# Continuous Motion - Anomaly Detection

Using: Continuous motion recognition - Edge Impulse Documentation as reference and mobile phone as a data Acquisition device - trying to configure Anomaly Detection Block.

Feature names are not given in Anomaly Detection Settings, so do not know what to choose for accX RMS, accY RMS and accZ RMS axes,

Request help.

I can see all the values now. Thanks.

1 Like

Welcome to the forum @mohitjain512 Please include screenshots, and project details when asking questions so we can see what you mean.

Best

Eoin

In the analysis of anomaly readings on vibration data in the x, y, z axes, the following terms have the following meanings:

Entropy: Entropy is a measure of the randomness or disorder in a data set. In the context of vibrations, entropy is used to assess the complexity of vibration patterns and can be useful in detecting significant changes or anomalies in data.

Z-Cross (Zero Crossing): Z-Cross is a measure that counts the number of times a signal crosses the zero axis. In vibration analysis, the Z-Cross can be useful in identifying changes in vibration direction and detecting abrupt events or anomalies in the data.

M-Cross (Mean Crossing): The M-Cross is similar to the Z-Cross, but counts the number of times a signal crosses the mean or average of the data. Like the Z-Cross, the M-Cross can help identify significant changes in vibration direction and detect anomalies.

Skew (Skewness): Skewness or asymmetry is a statistical measure that describes the lack of symmetry in the distribution of data. In the context of vibrations, the skew can provide information about the shape of the data distribution and help identify possible anomalies or deviations from a symmetric distribution.

Kurtosis (Kurtosis): Kurtosis is a statistical measure that describes the shape of the data distribution relative to the normal distribution. A high value of kurtosis indicates a distribution with heavy tails or sharp peaks, while a low value of kurtosis indicates a more flattened distribution. In vibrations, kurtosis can help identify unusual events or significant spikes in the data.

Median: The median is the middle value of a set of data ordered in ascending or descending order. It is a measure of central tendency that is not affected by extreme values. In vibration analysis, the median can be used to characterize the central vibration level and help detect significant deviations.

Mean: The mean is the sum of all the values in a data set divided by the number of values. It is also a measure of central tendency and provides information about the average level of vibration in the x, y, z axes.

Standard Deviation: The standard deviation is a measure of dispersion that indicates how far individual values are from the mean. In vibration analysis, the standard deviation can be used to assess data variability and detect anomalies or significant changes.

Variance: The variance is the square of the standard deviation and provides a measure of the spread of the data around the mean. Like the standard deviation, the variance is used to assess vibration variability and detect abnormalities.

RMS (Root Mean Square): The RMS is a measure that represents the square root of the mean of the squared values of a data set. In vibration analysis, RMS is used to obtain a measure of the effective value of vibration, that is, the average magnitude of vibration in the x, y, z axes.

L0-n5: L0-n5 represents the 5th percentile level of the data. The 5th percentile is the value below which 5% of the ordered data falls. In vibration analysis, L0-n5 can be used to identify the lowest vibration levels and detect events or time periods where data is below a specific threshold.

L0-n25: L0-n25 represents the 25th percentile level of the data. The 25th percentile is the value below which 25% of the ordered data falls. In vibration analysis, L0-n25 can be used to identify low vibration levels and detect events or time periods where data is below a specific threshold.

L0-n75: L0-n75 represents the 75th percentile level of the data. The 75th percentile is the value below which 75% of the ordered data falls. In vibration analysis, L0-n75 can be used to identify high vibration levels and detect events or time periods where data exceeds a specified threshold.

L0-n95: L0-n95 represents the 95th percentile level of the data. The 95th percentile is the value below which 95% of the ordered data falls. In vibration analysis, L0-n95 can be used to identify high vibration levels and detect events or time periods where data exceeds a specified threshold.