Mario Innocenti

Autonomous formation flight

This article describes an approach to close-formation flight of autonomous aircraft. A standard LQ-based structure was synthesized for each vehicle and for formation position error control using linearized equations of motion and a lifting line model of the aircraft wake. We also consider the definition of a formation management structure, capable of dealing with a variety of generic transmission and communication failures among aircraft. The procedure was developed using a decentralized approach and relies on the Dijkstra algorithm. The algorithm provides optimal path information sequencing in the nominal case, as well as the redundancy needed to accommodate failures in data transmission and reception. Several simulations were carried out, and some of the results are presented. The overall scheme appears to be a valuable starting point for further research, especially specialization to situations representing more detailed and operational failures.

A sliding mode missile pitch autopilot synthesis for high angle of attack maneuvering

A new approach to the synthesis of longitudinal autopilots for missiles flying at high angle of attack regimes is presented. The methodology is based on sliding mode control, and uses a combination of aerodynamic surfaces and reaction jet thrusters, to achieve controllability beyond stall. The autopilot is tested on a small section of the flight envelope consisting of a fast 180/spl deg/ heading reversal in the vertical plane, which requires robustness with respect to uncertainties in the systems dynamics induced by large variations in dynamic pressure and aerodynamic coefficients. Nonlinear simulation results show excellent performance and capabilities of the control system structure.

Kalman filters and neural-network schemes for sensor validation in flight control systems

Detection, identification, and accommodation of sensor failures can be a challenging task for complex dynamic systems. This paper presents the comparison of two different approaches for the task of sensor failure detection, identification, and accommodation in a flight control system assumed to be without physical redundancy in the sensory capabilities. The first approach is based on the use of a set of online learning neural networks; the second approach is based on the use of a bank of Kalman filters. The objective is to evaluate the robustness of both schemes; the comparison is performed through testing of the schemes for several types of failures presenting different level of complexity in terms of detectability. The required computational effort for both schemes is also evaluated. For each of these failure types this comparison is performed at nominal conditions, that is with the system model and its noise perfectly modeled (with the Kalman filter scheme performing at nominal conditions), and at conditions, where discrepancies occur for the modeling of the system as well as the system and measurement noises. While the Kalman-filter-based scheme takes advantage of its robustness capabilities, the neural-network-based scheme, starting from a random numerical architecture, relies on the learning accumulated either online or from off-line simulations. The study reveals that online learning neural architectures have potential for online estimation purposes in a sensor validation scheme, particularly in the case of poorly modeled dynamics.

Neural-Network-Based Scheme for Sensor Failure Detection, Identification, and Accommodation

This paper presents a neural-network-based approach for the problem of sensor failure detection, identification, and accommodation for a flight control system without physical redundancy in the sensors. The approach is based on the introduction of on-line learning neural network estimators. For a system with n sensors, a combination of a main neural network and a set of n decentralized neural networks achieves the design goal. The main neural network and the ith decentralized neural network detect and identify a failure of the ith sensor, whereas the output of the ith decentralized neural network accommodates for the failure by replacing the signal from the failed ith sensor with its estimate. The on-line learning for these neural network architectures is performed using the extended back-propagation algorithm. The document describes successful simulations of the sensor failure detection, identification, and accommodation process following both soft and hard sensor failures. The simulations have shown remarkable capabilities for this neural scheme.

Experimental application of extended Kalman filtering for sensor validation

A sensor failure detection and identification scheme for a closed loop nonlinear system is described. Detection and identification tasks are performed by estimating parameters directly related to potential failures. An extended Kalman filter is used to estimate the fault-related parameters, while a decision algorithm based on threshold logic processes the parameter estimates to detect possible failures. For a realistic evaluation of its performance, the detection scheme has been implemented on an inverted pendulum controlled by real-time control software. The failure detection and identification scheme is tested by applying different types of failures on the sensors of the inverted pendulum. Experimental results are presented to validate the effectiveness of the approach.

A synthetic environment for dynamic systems control and distributed simulation

Rapid prototyping and controlled motion evaluation of complex human-machine interfaces, from nuclear plant operation panels to deep submerged underwater vehicles to advanced airplane cockpits, require hardware-in-the-loop, man-in-the-loop, and software integration. What appears to be needed is specific software to give designers tools for analyzing and simulating complex and integrated projects. The research software described in this article promises to fill that need, providing a new synthetic environment for simulation and control synthesis of dynamic systems. The article addresses problems of high performance, realistic environments, and vehicle simulation, with particular attention to synthetic world creation and visualization. The new software is capable of handling most of the simulation and visualization requirements highlighted.

Online Learning Neural Architectures and Cross-Correlation Analysis for Actuator Failure Detection and Identification

This paper describes a study related to the testing and validation of a neural-network based approach for the problem of actuator failure detection and identification following battle damage to an aircraft control surface. Online learning neural architectures, trained with the Extended Back-Propagation algorithm, have been tested under nonlinear conditions in the presence of sensor noise. In addition, a parametric study has been conducted that addresses the selection of ‘near optimal’ neural architectures for online state estimation purposes. The Failure Detect-ability/False Alarm Rate ratio problem has also been considered in this study. The testing of the approach is illustrated through typical highly nonlinear dynamic simulations of a high performance aircraft.

Vision Algorithms for Formation Flight and Aerial Refueling with Optimal Marker Labeling

This pa per presents the experimental results of an artificial vision system prototype for application to unmanned formation flight and aerial refueling. In the former, a camera on the wingman captures leader images, estimating the relative position; in the latte r, using probe -and -drogue refueling, the aircraft camera acquires basket images, and from that estimating the relative position. Position estimation is based on localization of infrared markers which hav e a known geometry distribution over the leader airfr ame or drogue body . Experimental results using a low cost simulated formation flight setup are shown, to validate the procedure.

Fast unmanned vehicles task allocation with moving targets

This paper presents a fast algorithm for allocation at mission-time of moving targets to a group of unmanned vehicles. A fleet of UAVs must fly through a known environment to reach partially unknown locations, or targets, where three tasks: identification, attack and verification must be performed sequentially. The total mission cost is identified to be the sum of the total times that the UAVs spend completing their tasks, while respecting the task priorities and ensuring the task precedence laws. The problem is solved in two steps; the first step is performed off-line and is the most computationally intensive: the environment is subdivided into triangle-shaped areas forming the tessellation graph (TG), and the shortest path between each two vertexes couples of the plane is computed using the all-pairs-nodes Dijkstra algorithm. The second step, at mission-time, regards management of moving targets and adaptation to the results of the identification phase. Optimal task assignment is performed using the Hungarian algorithm; exact path lengths between vehicles and targets are computed from the off-line computed Dijkstra paths. One parameter is available to tune the optimal task allocation algorithm with respect to desired aggressive/selfish or cooperative behavior.

Sensor validation using hardware-based on-line learning neural networks

The objective of this document Is to show the capabilities of parallel hardware-based on-line learning neural networks (NNs). This specific application is related to an on-line estimation problem for sensor validation purposes. Neural-network-based microprocessors are starting to be commercially available. However, most of them feature a learning performed with the classic back-propagation algorithm (BPA). To overcome this lack of flexibility a customized motherboard with transputers was implemented for this investigation, The extended BPA (EBPA), a modified and more effective BPA, was used for the on-line learning, These parallel hardware-based neural architectures were used to implement a sensor failure detection, identification, and accommodation scheme in the model of a night control system assumed to be without physical redundancy in the sensory capabilities. The results of this study demonstrate the potential for these neural schemes for implementation in actual flight control systems of modern high performance aircraft, taking advantage of the characteristics of the extended back-propagation along with the parallel computation capabilities of NN customized hardware.