集约化水产养殖水质预警系统的设计与实现  

Design and Implementation of Intensive Aquaculture Water Quality Early Warning System

作　　者：于承先[1] 徐丽英[1] 邢斌[1] 王剑秦[1] 李道亮[1] 

机构地区：[1]中国农业大学信息与电气工程学院,北京100083

出　　处：《计算机工程》2009年第17期268-270,共3页Computer Engineering

基　　金：国家"863"计划基金资助项目(2007AA10Z238);国家科技支撑计划基金资助项目(2006BAD10A02-05)

摘　　要：针对集约化水产养殖场因水质导致的养殖风险问题,提出集约化水产养殖的水质预警系统。该系统利用系统动力学原理分析水中溶氧变化的影响因子,确定水质预测网络的输入输出量,通过反向传播神经网络优化算法和规则预警策略减少误报次数。实验结果证明,该系统实现水产养殖语音预警和短信预警,已在山东某水产养殖场应用,预警准确率为94.9%。Aiming at the problems of risk caused by water quality in aquaculture farm, this paper proposes a water quality early warning system for intensive aquaculture. It uses system dynamic principle to analyze the factors affecting dissolved oxygen in the water and confirm the input and output parameters of the water quality predicting network, and decrease the time of error alarm through an optimized back propagation neural network algorithm and a rnle based early warning strategy. Experimental result proves the system implements voice and short-message early warning for aquaculture, it is applied in aquaculture farm of Shandong province, and accuracy of the early warning is 94.9%.

关 键 词：集约化水产养殖 水质 预警系统 

分 类 号：TP311.1[自动化与计算机技术—计算机软件与理论]