This repo dedicates to codes for assessing the reliability of air quality sensor network
The file DS.py contains implementation of Dempster-Shafer algorithm. It is used to derive the probability of events, scenarios based on the given sampling matrix and feature matrix. The constructor of class DempsterShafer is initiated by passing the sampling and feature matrices with the type numpy.array. The method result of the class abstracts the Dempster-Shafer combination rule and returns the probability of the hypotheses in the feature matrix.
The plots illustrating results from the reliability assessment of Dempster-Shafer are encoded in ``.
The probability assigned to each hypothesis is plotted as 100% stacked graph.
Probability of the normal operation is plotted for comparison.
Switching is performed based on the highest probability of normal operation within all sensors of the dependable system. A form of Gantt chart is plotted to visualize the time interval when a sensor is selected.
In this graph, a continuous data flow of measured parameters is established through the switching mechanism. A continuous flow is constructed by stitching intervals of data from different sensors in the dependable system.