For absolute process safety in diverse machine applications, timely and reliable anomalous behavior detection is very crucial. Different machine applications have different normal behavior patterns and safety standards thus require adjustable and adaptive anomaly detection techniques. In this paper an autonomous behavior modeling approach for anomaly detection has been presented. In this approach time segmented vibration signals from the machines are transformed into spectral contents. After normalization, these frequency domain contents are divided into weighted frequency bins and then Gaussian models are achieved for these frequency bins over the entire training set. Using summation rule on the outputs of Gaussian models a single indicative measure of the machine health: normality score is obtained. The sensitivity of the normality score and anomaly detector towards potential anomalous signals can be controlled by using different number of bins and weights. Suitable parameters values, number of bins and weights profile, for anomaly detector model are selected autonomously using minimum value of the cost function. The increase of normality score of this model above a certain threshold is considered an alarm indicating anomalous behavior. Thus the proposed method enables us to achieve autonomously a suitable anomaly detection model with suitable parameters with controlled sensitivity during the test phase.
Safety has always been of vital concern in both industrial and home applications. Ensuring safety often requires certain quantifications regarding the inclusive behavior of the system under observation in order to determine deviations from normal behavior. In machine health monitoring, the vibration signal is of great importance for such measurements because it includes abundant information from several machine parts and surroundings that can influence machine behavior. This paper proposes a unitary anomaly detection technique (UAD) that, upon observation of abnormal behavior in the vibration signal, can trigger an alarm with an adjustable threshold in order to meet different safety requirements. The normalized amplitude of spectral contents of the quasi stationary time vibration signal are divided into frequency bins, and the summed amplitudes frequencies over bin are used as features. From a training set consisting of normal vibration signals, Gaussian distribution models are obtained for each feature, which are then used for anomaly detection.