Abstract: Sales forecasts for auto parts manufacturers are critical to the overall health and sustainability of the auto industry. As a result, it has become critical to design a convenient and accurate forecasting model based on little historical data. By examining a modest amount of valid data, gray prediction theory can investigate the law of change. The Improved Grey Model (1,1) (IGM (1,1)) model is introduced in this study, which conducts a functional transformation on the original data series in order to create a new one with a high degree of smoothness. A genetic algorithm is utilized to establish the optimal parameter values for the background values, which enhances prediction accuracy. The model's predictive accuracy was evaluated using the annual sales of Company B, a Changchun-based auto parts manufacturer, from 2009 through 2020. The numerical findings indicate that the proposed method outperforms the four models regarding forecasting performance. Additionally, the proposed method is critical for conducting in-depth research, promoting, and implementing the gray model in auto parts manufacturing firms.  

Key words: GM (1,1), Smooth Degree, Background Value, Genetic Algorithm, Auto Parts Manufacturer