The invention discloses a
system for detecting and analyzing water-soluble
heavy metals in an ecological
system based on
machine learning. The
system is used for detecting water-soluble harmful heavy
metal elements in atmospheric
particulates. The system comprises a sampling preprocessing module, an
electrochemical detection module, a
machine learning analysis module and a
data verification output module which are connected in sequence, the sampling pretreatment module is used for collecting PM2.5 and PM10 particles by adopting a
quartz fiber filter membrane and carrying out dissolving and filtering pretreatment; the
electrochemical detection module is used for heavy
metal enrichment and detection; the
machine learning analysis module adopts an improved CNN (
Convolutional Neural Network) to automatically identify the
volt-
ampere spectrogram, and establishes a nonlinear mapping model of the relationship between the
metal ion concentration and the
peak current; and the
data verification output module performs quantitative analysis by adopting a
standard addition method, and compares and identifies the
pollution source through a multi-element
fingerprint feature
database. According to the invention, eight heavy metal elements such as Cd, V, Cr, Ni, Se, As, Mn and Pb can be detected at the same time, the detection range is 0.1-1000 [mu] g / L, and intelligence and
automation of heavy metal detection are realized.