A novel immune algorithm for resolution and quantitative determination of t
he components in overlapping chromatograms was proposed by imitating biolog
ical immune systems. The algorithm takes an overlapping chromatogram as its
input and subtracts the chromatograms of standard samples from the input b
y iteration of a network. When the residual does not change, the network wi
ll converge and chromatographic information of the components in overlappin
g chromatogram will be obtained. Both simulated and experimental data sets
were investigated by the method. Results showed that both resolved results
and recoveries of quantitative determination are satisfactory. Comparing wi
th conventional least-square method, the immune algorithm is fast in calcul
ation.