefits can be fully utilized through sustainable operation and management. In this paper, after analyzing the traditional pattern generation method, we combine the semantic generation method with the design of the loss function and network structure of the generative adversarial network and construct the generation algorithm of architectural residential decoration pattern based on semantic perception. The constructed pattern generation algorithm is used to manage the sustainability of architectural residential decoration patterns and optimize the rate of return on the investment of sustainability management through fuzzy theory. The results show that the color similarity, pattern similarity, and organization form similarity are (0.62, 0.91), (0.58,0.89), and (0.59,0.88), respectively, and the return efficiency of the investment in the architectural residential decoration of Shanghai Shikumen architecture before and after the optimization fluctuates between (-0.26, 0.30) and (-0.18, 0.26). The ROI of sustainable business management of decorative patterns in Shanghai Shikumen architectural dwellings in this study can develop stably.