Failure Accommodation in Digital Flight Control Systems Accounting for Nonlinear Aircraft Dynamics
Raymond C. Montgomery, Douglas B. Price · Journal of Aircraft · 1976
This paper presents a design method for optimal redundancy management for nonlinear systems with application to highly maneuvering aircraft. The approach taken is based on selecting the failure states to be covered by the system design and constructing a cost function that represents the cost of making an incorrect decision. The decision logic that minimizes the cost requires a bank of extended Kalman filters running in parallel. This produces a severe computational requirement. To reduce this requirement, a suboptimal logic is developed based on using a nonlinear single-stage prediction algorithm in the filters with filter gains and decision logic selected using steady-state results obtained from a linearization of the vehicle and sensor dynamics. The design process then is applied to designing a redundancy management system for the F8-C aircraft. The performance of the system is indicated with results of a nonlinear, piloted, six-degree-of-freedom simulation of the F8-C aircraft. Results indicate that the system is superior in failure detection to a system using the same structure but using a linear single-stage prediction algorithm in the filters.