In this paper we evaluate the presence of anomalies in the Engine Coolant Temperature (ECT) sensor operation by collecting telemetry data of a single car in two different operational modes. The proposed approach has evaluated ten different one-class classifiers in three different anomaly levels, defined from the sensor's malfunctioning. Based on the results from the experimental data, the evaluation has shown: the One-Class Support Vector Machine with third- degree polynomial kernel function as the best anomaly detection technique for the vehicle operation in movement trajectory and the k-nearest neighbor as the best technique for the vehicle stopped, but with the engine running.
Detecting Anomalies in the Engine Coolant Sensor Using One-Class Classifiers
2019-09-01
269039 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
ENGINE COOLING SYSTEM HAVING COOLANT TEMPERAUTRE SENSOR
Europäisches Patentamt | 2017
|Online Contents | 1993
Effects of Nanofluid Coolant in a Class 8 Truck Engine
SAE Technical Papers | 2007
|Kraftfahrwesen | 1993
SAE Technical Papers | 2002