基于光谱技术的土壤环境污染物成分检测方法

A Method of Soil Environmental Pollutant Composition Detection Based on Spectral Technology

土壤污染会严重威胁人类生活,因此有必要对土壤环境污染物成分进行检测,以寻求正确的处理方法。但目前检测手段较为落后,使得污染物成分检测精度较低。为解决这一问题,本文提出了一种基于光谱技术的土壤环境污染物成分检测方法。首先使用光谱仪器扫描土壤样品,再利用差分吸收光学光谱技术测量土壤环境污染物含量,将该含量作为改进深度学习网络的输入向量;然后将粒子群算法优化权值与Softmax分类器相结合,获得的深度学习自动编码器,对输入向量编码进行解码,并训练输入向量样本合集,至此,完成基于深度神经网络的土壤污染物检测模型的构建。实验结果证明,该方法具有较强的样本数据分类能力,具有较高的污染物成分检测精度。

Soil pollution will seriously threaten human life.Therefore,it is necessary to detect the components of pollutants in soil environment in order to find correct treatment methods.However,the current detection method are relatively backward,which leads to the lower detection accuracy for pollutant components.In order to solve this problem,a method for detecting the components of soil environmental pollutants based on spectral technology is proposed in this paper.Firstly,the soil samples were scanned using a spectral instrument,and then the soil environmental pollutants was measured using differential absorption optical spectroscopy,which was used as the input vector for the improved deep learning network.In addition,the optimized weight of particle swarm optimization algorithm is combined with Softmax classifier to obtain the depth learning automatic encoder,decode the input vector compilation,and train the input vector sample set.Therefore,the construction of soil pollutant detection model based on depth neural network is completed.The experimental results show that this method has strong sample data classification ability and high accuracy of pollutant components.