Abstract:Abstract In this paper, propose a BP artificial neural network optimized by artificial bee colony algorithm with a PM2.5 value for the current Nanning area. In this paper, aiming at the shortcomings of BP neural network prediction, the global optimization performance of artificial bee colony algorithm is used to provide the optimal global initial value for BP neural network and avoid the neural network falling into local optimization. To make the prediction model more accurate. PM2.5 is a small part of the atmospheric chemical composition, which has a great impact on the air quality and thus on the human body. With people’s attention to their own health nowadays, it is of great practical significance to predict the PM2.5 content in a certain period in the future. The results show that the average square error and average relative error of the prediction of PM2.5 content based on BP artificial neural network are only 1.25% and 2.46% respectively. BP artificial neural network based on bee colony can be well applied to the prediction of chlorophyll a content, providing a reference for the prediction of PM2.5 content in a certain period in the future.

Prediction of PM2.5 value by BP artificial neural network in honeybee colony