Study of the test flow optimization method based on utility function with discounted reward factor

A novel software framework for radar fault diagnosis is proposed. The framework separates all diagnosis algorithms from the test flows in fault isolation process. In the framework, the test flow and the executing loop of diagnostic algorithms are parallel processing, and test flow can be optimized dynamically. Define a utility function with discounted reward factor. Based on the utility function and framework, the paper utilizes an example to explains how to search the best flow from candidate flows.