本文利用干扰效应并采用模拟退火 误差反向传播人工神经网络算法辅助原子吸收分光光度法在单波长下对粮谷中痕量镉和砷的同时测定时行了研究 ,取得了满意结果。标准加入法的回收率分别为 91%～98%和 87%～ 92 % ,7次平行测定结果的相对标准偏差均小于 4%。该法快速、简单 In this paper, the interference effects and simulated annealing error backpropagation artificial neural network algorithm assisted by atomic absorption spectrophotometry at a single wavelength of trace amounts of trace amounts of cadmium and arsenic in the study, and achieved satisfactory results. The recoveries of standard addition method were 91% -98% and 87% -92%, respectively. The relative standard deviations of seven parallel determinations were less than 4%. The method is fast and easy