VISTA X-62A Model Following Algorithm Overview

The United States Air Force Test Pilot School (USAF TPS) operates a uniquely modified F-16 for research and development of the world’s most advanced flight control systems. This aircraft is known as the Variable In-Flight Simulator Test Aircraft or VISTA. In the summer of 2021, VISTA was redesignated from an NF-16D to the X-62A to highlight its unique capabilities to perform advanced experimental flight tests and flight simulation of other aircraft. VISTA is able to achieve these flight tests through a simulation using a technique called Model Following. As part of an upgrade of the VISTA aircraft, known as GEN2020, Lockheed Martin’s Skunk Works® developed new capabilities of the Model Following Algorithm (MFA) that has provided for the ability to drive VISTA X-62A to simulate the dynamics of another aircraft through the control augmentation process of Non-Linear Dynamic Inversion. The MFA architecture consists of four major components; 1) The Commands subsystem, which contains the desired dynamics that VISTA is to simulate represented by a closed loop simulation of a selected aircraft. 2) The Regulator subsystem, which regulates the Control Variable errors that increment the desired dynamics. 3) The On-Board Model subsystem, which calculates the current accelerations and control derivative matrix (i.e., the vehicles B-matrix). And 4) The Control Allocator subsystem, which converts the desired dynamics into X-62A effector commands. The new MFA allows for multiple vehicle model types to be implemented by the researcher which are then broken down into two different components within the MFA called the Research Control Laws (RCLAWs) and Research Plant Models (RPMs). Then, the dynamic inversion technique is applied to 5 degrees of freedom: axial, heave, roll, pitch, and yaw. No lateral side-force degree of freedom is included as the X-62A does not have a direct side-force effector. A specific RPM architecture, the Quasi-Linear Parameter Varying, is discussed in detail in this paper, as well as the effectiveness and limitations of the MFA implementation via 1/s analysis and the real-time attainable moment assessment.