Non-contact real-time in-situ water quality monitoring method

Abstract

The invention discloses a non-contact real-time in-situ water quality monitoring method which comprises the following steps: acquiring spectral data and data of a water sample to be detected, and constructing a water quality parameter deep learning model based on hyperspectrum; based on the water quality parameter deep learning model, the spectral data and the to-be-detected water sample data, obtaining an optimal algorithm model of the to-be-detected water sample parameters; and obtaining a monitoring result through a multi-parameter water quality inversion algorithm based on the optimal algorithm model of the parameters of the water sample to be detected. The monitoring method does not need to be in direct contact with a water body to be detected, and compared with traditional underwater probe contact type monitoring, the monitoring method is low in energy consumption, small in loss, easy and convenient to maintain and very limited in influence of external environments such as stormy waves, and the monitoring precision is improved.

Description

Technical field [0001] The present invention belongs to the field of water quality, in particular, to a non-contact real-time in situ water quality monitoring method. Background technique [0002] With the rapid development of my country's economy, water pollution problems are increasingly serious. In order to monitor the changes of water quality, the number of national monitoring and the number of monitoring points, as of June 2019, the country has a total of 2050 national surface water assessment, and the provincial control and local control section is more star row. Relying on traditional ground surveys and methods of considering water quality to supervise these important sections will cost a lot of human and financial resources. In particular, due to the traditional monitoring method, most of the sensor is immersed in the water body to be tested, it is often limited to the measurement; in addition, since the sensor is susceptible to corrosion or due to the present, it is nece...

AI Technical Summary

Problems solved by technology

[0006] (1) Insufficient temporal and spatial resolution: more than 2 / 3 of the lakes in my country are less than 10km 2 , for rivers, the width is basically less than 1km

For small water bodies, at present, it can only be observed by Landsat, Sentinel, Gaofen and other medium- and high-resolution satellites (10-30m). However, Landsat satellites have a long revisit cycle and it is difficult to capture short-term changes in the water environment, thus limiting lakes, reservoirs, and rivers. and remote sensing applications of water environment in important sections

[0007] (2) The radiation dynamic range is narrow, and the signal is easy to be saturated: the ocean water body is relatively clean, and the water body signal is weak. The water color sensor is set to receive the water body signal better, and the radiation dynamic range setting is low

However, for inland turbid water bodies such as lakes and rivers, the radiation signal is much higher than that of ocean water bodies, which often exceeds the threshold range of the sensor, resulting in saturation of the band signal reflecting water quality changes, thereby reducing the accuracy of remote sensing estimation

[0008] (3) The band setting does not meet the requirements of inland water bodies: the concentration of chlorophyll a in inland water bodies is high and is affected by suspended solids and CDOM. Sensitive band of prime monitor is around 620nm

However, the empirical factors in the MUMM algorithm are affected by the water conditions, and the universality is not good; while the SWIR algorithm requires two short-wave infrared bands, except for MODIS, VIIRS, and Landsat, most sensors do not have this condition

Ground-based (shore-based) monitoring can effectively avoid the influence of the atmosphere on the accuracy of remote sensing signal extraction of water bodies. Although ground-based monitoring technologies have been widely used, such as traffic cameras, video surveillance, etc., these monitoring equipment can only provide very limited Spectral bands, and these bands are not designed for water bodies, so remote sensing of water quality parameters cannot be extracted through them

Images