PM2.5 Prediction Model Based on ABC-BP

Now days, the concentration of PM2.5 is mostly predicted by BP-neural network. But the BP-neural network is sensitive to the initial value, and it is easy to fall into the local minimum value. Artificial Bee Colony Algorithm (ABC) is adopted in this paper for solving the problems and optimizing the BP-neural network, using the idea and method of that bees are looking for high-quality food efficiently, to find the best value of weights and thresholds of network. The optimized neural network not only possess generalized mapping ability of neural network, but also possess global iteration and local search ability like Artificial Bee Colony Algorithm. Through comparative experiments, it is proved that the PM2.5 prediction model based on the neural network optimized by Artificial Bee Colony Algorithm effectively improves the accuracy of PM2.5 prediction.