Atmospheric pollution source identification method and device, computer equipment and storage medium

Abstract

The invention relates to artificial intelligence, the invention provides an atmospheric pollution source identification method and a device, computer equipment and a storage medium, and the method comprises the steps: obtaining various atmospheric pollution components of a plurality of pollution sources, constructing a pollution component matrix, and initializing a first target function as follows: the pollution component matrix = a pollution component contribution matrix * a pollution component content matrix + a residual matrix; calculating a pollution component contribution matrix and a pollution component content matrix according to the residual matrix when the residual matrix is iteratively calculated through a gradient descent algorithm to be smaller than a threshold value; calculating the mass concentration contributed by each pollution source according to the pollution component matrix and the pollution component content matrix so as to determine the source spectrum of each pollution source; and identifying the pollution source type corresponding to the source spectrum of each pollution source according to the mapping relationship between the source spectrum of the pollution source and the pollution source type. According to the invention, each source of the atmospheric pollution components can be accurately identified. In addition, the invention also relates to the blockchain technology, and the mapping relationship is stored in the blockchain.

Description

technical field [0001] The invention relates to the technical field of artificial intelligence, in particular to an air pollution source identification method, device, computer equipment and storage medium. Background technique [0002] Air pollution has become a major bottleneck and potential risk that affects economic and social development. The effective prediction of air pollutants is of great significance for fighting the "blue sky defense war" and effectively coping with heavy pollution weather. [0003] The artificial neural network prediction method is the focus of research in the field of artificial intelligence in the world in recent years. The artificial neural network prediction method does not need a clear functional relationship between input and output. It mainly completes the simulation process by training and learning a large amount of data, and Use the trained network to make predictions on new input data. However, the neural network can only predict the ...

AI Technical Summary

Problems solved by technology

However, the neural network can only predict the degree of atmospheric pollution through the current state of the influencing factors, and cannot effectively predict the source of pollution.

Images