Data from sensor array are often arranged in three-dimension as sample × time × sensor. Traditional methods are mainly used for two-dimension data. When such methods are applied, some time-profile information will lost. To acquire the information of samples, sensors and times more exactly, parallel factor analysis (PARAFAC) is investigated to deal with three-way data array. Through the analysis and classification of three kinds of oil odor samples, the performance of PARAFAC in gas sensor array signal analysis is verified and validated.

Magnetic anomaly detection is a passive method for detection of a ferromagnetic target, and its performance is often limited by external noise with a power spectral density of 1/fa, (0<a<2). In consideration of this kind of noise is non-stationary, self-similarity and long-range correlation, an effective noise reduction method based on the wavelet transform is proposed in this paper. The proposed method is only take one parameter into account, while the hard thresholding and soft thresholding methods utilize the relationship of the variance of the noisy signal. The simulation results show that the performance of our proposed method is superior to that of other methods.

Magnetic anomaly detection is a passive method for detection of a ferromagnetic target, and its performance is often limited by external noise with a power spectral density of 1/fa, (0<a<2). In consideration of this kind of noise is non-stationary, self-similarity and long-range correlation, an effective adaptive detection based on the wavelet transform is proposed in this paper. The discrete wavelet transform of the noisy signal is chosen as the inputs of the adaptive whitening filter, and then decomposed by the orthonormal basis functions (OBFs) and the energy signal was taken out for threshold detection. The simulation results show that the algorithm improves the effect for detecting weak magnetic anomaly signal contaminated by 1/fanoise.