In order to solve the problem that the actual description of the water bloom behavior is not entirely conform to reality and water bloom prediction is not accurate enough by existing algae growth dynamics model due to the neglect of model parameters change with time

this paper builds cyanobacteria feeding and nutrient cycling model

and proposes algae growth dynamics model with time-varying parameters based on time variation influences of water temperature and illumination on model parameters. The calibration method for constant parameters of the model is optimized based on genetic and numerical algorithm. And the model time-varying parameters is modeled and predicted by multivariate time series method. Nonlinear dynamic mechanism of cyanobacteria bloom behavior is analyzed by bifurcation theory and time varying system theory

and then a new method of cyanobacteria bloom prediction is put forward. The monitoring data analysis of the Taihu River Basin in Jiangsu shows that cyanobacterial growth dynamics model with time-varying parameters can reflect nonlinear dynamic characteristics of cyanobacteria bloom behavior in cyanobacteria growth time-varying system. The model is more consistent with the actual situation

and cyanobacteria bloom prediction result is more accurate.