模拟退火-BP神经网络法测定无机铬的形态  

Speciation of Chromium by Simulated Annealing-BP Neural Network

作　　者：林亚萍[1] 金继红[1] 

机构地区：[1]中国地质大学武汉材料科学与化学工程学院,武汉430074

出　　处：《分析科学学报》2006年第1期62-64,共3页Journal of Analytical Science

摘　　要：提出了用神经网络方法测定无机铬的形态。在Back Propagation(BP)算法的基础上,吸取了模拟退火算法的优点,解决了BP算法容易陷入局部极小点的问题。经实际样品分析,结果满意。This paper presents an artificial neural network for the speciation of Chromium. The method takes advantage of global minimum property of Simulated procedure on the basis of BP algorithm, and can jump out of the local minimum and converge to the global minimum. The proposed method was applied to the determination of synthetic sampled with satisfactory results.

关 键 词：铬 形态分析 神经网络 模拟退火 同时测定 

分 类 号：O657.3[理学—分析化学]