James C. Deckert

F-8 DFBW sensor failure identification using analytic redundancy

In this paper, we outline the structure of a sensor failure detection and identification system designed for the NASA F-8 DFBW aircraft. The system is for use in a dual-redundant environment, and it takes maximal advantage of all functional relationships among the sensed variables. The identification logic uses the quality sequential probability ratio, which provides a useful on-line measure of confidence in the various forms of analytic redundancy. Preliminary simulation results indicate good behavior of the analytic decision statistic, based on the sequential probability ratio test.

Dual-Sensor Failure Identification Using Analytic Redundancy

In this paper we present a reliable technique for failure detection and identification for dual flight control sensors aboard the F-8 digital fly-by-wire aircraft, and we discuss the successful application of the technique to identifying failures injected on test flight downlink data. The technique exploits the analytic redundancy which exists as relationships among variables being measured by dissimilar instruments, and it accommodates both modeling errors and the allowable errors on unfailed instruments. With straightforward modification the technique may be extended to provide failure monitoring of a single remaining sensor after the identified failure of its companion sensor. Nomenclature = SPRT failure threshold ( 0) = DG case orientation angle

Resource allocation and performance evaluation in large human-machine organizations

A methodology is presented for mapping the processes comprising a mission onto the capable resources within the organization such that the completion time of the terminal process in the mission is minimized. The authors then build a Petri net simulation directly from the output of the mapping algorithm to perform sensitivity analyses on the solution. This capability enables the analyst to study in an interactive way variations in the performance of the organization as a function of its workload capacity, the expertise distribution of its members, the task requirements, and the communication network linking the different resources in the organization. >

A design methodology for robust failure detection and isolation

A decentralized failure detection and isolation (FDI) methodology, which is robust with respect to model uncertainties and noise, is presented Redundancy metrics are developed, and optimization problems are posed for the choices of robust parity relations. Closed-form solutions for some special failure cases are given. Connections are drawn with other disciplines, and the use of the metrics to evaluate alternative FDI schemes is discussed.

Application of likelihood ratio methods to failure detection and identification in the NASA F-8 DFBW aircraft

A system for on-line detection and identification of aircraft sensor and effector failures is developed. The heart of the system is a state estimator which provides accurate, real time estimates of the aircraft states. These estimates are used both to provide failure analysis and as inputs to the flight control system. Because the sensors measure functions of the aircraft state, the state estimator also provides running estimates of what it believes each sensor output ought to be, based upon the previous history of sensor outputs and commanded control inputs. Because of the relatively large number of sensors, of various types, that are available; there is an abundance of observability. Thus, failure of a single sensor will not greatly degrade the state estimates and in the event of a sensor failure the output of the failed sensor will diverge from the estimated value. The divergence is monitored and decision logic, based upon likelihood ratio tests, is employed for sensor failure detection and identification (FDI). The likelihood ratio methods provide a systematic, quantitative means for design of the decision logic. This FDI technique is in direct contrast to techniques which employ voting among like sensors and therefore require three sensors of every type in order to identify the single failure of any one of them. The present technique, employing likelihood ratio methods, requires fewer sensors because of its utilization of the redundant information available from sensors of different types which are coupled through the dynamics of the aircraft.

RCS Jet Failure Identification for the Space Shuttle

Abstract A software technique for onboard detection and identification of hard failures and leaks of the shuttle orbiter reaction control subsystem jets during the orbital flight phase is presented. The method uses only the gimbal angle and linear velocity measurements available from the orbiter inertial measurement unit. Uncoupled steady-state constant covariance extended Kalman filters with residual traps are employed for rotational and translational state estimation, and generalized likelihood techniques are used for failure identification. Rigid body simulations indicate station-level identification times of 1.2 seconds for primary jet hard failures and less than 70 seconds for primary jet leaks.

A Signal Validation Methodology for Nuclear Power Plants

A signal validation methodology to improve the reliability of the information displayed to the operator of a nuclear power plant is presented. The general design methodology is developed and then applied to the steam generator and feedwater subsystem of a typical pressurized water reactor. Using steady-state and transient simulations of this subsystem, including realistic additive sensor noise, the developed algorithm successfully isolates a variety of injected sensor faults and demonstrates its inherent capability to isolate common-mode sensor failures and plant component failures.

Human cognitive performance in antisubmarine warfare: situation assessment and data fusion

A cognitive simulation model of the antisubmarine warfare commander (ASWC) in a Naval Battle Group was developed, with particular emphasis on modeling his data fusion activities during situation assessment. These activities involve the estimation of enemy submarine tracks (position and velocity) and their likelihoods in a passive sonar convergence zone environment. The simulation is normative-descriptive in nature. A normative (optimal) mathematical model was first developed based on modern estimation and control theory. Experimental results with 20 experienced naval subjects on four scenarios, involving 12 contacts and three submarines each, indicated that the optimal or normative model always demonstrated superior performance to that of the human subjects. Known cognitive limitations and biases, drawn from the cognitive and behavioral sciences and/or directly observed in human subject performance, were then incorporated in the model. >

Resource allocation in large man-machine organizations

The authors present an algorithm to map processes onto organizations. The organization consists of resources and communication links that must be used to complete all the processes. The objective is to minimize the completion time of the terminal process without violating the constraints which are defined in terms of the requirements of the processes and the capabilities of the resources. The organization performing the processes, with the process mapping specified by the algorithm, is then simulated using Petri nets. The simulation model can be generated automatically from the definition of the organizational structure and the solution of the process mapping problem. Then simulations can be performed to examine the sensitivity of the overall performance to changes in the parameters of the mission and of the organization.<<ETX>>

A normative-descriptive study of team detection with communication alternatives

A model of human decision-making in distributed detection problems is developed and described. The model, called TOSCA, involves a team of two decision-makers (DMs), a primary DM and a consultant, whose goal is to solve a sequential binary hypothesis-testing problem. The model is based on the work of J.D. Papastavrou and M. Athans (1986), which modeled the primary DM and the consultant as members of a team with a common goal. TOSCA extends their work by including other features involved in distributed detection and team decision-making. TOSCA extends the model of Papastavrou and Athans by not only developing a normative mathematical formulation with these features, but also designing and running an experimental paradigm with human subjects to generate descriptive data. After analyzing how humans deviated from predicted normative behavior, the parameters in the normative model were adjusted to reflect human statistical misperception, resulting in a normative-descriptive model.<<ETX>>