大鹏湾夜光藻种群动态机理模型辨识

THE IDENTIFICATION OF A MECHANISM MODEL ON THE POPULATION DYNAMICS OF NOTILUCA SCINTILLANS IN THE DAPENG BAY IN THE SOUTH CHINA SEA

本文建立了大鹏湾夜光藻种群动态机理模型。以1990年4月2日至1990年6月10日期间每隔2d采样1次的数据为基础在浮游植物、温度、溶解氧(DO)、可溶性无机磷(DIP)、可溶性无机氮(DIN)(包含硝态氮和亚硝态氲)和叶绿素a等6个因子中,辨识出夜光藻应以温度、可溶性无机磷和可溶性无机氮为控制变量,并以递阶的方式对夜光藻种群动态进行回报。另在模型辨识的过程中引入自回归模型阶次的辨识,从而建立3个断面6个层次的6个自回归与非线性回归并联混合回归模型。计算结果表明,回报值对实测值的拟合率均在87％以上。根据模型指出夜光藻生长的最适温度(17—22℃)、可溶性无机氮的浓度(100—500μg/l),可溶性无机磷的浓度(10—40μg/l)的范围。

A mechanism model for simulating the population dynamics of Noctiluca scintillans in Dapeng Bay is described.The data are based on the samples taken at intervals of fortnight from 2 April to 10 June 1990. Amongst the factors of phytoplankton, temperature, dissolved oxygen (DO),dissolved inorganic phosphate (DIP), dissolved inorganic nitrogen (DIN, including NO3--N and NO2--N) and chlorophyll-a, temperature, DIP and DIN are identified as tha variables of controlling the growth of Noctiluca scintillans. By applying hierachy to the reciprocation model and introducing auto-regression ordering into the model identification, a mixed simulaneous regression model derived from six auto-regression and non-linear regression models was established. The formula is.The fitting degree between the prediction and field data is more than 87%. The optimum temperature, concentraions of DIN and DIP are 17-20℃,100-500 μg/1 and 10-40 μg/1, respectively.