Multi-influence factor prediction for water bloom based on multi-sensor system

Authors

  • Li Wang
  • Xiaoyi Wang School of Computer and Information Engineering, Beijing Technology and Business University, Beijing, China
  • Jiping Xu School of Computer and Information Engineering, Beijing Technology and Business University, Beijing, China
  • Yan Shi School of Computer and Information Engineering, Beijing Technology and Business University, Beijing, China
  • Jiabin Yu School of Computer and Information Engineering, Beijing Technology and Business University, Beijing, China

Keywords:

prediction model, water bloom, parameter estimation

Abstract

This paper proposes a new multi-influence factors prediction method for water bloom prediction based on a remote monitor system and multi-sensor data taking into account the integrated effect of multiple influential factors along with the periodicity and random effect of environmental variables. Valid and accurate water-bloom prediction can be obtained by combining various multidimensional time series methods. Comparing the proposed model based on multi-sensors data to a traditional one-dimensional time series model based on one-sensor data, it has been found that a multidimensional model is a useful and accurate model for establishing multiple influential factors time series of water bloom. The optimum model can be used not only to predict water bloom but also to determine the period and random change rule of multiple influential factors.

Downloads

Published

2014-06-04