PSO-BP网络模型在温室数据融合中的应用研究

Application research of PSO-BP neural network in data fusion of greenhouse

由于温室的众多要素之间是相互制约、互相配合的,为了形成一个准确、合理的判断,将PSO算法的全局优化能力和BP神经网络良好的非线性映射能力相结合,优化BP神经网络的权值和阈值,提出了一种基于PSO的BP网络数据融合算法,并利用该算法对温室多传感器(温度传感器、湿度传感器和光照度传感器)同时检测到的数据进行融合。仿真结果表明:基于PSO-BP网络的数据融合算法能够获得温室准确有效的信息,提高温室控制的有效性与准确性。

Many elements of the greenhouse are constraint and coordinate,in order to make an effective, rational decision,the global optimization capability of PSO algorithm and the nonlinear reflecting capability of BP neural network are applied to optimizate the weights and thresholds of BP neural network and fuse the data synchronously detected by multi-sensors（ temperature sensors ,humidity sensors and light sensor）. The result of simulation shows that the data fusion algorithm of BP neural network based on PSO can get accurate information of the greenhouse, and improve the accuracy and effectiveness of greerihouse control.