论文部分内容阅读
根据苯并(a)芘在特定光的照射下能够发出荧光的特性,采用紫外荧光法检测苯并(a)芘的浓度,构建了检测苯并(a)芘浓度的实验系统。为了精确地检测出苯并(a)芘的浓度,需要将实验中所得数据中所包含的噪声滤除掉,得到有效的荧光信号。将归一化最小均方自适应滤波(NLMS)的原理应用到光谱数据的噪声处理中,并将其与小波去噪的效果相比较,用Matlab仿真出去噪效果图,从不同的角度分析去噪效果。经过对比得出,用归一化最小自适应滤波能够继续保持信号的非平稳特性,达到更加理想的去噪效果,去噪性能更优,此种方法适用于荧光法检测物质浓度的信号处理中。
According to the fact that benzo (a) pyrene can emit fluorescence under the irradiation of specific light, the concentration of benzo (a) pyrene was detected by ultraviolet fluorescence spectroscopy, and the experimental system for detecting the concentration of benzo (a) pyrene was established. In order to accurately detect the concentration of benzo (a) pyrene, it is necessary to filter out the noise contained in the data obtained in the experiment to obtain an effective fluorescence signal. The principle of normalized least mean square adaptive filtering (NLMS) is applied to the noise processing of spectral data, and compared with the effect of wavelet denoising. The effect diagram of denoising is simulated by Matlab, and analyzed from different angles Noise effect. After comparison, it can be concluded that the normalized minimum adaptive filter can keep the non-stationary characteristics of the signal to achieve a more ideal denoising effect and better denoising performance. This method is suitable for the signal processing of fluorescent substance detection concentration .