적응격자 알고리즘을 이용한 대기오염 예측에 관한 연구

A Study on Air Pollution Prediction Using Adaptive Lattice Altorithm

In this paper a adaptive LMS(least mean-square) lattice predictor, which is composed of the adaptive lattice algorithm and LMS algorithm by Widrow-Hopf, is used to predict the future air pollution of the extraordinary levels in the environmental system. This prediction algorithm is applied to the one-step forward prediction of atmospheric CO concentration by using real observed data. Computer simulation proves that the power in the forward error sequences decreases as the number of stages in the lattice is increased.