A N-Order Grey-GA Optimizer to Forecast Taiwan Pollution Trends

This research designs a novel n-order grey model (GM) to predict import and export effects on air pollution trends in Taiwan. We also present an optimal refiner of Grey-Genetic Algorithms (GGA) to improve GM efficiency. In our test-beds, we propose several GM models for the prediction of air pollution trends. The simple model (GM(1, 1)) and the n-order multiple model (GM(1, N) and RGM(1, N)) are utilized in air pollution prediction that was evaluated on their performances. These GM models are compared with those of the import and export data from Taiwan manufacturers. The results show that the prediction accuracy of the (GM(1, N) and RGM(1, N)) model presented the better accuracy than GM(1, 1). The experimental results also show that the GGA controller can refine the prediction accuracy of GM models. This research is successfully applied these N-order multiple GM models for producing more accuracy information for pollution control.