Method of optimal sensor selection and fusion for heat exchanger fouling diagnosis in aerospace systems
A system and method that determines whether a heat exchanger within a complex networked system is fouling is provided. The system and method includes training classifiers indicative of a plurality of fouling conditions associated with the heat exchanger and testing the classifiers with optimal sensor data from optimal sensors to determine whether the fouling is being experienced by the heat exchanger.
1. A processor-implemented method of determining whether a heat exchanger is fouling, the processor-implemented method executable by a processor of a diagnosis system in communication with an environmental control system, the processor-implemented method comprising:
a) training, by the processor, classifiers to indicate a plurality of fouling conditions associated with the heat exchanger, the heat exchanger being within the environmental control system by utilizing a model to generate sensor data based on the plurality of fouling conditions and parametric combinations of ambient temperature and occupancy count, wherein the ambient temperature is a temperature external to and affects the operation of the environmental control system, the classifiers being encoded as computer executable instructions on a tangible computer readable memory that when executed by the processor causes the processor to carry out the computer executable instructions to indicate the plurality of fouling conditions associated with the heat exchanger, wherein the training of the classifiers further comprises performing an extraction of features from each sensor and processing the features by the classifiers to determine a fouling severity for each of the fouling conditions; and
b) testing, by the processor, the classifiers with sensor data from sensors to determine whether the fouling is being experienced by the heat exchanger.
2. The processor-implemented method of claim 1 , further comprises:
performing a selection of the sensors from a plurality of sensors, each of the sensors being associated with at least one component within the environmental control system.
3. The processor-implemented method of claim 2 , wherein the selection of the sensors is executed by an unsupervised embedded algorithm that relies on a filter algorithm to select a candidate list of sensors and then applies a K-means clustering algorithm to rank sensors.
4. The processor-implemented method of claim 1 , further comprises:
outputting a decision with respect to the testing of the classifiers to determine whether the fouling is being experienced by the heat exchanger.
5. The processor-implemented method of claim 1 , wherein the environmental control system is an air management system of an aircraft.