Time-varying nonlinear modeling and analysis of algal bloom dynamics

The ecological dynamic models of algal bloom in existence are of poor environmental adaptability and hard to reflect nonlinear dynamic change of algal bloom formation mechanism. To solve this problem, time variable is applied to mechanism-based model according to algal bloom nonlinear dynamic in this paper, and mechanism feature of algal bloom is represented by time function model and effecting function model. Data-driven modeling approaches including tabu search and genetic algorithm are also adopted to optimize structure and parameter for the time-varying nonlinear modeling to improve environmental adaptability and accuracy of the model. High-precision numerical solutions of the optimized time-varying nonlinear model is obtained by fourth-order Adams predictor–corrector method. This method finally realizes effective prediction of algal bloom.