基于优化深度置信网络的多传感器水质监测研究

Research on multi-sensor water quality monitoring based on optimized deep confidence networks

摘要:

为了满足多种水环境的大范围、精准监测需求,提出了基于优化深度置信网络的多传感器水质监测方法。设置水质监测标准,作为水质等级的判定条件。优化设计水体温度、PH值、溶解氧、浊度等传感器设备,利用优化深度置信网络选择多传感器的安装位置。利用多传感器采集水环境数据并完成融合处理,通过多个水质监测指标的计算以及与设置标准的比对,得出多传感器水质监测的可视化输出结果。通过性能测试实验得出结论：优化设计方法的水质监测范围为2041.79平方千米,浊度、pH值、溶解氧和氨氮浓度指标的监测误差分别为0.005FTU、0.07、0.05mg/L和0.007mg/L,均低于传统方法,且满足预设条件。

Abstract:

In order to meet the needs of large-scale and accurate monitoring of various water environments, a multi-sensor water quality monitoring method based on optimized deep confidence networks is proposed. Set water quality monitoring standards as criteria for determining water quality levels. Optimize the design of sensor equipment for water temperature, pH value, dissolved oxygen, turbidity, etc., and use the optimized deep confidence network to select the installation location of multiple sensors. Utilize multiple sensors to collect water environment data and complete fusion processing. Through the calculation of multiple water quality monitoring indicators and comparison with set standards, obtain visual output results of multi-sensor water quality monitoring. Through performance testing experiments, it was concluded that the water quality monitoring range of the optimized design method is 2041.79 square kilometers, and the monitoring errors of turbidity, pH value, dissolved oxygen, and ammonia nitrogen concentration indicators are 0.005FTU, 0.07, 0.05mg/L, and 0.007mg/L, respectively, which are lower than traditional methods and meet the preset conditions.