PM2.5 Concentration Detection Method Based on Interval Radial Basis Function Neural Network

Abstract

The present invention is a PM2.5 concentration detection method based on an interval radial basis function neural network, which belongs to the field of air quality detection; the method first converts the air into a photoelectric signal, collects the photoelectric signal and the corresponding actual PM2.5 concentration and normalizes it , then use the normalized interval photoelectric signal as the input value, and the interval PM2.5 concentration as the output value, train the interval radial basis function neural network, and finally use the trained interval radial basis function neural network to obtain the final Interval PM2.5 concentration, the present invention overcomes the shortcomings of difficulty in solving the matrix equation algorithm and low accuracy. It can not only effectively detect the PM2.5 concentration, but also detect the PM2.5 concentration detected under uncertainty conditions. Within which range it floats, the detection accuracy is high and the detection method is simple.

Description

technical field [0001] The invention belongs to the field of air quality detection, in particular to a PM2.5 concentration detection method based on an interval radial basis function neural network. Background technique [0002] In recent years, many areas in my country have been caught in severe smog weather, and PM2.5 is the culprit of smog weather; PM2.5 refers to the particulate matter in the atmosphere with an aerodynamic diameter less than or equal to 2.5 μm, which has also become an inhalable Lung particulate matter; compared with coarser atmospheric particulate matter, PM2.5 has a small particle size, is rich in a large amount of toxic and harmful substances, has a long residence time in the atmosphere, and has a long transportation distance, pollutes the atmospheric environment, and poses a serious threat to people's health. Therefore, it is an important and meaningful work to carry out PM2.5 environmental quality detection research. [0003] Due to the simplified ca...

AI Technical Summary

Problems solved by technology

[0003] Due to the simplified calculation in the derivation and the limitation of the physical size of the equipment, it is difficult for the traditional PM2.5 concentration detection method to solve the matrix equation algorithm, and the detection accuracy is low; and due to the influence of environmental conditions, detection equipment, random factors, etc., the detection equipment When the air is converted into a photoelectric signal, the converted photoelectric signal has a certain degree of uncertainty; at the same time, the real-time PM2.5 concentration announced on the official website of the National Environmental Monitoring is also measured by some monitors, and there is also uncertainty

[0004] At present, the training data of the existing PM2.5 concentration detection method based on the basis function neural network is point data, which also does not take into account the uncertainty of the detected photoelectric signal and real-time PM2.5 concentration, and cannot obtain When there is uncertainty in the photoelectric signal, what range does the detected PM2.5 concentration fluctuate in? Under the condition of considering this uncertainty, the traditional neural network has become powerless

Images