Automatic Pattern Generation Based on Genetic Algorithm and its Application in Design Platform
Yue He, Lingxia Zhang · Procedia Computer Science · 2025
In order to solve the problems of limited creativity and low design efficiency in the traditional pattern design process, this paper introduces genetic algorithm for automatic pattern generation. This method continuously optimizes the pattern genome by simulating the operations in the biological evolution process to generate patterns that meet specific aesthetic standards and design requirements. This paper first initializes random patterns as the initial population, and each pattern is regarded as an individual with its own genome. Next, a fitness function is defined to evaluate the conformity of each pattern. Then, based on the fitness score, the patterns with better performance are selected to enter the next generation, simulating the "survival of the fittest" law in nature. Through mutation operations, the gene values of some patterns are randomly changed to increase the diversity of the population and avoid falling into the local optimal solution. In the design platform, the generation technology proposed in this paper is integrated into a functional module. Users can set pattern parameters through a simple interface and then start the automatic generation process. The system will automatically run the genetic algorithm based on the user’s input and display the generated patterns and their evolution process in real time on the design platform. Compared with the traditional rule-based pattern generation method, the proposed method has advantages in generation time, color consistency and innovation score, and the maximum generation time is only 399 milliseconds. The proposed method provides stronger support for the development of the design industry.