A Study on Dynamic Model of Dissolved Oxygen in Fishpond

A pond is a complex ecosystem. Each factor in the system must be surveyed and supervised dynamically to increase the production and lower the cost of aquiculture. The dissolved oxygen (DO) is one of the important parameters describing pond ecosystem. DO changes continuously with a period of 24 hours,so that the ARMA(p,q) model of pond DO may be founded by the method of time series analysis. The modeling principle and steps were explained with an example, and the model, which is founded by the method of pattern recognition with autocorrelation function and model parameters is estimated by the least square method. Results fit well with real measurement and can be used as real-time forecast of pond DO. Compared with other methods,this method is simple, well real-time, highly accurate and has the characteristics of dynamic survey and supervision of multi -ponds DO suitable to be used with computer.