| Forecasting the coverage area and the pollution level is an important procedure of biological protection. Accurately predict the concentration and activity of warfare agents is useful to large-scale field disinfectant. But for a long time, the traditional method has the characteristics of qualitative and uncertainty. Therefore, the effective and feasible measurements are urged to build.In this study, Bacillus subtilis had been used as the test bacteria, the methods of biological information technology, electron microscopy, aerosol technology, viable count analysis and computer tools for intelligence analysis have been employed to construct trhe artificial intelligent forecasting modle for activity of spore staying on object sureface. The results are as follows:1. There are more similarities among Bacillus cereus, Bacillus subtilis and Bacillus anthracis in their genetic homology, structures and resistance to heat, UVC and Chlorine disinfectant. As a good substitution for Bacillus anthracis spores, Bacillus cereus and Bacillus subtilis spores could be employed in experiments.2. In a simulated environment of natural surface (leaf, stone, tile, cloth), the resistance of spore decreased with the exposure to temperature, humidity and UVC irradiation. UVC exposure was the most sensitive factor to reduce the activity of spore. And the spores staying on the leaf had stronger resistance than other surfaces. .3. According to the mechanism of Matlab neural network, the prediction model of resistance of spores staying on object sureface has been studied. Based on the characteristics of research purposes, simulated environmental conditions and training smooth curves, five input neurons, eight hidden layer nodes and one output neurons have been designed. The 'Tansig', and 'purelin' were transfer function. The,"trainlm"was training function. The network iteration was 100. The forecasting efficiency to retrospective data was 95% and the efficiency to prospective data was 85%. |
PDF Full Text Request