燕子掌微弱电信号的计量研究

Testing and measurements of weak electrical signals in Crassula portulacea

利用自制的屏蔽室和屏蔽箱以及自制的铂金传感器接触式微弱电信号测试仪,初次获得了燕子掌(Crassula portulacea)自适应微弱电信号特性.采用小波软阈值消噪法对测试电信号进行消噪,并进行了时间序列的高斯径向基函数(RBF)神经网络预测.结果表明,采用RBF人工神经网络对植物微弱电信号进行短期预测是可行的.实现预测是在温室和/或塑料大棚生产中建立植物自适应智能化自动控制系统的关键环节.

The character of original weak electrical signals in Crassula portulacea was tested by a touching testing using platinum sensors in a system of self-made double shields. Tested data of electrical signals were denoised by the wavelet soft threshold using Gaussian radial base function （RBF） as the time series at a delayed input window chosen at 50. An intelligent RBF forecasting system was set up to predict signal in the plant. Results show that it is feasible to predict the weak electrical signal in the plant. Predicting data can be used as a key tache for constructing an intelligent automatic control system based on adaptive characteristics of plants to achieve energy savings of agricultural production in greenhouses and/or plastic lookums.