Simultaneous Determination of p-Nitrochlorobenzene and o-Nitrophenol in Mixture by Single-Sweep Oscillopolarography Based on Cascade Neural Network
By combining the improved wavelet neural network and BP neural network, a new structure based on mixed cascade neural network was established. The novel cascade neural network has been used to the oscillopolargriphic signals analysis. By the figure fitting and parameters extracting, we realized the prediction of the simulation samples.The training speed and the predication accuracy can be enhanced by optimizing the network structure and parameters. The result of concentration prediction is satisfied . The method has been applied to the simultaneous determination of p- Nitrochlorobenzene (p-NCB) and o-Nitrophenol (o-NP) in simulation samples with satisfactory results. The Relative error and Recovery of p-NCB、o-NP were 3.76%、96.2%； 4.05%、96.0%, respectively. This novel cascade neural network combines the advantage of wavelet neural networks and BP neural networks, and performs its own functions respectively. It has shown a unique advantage in the overlap peak analyze.
S. Duan et al., "Simultaneous Determination of p-Nitrochlorobenzene and o-Nitrophenol in Mixture by Single-Sweep Oscillopolarography Based on Cascade Neural Network", Advanced Materials Research, Vols. 217-218, pp. 1469-1474, 2011