摘要
The research for the new algorithm is in the forward position and an issue ofgeneral interest in chemometrics all along. A novel chemometrics method, Chemical Ant ColonyAlgorithm, has first been developed. In this paper, the basic principle, the evaluation function,and the parameter choice were discussed. This method has been successfully applied to the fitting ofnonlinear multivariate function and the optimization of test conditions in chrome-azure-S-Alspectrophotometric system. The sum of residual square of the results is 0.0009, which has reached agood convergence result.
The research for the new algorithm is in the forward position and an issue ofgeneral interest in chemometrics all along. A novel chemometrics method, Chemical Ant ColonyAlgorithm, has first been developed. In this paper, the basic principle, the evaluation function,and the parameter choice were discussed. This method has been successfully applied to the fitting ofnonlinear multivariate function and the optimization of test conditions in chrome-azure-S-Alspectrophotometric system. The sum of residual square of the results is 0.0009, which has reached agood convergence result.
基金
ProjectsupportedbytheNationalNaturalScienceFoundationofChina (No .2 0 175 0 13 )andtheHigherEducationDevelopmentFounda tionofShanghaiinChina (No .0 1A17)
