Karen Gundy-Burlet

Combining unit-level symbolic execution and system-level concrete execution for testing nasa software

We describe an approach to testing complex safety critical software that combines unit-level symbolic execution and system-level concrete execution for generating test cases that satisfy user-specified testing criteria. We have developed Symbolic Java PathFinder, a symbolic execution framework that implements a non-standard bytecode interpreter on top of the Java PathFinder model checking tool. The framework propagates the symbolic information via attributes associated with the program data. Furthermore, we use two techniques that leverage system-level concrete program executions to gather information about a units input to improve the precision of the unit-level test case generation. We applied our approach to testing a prototype NASA flight software component. Our analysis helped discover a serious bug that resulted in design changes to the software. Although we give our presentation in the context of a NASA project, we believe that our work is relevant for other critical systems that require thorough testing.

An Adaptive Critic Approach to Reference Model Adaptation

Neural networks have been successfully used for implementing control architectures for different applications. In this work, we examine a neural network augmented adaptive critic as a Level 2 intelligent controller for a C- 17 aircraft. This intelligent control architecture utilizes an adaptive critic to tune the parameters of a reference model, which is then used to define the angular rate command for a Level 1 intelligent controller. The present architecture is implemented on a high-fidelity non-linear model of a C-17 aircraft. The goal of this research is to improve the performance of the C-17 under degraded conditions such as control failures and battle damage. Pilot ratings using a motion based simulation facility are included in this paper. The benefits of using an adaptive critic are documented using time response comparisons for severe damage situations.

Control Reallocation Strategies for Damage Adaptation in Transport Class Aircraft

This paper examines the feasibility, potential benefits and implementation issues associated with retrofitting a neural-adaptive flight control system (NFCS) to existing transport aircraft, including both cable/hydraulic and fly-by-wire configurations. NFCS uses a neural network based direct adaptive control approach for applying alternate sources of control authority in the presence of damage or failures in order to achieve desired flight control performance. Neural networks are used to provide consistent handling qualities across flight conditions, adapt to changes in aircraft dynamics and to make the controller easy to apply when implemented on different aircraft. Full-motion piloted simulation studies were performed on two different transport models: the Boeing 747-400 and the Boeing C-17. Subjects included NASA, Air Force and commercial airline pilots. Results demonstrate the potential for improving handing qualities and significantly increased survivability rates under various simulated failure conditions.

Automatically finding the control variables for complex system behavior

Testing large-scale systems is expensive in terms of both time and money. Running simulations early in the process is a proven method of finding the design faults likely to lead to critical system failures, but determining the exact cause of those errors is still time-consuming and requires access to a limited number of domain experts. It is desirable to find an automated method that explores the large number of combinations and is able to isolate likely fault points.Treatment learning is a subset of minimal contrast-set learning that, rather than classifying data into distinct categories, focuses on finding the unique factors that lead to a particular classification. That is, they find the smallest change to the data that causes the largest change in the class distribution. These treatments, when imposed, are able to identify the factors most likely to cause a mission-critical failure. The goal of this research is to comparatively assess treatment learning against state-of-the-art numerical optimization techniques. To achieve this, this paper benchmarks the TAR3 and TAR4.1 treatment learners against optimization techniques across three complex systems, including two projects from the Robust Software Engineering (RSE) group within the National Aeronautics and Space Administration (NASA) Ames Research Center. The results clearly show that treatment learning is both faster and more accurate than traditional optimization methods.

TWO-DIMENSIONAL COMPUTATIONS OF MULTI-STAGE COMPRESSOR FLOWS USING A ZONAL APPROACH

A clear understanding of the fluid dynamics associated with rotor/stator configurations can be very helpful when optimizing the performance of turbomachinery. In this study, a two-dimensional, implicit, thin-layer, Navier-Stokes zonal approach has been used to investigate the flow within a 2 1/2-stage compressor. Relative motion between the rotor and stator airfoils is made possible with the use of systems of patched and overlaid grids that move with respect to each other. The treatment of multistage turbomachines with arbitrary numbers of airfoils per row is made possible by the use of a flexible database system. Results in the form of instantaneous pressure and entropy contours and time-averaged pressures are presented for the 2 1/2-stage compressor. Time-averaged pressures and pressure amplitudes for a single-stage turbine configuration are also presented. The numerical results compare well with experimental data.

A SURVEY OF HOT STREAK EXPERIMENTS AND SIMULATIONS

Experimental and computational data have shown that the flow exiting gas-turbine combustors can contain large circumferential and radial temperature non-uniformities. The temperature non-uniformities, or hot streaks, can have a significant impact on the performance and durability of first-stage turbine airfoils. This paper contains a survey of the hot streak experiments and simulations that have been performed during the last two decades, and the impact they have had on the design of high-pressure turbine stages.

INVESTIGATION OF AIRFOIL CLOCKING AND INTER-BLADE-ROW GAPS IN AXIAL COMPRESSORS

Axial compressors have inherently unsteady flow fields because of relative motion between rotor and stator airfoils. This relative motion leads to viscous and inviscid (potential) interactions between blade rows. As the number of stages increases in a turbomachine, the buildup of convected wakes can lead to progressively more complex wake/wake and wake/airfoil interactions. Variations in the relative axial circumferential positions of stators or rotors can change these interactions, leading to different unsteady forcing functions on airfoils and different compressor efficiencies. In addition, the axial gaps between adjacent blade rows affect the unsteady forcing functions by modulating the potential interaction of the rows. The current study uses an unsteady, two-dimensional thin-layer Navier-Stokes procedure to investigate the combined effects of stator clocking and varying axial gaps in a low-speed axial compressor design. Relative motion between rotors and stators is made possible by the use of systems of patched and overlaid grids. Results include surface pressures, instantaneous forces and efficiencies for the compressor.

Software V&V support by parametric analysis of large software simulation systems

Modern aerospace software systems simulations usually contain many (dependent and independent) parameters. Due to the large parameter space, and the complex, highly coupled nonlinear nature of the different system components, analysis is complicated and time consuming. Thus, such systems are generally validated only in regions local to anticipated operating points rather than through characterization of the entire feasible operational envelope of the system. We have addressed the factors deterring such a comprehensive analysis with a tool to support parametric analysis and envelope assessment: a combination of advanced Monte Carlo generation with n-factor combinatorial parameter variations and model-based testcase generation is used to limit the number of cases without sacrificing important interactions in the parameter space. For the automatic analysis of the generated data we use unsupervised Bayesian clustering techniques (AutoBayes) and supervised learning of critical parameter ranges using the treatment learner TAR3. This unique combination of advanced machine learning technology enables a fast and powerful multivariate analysis that supports finding of root causes.

Physics of Airfoil Clocking in Axial Compressors

Axial compressors have inherently unsteady flow fields because of relative motion between rotor and statnr airfnils. This relative motion leads to viscous and inviscid (potential) interactions between blade rows. As the number of stages increases in a turbomachine, the buildup of convected wakes can lead in progressively more complex wake/wake and wake/airfnil interactions. Variations in the relative circumferential positions of stators or rotors can change these interactions, leading to different unsteady forcing functions on airfoils and different compressor efficiencies. The current study uses an unsteady, two-dimensional thin-layer Navier-Stokes zonal approach to investigate the unsteady aerodynamics of stator clocking in a low-speed 2 ½-stage compressor. Relative motion between rotors and stators is made possible by the use of systems of patched and overlaid grids. Results include surface pressures instantaneous forces and efficiencies for a 2 ½-stage compressor configuration.Copyright

Parametric Analysis of a Hover Test Vehicle using Advanced Test Generation and Data Analysis

Large complex aerospace systems are generally validated in regions local to anticipated operating points rather than through characterization of the entire feasible operational envelope of the system. This is due to the large parameter space, and complex, highly coupled nonlinear nature of the different systems that contribute to the performance of the aerospace system. We have addressed the factors deterring such an analysis by applying a combination of technologies to the area of flight envelop assessment. We utilize n-factor (2,3) combinatorial parameter variations to limit the number of cases, but still explore important interactions in the parameter space in a systematic fashion. The data generated is automatically analyzed through a combination of unsupervised learning using a Bayesian multivariate clustering technique (AutoBayes) and supervised learning of critical parameter ranges using the machine-learning tool TAR3, a treatment learner. Covariance analysis with scatter plots and likelihood contours are used to visualize correlations between simulation parameters and simulation results, a task that requires tool support, especially for large and complex models. We present results of simulation experiments for a cold-gas-powered hover test vehicle.