SENSOR DEVICE FOR ANOMALY DETECTION THROUGH MACHINE LEARNING AND RELATED CONTROL METHOD
Sensor device with a microcontroller unit and a sensor including a transducer, which is coupleable to a device and generates a signal indicative of a physical quantity, and a processing circuit including: a conversion stage which generates samples of the physical quantity; a data generation stage which generates data vectors as a function of the samples, each data vector being formed by programmable quantity values; and a decision stage. The microcontroller unit programs the decision stage so that it classifies the data vectors by executing a decision tree having a structure and thresholds. In a configuration mode, the microcontroller unit programs the data generation stage; in a calibration mode, the microcontroller unit acquires a corresponding set of data vectors, determines, for each programmable quantity, a corresponding range of admissible values and programs the thresholds as a function of the ranges of admissible values; in a detection mode, the decision stage classifies the data vectors, by executing the decision tree on the basis of the programmed thresholds.
1 . A sensor device comprising a microcontroller unit and a sensor comprising a transducer coupleable to a device to be monitored and configured to generate at least one signal indicative of a physical quantity that depends on the functioning of the device to be monitored, the sensor further comprising a processing circuit comprising:
a conversion stage configured to generate a succession of samples of the physical quantity as a function of the signal;
a data generation stage configured to generate, in a manner programmable by the microcontroller unit, a succession of data vectors as a function of the succession of samples, each data vector being formed by a corresponding set of values of a number of programmable quantities; and
a decision stage;
and wherein the microcontroller unit is configured to program the decision stage in such a way that the decision stage is configured to classify the data vectors by executing a decision tree having a structure and a plurality of thresholds; the sensor device being further configured to operate:
in a configuration mode, wherein the microcontroller unit programs the data generation stage; and subsequently
in a calibration mode, wherein the microcontroller unit acquires a corresponding set of data vectors, determines, for each programmable quantity and on the basis of the corresponding set of data vectors, a corresponding range of admissible values and subsequently programs the thresholds as a function of the ranges of admissible values; and subsequently
in a detection mode, wherein the decision stage classifies corresponding data vectors alternatively as normal or abnormal, by executing the decision tree on the basis of the programmed thresholds.
2 . The sensor device according to claim 1 , wherein the data generation stage comprises a filtering stage configured to generate at least a succession of filtered samples by applying a digital filter to the samples; and wherein the data generation stage is further configured in such a way that the data vectors are a function of the filtered samples; and wherein, when the sensor device operates in the configuration mode, the microcontroller unit programs the digital filter.
3 . The sensor device according to claim 1 , wherein the data generation stage further comprises a feature extraction stage configured to generate a succession of feature vectors that depend on the samples; and wherein the data generation stage is further configured in such a way that the data vectors are a function of the feature vectors; and wherein, when the sensor device operates in the configuration mode, the microcontroller unit programs the features.
4 . The sensor device according to claim 3 , wherein the feature extraction stage is configured to calculate the value of each feature of the feature vectors on the basis of a corresponding set of samples or a corresponding set of filtered samples; and wherein each data vector comprises a respective feature vector.
5 . The sensor device according to claim 1 , configured to couple to an external sensor configured to generate a succession of external samples indicative of a corresponding physical quantity; and wherein the data generation stage is configured in such a way that the data vectors also depend on the external samples.
6 . The sensor device according to claim 1 , wherein the transducer comprises at least one of an acceleration transducer or an angular speed transducer.
7 . The sensor device according to claim 1 , wherein the microcontroller unit is configured to detect an anomaly of the device to be monitored when a data vector is classified as abnormal.
8 . A method for controlling a sensor device comprising a microcontroller unit and a sensor comprising a transducer coupleable to a device to be monitored and configured to generate at least one signal indicative of a physical quantity that depends on the functioning of the device to be monitored, the method comprising:
generating, by the sensor, a succession of samples of the physical quantity as a function of the signal;
generating, by the sensor and in a manner programmable by the microcontroller unit, a succession of data vectors as a function of the succession of samples, each data vector being formed by a corresponding set of values of a number of programmable quantities;
programming the sensor, through the microcontroller, to classify the data vectors by executing a decision tree having a structure and a plurality of thresholds;
through the microcontroller unit, programming the generation of the succession of data vectors; and subsequently
acquiring, by the microcontroller unit, a set of data vectors, determining, for each programmable quantity and on the basis of the set of data vectors, a corresponding range of admissible values and subsequently programming the thresholds as a function of the ranges of admissible values; and subsequently
classifying, by the sensor, data vectors alternatively as normal or abnormal, by executing the decision tree on the basis of the programmed thresholds.
9 . The method according to claim 8 , further comprising generating at least a succession of filtered samples by applying a digital filter to the samples and wherein the data vectors are a function of the filtered samples.
10 . The method according to claim 8 , further comprising generating a succession of feature vectors which depend on the samples; and wherein the data vectors are a function of the feature vectors.
11 . The method according to claim 10 , further comprising calculating the value of each feature of the feature vectors on the basis of a corresponding set of samples or of a corresponding set of filtered samples and wherein each data vector comprises a respective feature vector.
12 . The method according to claim 8 , further comprising coupling the sensor device to an external sensor configured to generate a succession of external samples indicative of a corresponding physical quantity and wherein the data vectors also depend on the external samples.
13 . The method according to claim 8 , further comprising detecting, through the microcontroller unit, an anomaly of the device to be monitored, when a data vector is classified as abnormal.
14 . A collection of non-transitory computer-readable media storing content that, when executed by one or more processors, cause a method for controlling a sensor device to be performed, the sensor device comprising a microcontroller unit and a sensor comprising a transducer coupleable to a device to be monitored and configured to generate at least one signal indicative of a physical quantity that depends on the functioning of the device to be monitored, the method comprising:
generating, by the sensor, a succession of samples of the physical quantity as a function of the signal;
generating, by the sensor and in a manner programmable by the microcontroller unit, a succession of data vectors as a function of the succession of samples, each data vector being formed by a corresponding set of values of a number of programmable quantities;
programming the sensor, through the microcontroller, to classify the data vectors by executing a decision tree having a structure and a plurality of thresholds;
through the microcontroller unit, programming the generation of the succession of data vectors; and subsequently
acquiring, by the microcontroller unit, a set of data vectors, determining, for each programmable quantity and on the basis of the set of data vectors, a corresponding range of admissible values and subsequently programming the thresholds as a function of the ranges of admissible values; and subsequently
classifying, by the sensor, data vectors alternatively as normal or abnormal, by executing the decision tree on the basis of the programmed thresholds.
15 . The collection of non-transitory computer-readable media according to claim 14 , wherein the method further comprises generating at least a succession of filtered samples by applying a digital filter to the samples and wherein the data vectors are a function of the filtered samples.
16 . The collection of non-transitory computer-readable media according to claim 14 , wherein the method further comprises generating a succession of feature vectors which depend on the samples; and wherein the data vectors are a function of the feature vectors.
17 . The collection of non-transitory computer-readable media according to claim 16 , wherein the method further comprises calculating the value of each feature of the feature vectors on the basis of a corresponding set of samples or of a corresponding set of filtered samples and wherein each data vector comprises a respective feature vector.
18 . The collection of non-transitory computer-readable media according to claim 14 , wherein the method further comprises coupling the sensor device to an external sensor configured to generate a succession of external samples indicative of a corresponding physical quantity and wherein the data vectors also depend on the external samples.
19 . The collection of non-transitory computer-readable media according to claim 14 , wherein the method further comprises detecting, through the microcontroller unit, an anomaly of the device to be monitored, when a data vector is classified as abnormal.