A Worst Case Benchmark Problem to Validate Voter Logic

TLDR: This research proposes a strategy that finds a worst case scenario for an autopilot namely using an orthogonal array based algorithm and Design of Experiments (DOE) and proves to improve the situation drastically.

Abstract: Aircraft flight control systems consist of primary controls and secondary control systems. Primary flight control systems provide the operation for the airplane's elevator, aileron rudder and horizontal stabilizer trim actuator. Simulink models are used to design and simulate such systems. Sensor failures happen very often and the control system is designed to be robust against sensor failures. Voting logic is used to select a good sensor in a multi sensor environment. The design of voter logic requires scenarios which provide catastrophic disastrous situations to study their effectiveness. This research proposes a strategy that finds a worst case scenario for an autopilot namely using an orthogonal array based algorithm and Design of Experiments (DOE). The voter logic is tested against such a situation and proves to improve the situation drastically.