提出用人工神经网络预报赤潮多发海区浮游植物生长趋势 ,并以 1998年 4～ 5月大亚湾澳头港实际监测数据为对象 ,以采样深度、水温、盐度、pH、DO、COD、浊度、营养盐、叶绿素a、微量元素、总碱度、气温、气压、风速、风向、光照、潮汐、总细胞密度等为参数 ,试验了人工神经网络的效果 .结果表明 ,人工神经网络可望成为赤潮预报的有效方法 The artificial neural network is used to forecast the phytoplankton growth trend in the frequent seafloor red tide. According to the actual monitoring data from April to May 1998 in the port of Daya Bay, the sampling depth, water temperature, salinity, pH, DO, COD, turbidity, Nutrient, chlorophyll a, trace elements, total alkalinity, temperature, barometric pressure, wind speed, wind direction, light, tide and total cell density were used as the parameters to test the effect of artificial neural network.The results showed that artificial neural network Effective method of forecasting