Kalman Filtering Method For The Inverse Problem Of Groundwater Pollution In The Liao River Basin

Liao river basin was in groundwater as the main source of drinking water in north China basin, and over the years due to the unreasonable development and utilization of groundwater resources, because a series of environmental geological problems, the "three nitrogen" pollution problem is particularly acute. In many model of groundwater pollution and the practical application, are greatly influenced by model parameters and the actual error, it is hard to get ideal results. Kalman filtering as a method of optimal state estimation is more and more attention in groundwater state estimation and parameter identification for its strong ability of noise suppression.In this paper, with the "three nitrogen" solute transport model which was established in the research area of the LiGuan water source area of the fort in Liao River Basin, the application of kalman filtering technology, combined with the finite difference method, and parameters in real-time water quality prediction in the studied area. The paper mainly includes two aspects: On the one hand, Parameter identification. The kalman filter technology combined with solute transport model, the pollutant migration state estimator model parameters as stochastic process, according to the real-time input of solute concentration measured information, the use of kalman filtering algorithm for parameter estimation. Due to the traditional kalman filter is based on an algorithm of linear system, requirements of linear state equation and measurement equation and as a measurement equation is nonlinear partial differential equation, need to be linearized processing first. In this paper, using the finite difference method, is widely popular extended kalman filtering algorithm in order to ensure accuracy under the premise of several basic same, avoiding the tedious Jacobian matrix and Hessian matrix calculation, reduces calculation difficulty. On the other hand, the water quality concentration and the related parameters are predicted simultaneously. Kalman filter algorithm is not only an effective parameter identification method, and the equation of state is augmented, parallel estimation of state and parameter, and achieves the goal of real-time estimation of pollutant concentrations and related parameters.