In recent years, domestic and foreign experts have begun to focus on exploring the field of environmental art. However, standard cultural algorithms have problems such as low convergence accuracy and poor application effects in innovative design of environmental art. The standard cultural algorithm no longer meets the requirements of environmental art for innovation, and the innovation effect provided is not ideal. For this purpose, this article establishes an innovative design environment art model based on pattern learning optimization of cultural algorithms. Firstly, genetic algorithms are used to provide the required population to the population space in cultural algorithms, and then pattern extraction is used to extract the feature information of excellent individuals. Within a certain period of time, all individuals in the group learn patterns on feature information, maximizing the guiding role of excellent patterns, Accelerate the convergence speed of the algorithm. Through simulation experiments, it has been confirmed that the cultural algorithm proposed in this article for pattern learning optimization has higher multi-objective optimization accuracy compared to the standard cultural algorithm, and is widely used in environmental art innovation design.