Johann Schumann and Pramod Gupta NASA Ames Research Center Bayesian Verification & Validation tools.

Slides:



Advertisements
Similar presentations
Chapter 12 Prototyping and Testing Design of Biomedical Devices and Systems By Paul H. King Richard C. Fries.
Advertisements

Model Checker In-The-Loop Flavio Lerda, Edmund M. Clarke Computer Science Department Jim Kapinski, Bruce H. Krogh Electrical & Computer Engineering MURI.
L OUISIANA T ECH U NIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK.
RISK INFORMED APPROACHES FOR PLANT LIFE MANAGEMENT: REGULATORY AND INDUSTRY PERSPECTIVES Björn Wahlström.
Robust and Efficient Control of an Induction Machine for an Electric Vehicle Arbin Ebrahim and Dr. Gregory Murphy University of Alabama.
Modeling Human Reasoning About Meta-Information Presented By: Scott Langevin Jingsong Wang.
Nonlinear rational model identification and control Professor Quan M. Zhu Bristol Institute of Technology University of the West of England Frenchay Campus.
Presented by: Hatem Halaoui
Project Work and Administration
1 Validation and Verification of Simulation Models.
University of Southern California Center for Software Engineering CSE USC ©USC-CSE 3/11/2002 Empirical Methods for Benchmarking High Dependability The.
CH 1 Introduction Prof. Ming-Shaung Ju Dept. of Mechanical Engineering NCKU.
Monté Carlo Simulation MGS 3100 – Chapter 9. Simulation Defined A computer-based model used to run experiments on a real system.  Typically done on a.
 Catalogue No: BS-338  Credit Hours: 3  Text Book: Advanced Engineering Mathematics by E.Kreyszig  Reference Books  Probability and Statistics by.
Unit 3a Industrial Control Systems
1 Prediction of Software Reliability Using Neural Network and Fuzzy Logic Professor David Rine Seminar Notes.
1 September 4, 2003 Bayesian System Identification and Structural Reliability Soheil Saadat, Research Associate Mohammad N. Noori, Professor & Head Department.
1 841f06parnas13 Evaluation of Safety Critical Software David L. Parnas, C ACM, June 1990.
©Ian Sommerville 1995 Software Engineering, 5th edition. Chapter 22Slide 1 Verification and Validation u Assuring that a software system meets a user's.
1. An Overview of the Data Analysis and Probability Standard for School Mathematics? 2.
Research Heaven, West Virginia Lyapunov Stability Analysis and On-Line Monitoring Bojan Cukic, Edgar Fuller, Srikanth Gururajan, Martin Mladenovski, Sampath.
ISE 352: Design of Experiments
CSCE 548 Secure Software Development Risk-Based Security Testing.
Research Heaven, West Virginia Verification and Validation of Adaptive Systems Online Failure Detection and Identification for IFCS through Statistical.
Assurance techniques for code generators Ewen Denney USRA/RIACS, NASA Ames Bernd Fischer ECS, U Southampton.
Independent Verification and Validation (IV&V) Techniques for Object Oriented Software Systems SAS meeting July 2003.
COMP3503 Intro to Inductive Modeling
QBM117 Business Statistics Estimating the population mean , when the population variance  2, is known.
1 Adaptive, Optimal and Reconfigurable Nonlinear Control Design for Futuristic Flight Vehicles Radhakant Padhi Assistant Professor Dept. of Aerospace Engineering.
© 2008 McGraw-Hill Higher Education The Statistical Imagination Chapter 10. Hypothesis Testing II: Single-Sample Hypothesis Tests: Establishing the Representativeness.
Slide 1V&V 10/2002 Software Quality Assurance Dr. Linda H. Rosenberg Assistant Director For Information Sciences Goddard Space Flight Center, NASA
West Virginia University Towards Practical Software Reliability Assessment for IV&V Projects B. Cukic, E. Gunel, H. Singh, V. Cortellessa Department of.
Verification and Validation in the Context of Domain-Specific Modelling Janne Merilinna.
V&V of COTS RTOS for Space Flight Projects The 1st Annual NASA Office of Safety and Mission Assurance (OSMA) Software Assurance Symposium (SAS) Michael.
Statistics and the Verification Validation & Testing of Adaptive Systems Roman D. Fresnedo M&CT, Phantom Works The Boeing Company.
TEKS (6.10) Probability and statistics. The student uses statistical representations to analyze data. The student is expected to: (B) identify mean (using.
Research Heaven, West Virginia Verification and Validation of Adaptive Systems Bojan Cukic, Eddie Fuller, Marcello Napolitano, Harshinder Singh, Tim Menzies,
Optimizing NASA IV&V Benefits Using Simulation Grant Number: NAG David M. Raffo, Ph.D College of Engineering and Computer Science School of Business.
Development of Methodologies for Independent Verification and Validation of Neural Networks NAG OSMA-F001-UNCLASS Methods and Procedures.
Safety Critical Systems 5 Testing T Safety Critical Systems.
September Bound Computation for Adaptive Systems V&V Giampiero Campa September 2008 West Virginia University.
Safety-Critical Systems 5 Testing and V&V T
M Global Software Group 1 Motorola Internal Use Only Better Software Quality at a Lower Cost: Testing to Eliminate Software Black Holes Isaac (Haim) Levendel,
Digital Intuition Cluster, Smart Geometry 2013, Stylianos Dritsas, Mirco Becker, David Kosdruy, Juan Subercaseaux Welcome Notes Overview 1. Perspective.
1 Quality Attributes of Requirements Documents Lecture # 25.
Intelligent Systems Software Assurance Symposium 2004 Bojan Cukic & Yan Liu, Robyn Lutz & Stacy Nelson, Chris Rouff, Johann Schumann, Margaret Smith July.
"Classical" Inference. Two simple inference scenarios Question 1: Are we in world A or world B?
February 2007 NASA Dryden Status Aerospace Control & Guidance Sub-committee Boulder, CO February 2007 John Bosworth (661)
Software Engineering1  Verification: The software should conform to its specification  Validation: The software should do what the user really requires.
ARO Workshop Wendy Roll - May 2004 Topic 4: Effects of software certification on the current balance between software/system modeling, analysis and testing.
Independent Verification and Validation. IV&V Independent group –Outside of the development shop –Responsible for verifying and validating a system –May.
Technological Educational Institute Of Crete Department Of Applied Informatics and Multimedia Neural Networks Laboratory 1 Introduction To Neural Networks.
Scalability and Development of Space Networks Vincenzo Liberatore, Ph.D. Disclaimer: the views expressed here are solely the author’s, not the presenter’s.
Csci 418/618 Simulation Models Dr. Ken Nygard, IACC 262B
1 SMART-T Briefing to OSMA SAS - July 19, 2004 SMART-T Project Overview Kurt D. Guenther AS&M / Dryden Flight Research Center July 19, 2004.
SwCDR (Peer) Review 1 UCB MAVEN Particles and Fields Flight Software Critical Design Review Peter R. Harvey.
ARO Workshop Wendy Roll - May 2004 Topic 4: Effects of software certification on the current balance between software/system modeling, analysis and testing.
Data Mining Practical Machine Learning Tools and Techniques By I. H. Witten, E. Frank and M. A. Hall Chapter 5: Credibility: Evaluating What’s Been Learned.
Statistical Inference for the Mean Objectives: (Chapter 8&9, DeCoursey) -To understand the terms variance and standard error of a sample mean, Null Hypothesis,
1 Software Testing and Quality Assurance Lecture 38 – Software Quality Assurance.
THEORY OF SAMPLING MMEA Certainty Seminar Markku Ohenoja 1 Markku Ohenoja / Control Engineering Laboratory
Analyze Wrap Up and Action Items
CSCE 548 Secure Software Development Risk-Based Security Testing
Saint- Petersburg State University of Aerospace Instrumentation
Intelligent Systems Software Assurance Symposium 2004
Measuring Service in Multi-Class Networks
Predictive Modeling for Patient Recruitment in Multicenter Trials
Neural Networks and Their Application in the Fields of Coporate Finance By Eric Séverin Hanna Viinikainen.
Biological Science Applications in Agriculture
Using Clustering to Make Prediction Intervals For Neural Networks
Presentation transcript:

Johann Schumann and Pramod Gupta NASA Ames Research Center Bayesian Verification & Validation tools for adaptive systems

Motivation for NN V&V Basis for Case Study I: Neuro-adaptive control (IFCS Gen-II) Network “learns” to compensate for deviations between plant and model Previous work: SW V&V process for NN-based control “Confidence tool” for dynamic monitoring The major obstacle to the deployment of adaptive and autonomous systems is being able to verify their correct operation – In Flight Fixed gain controllers cannot deal with catastrophic changes or degradation in plant Adaptive systems (e.g., NN) can react to unexpected situations through learning Relevance and potential: IFCS NN controlled aircraft (F-15, C-17) UAV control Space exploration Any safety-critical application of NN control

V&V Issues & our Approach Our approach combines mathematical analysis, intelligent validation, and dynamic monitoring and supports specific software V&V process, targets multiple aspects and phases of V&V of adaptive control systems, and uses a unique combination of research in –Neural Networks –Control Theory –Numerical Methods –Bayesian Statistics Verification: how to specify an unforseen event? Validation: not possible to test all configurations While traditional V&V methods will remain useful, these methods alone are insufficient to verify and certify adaptive control systems for use in safety-critical applications

Our Bayesian Approach How good is the network performing at the moment? Traditional: NN as a Black Box Here: Look at probability distribution of the NN output Variance (confidence measure) depends on: –How well is the network trained? –How close are we to “well-known” areas Large variance = bad estimate; no reliable result, just a guess Small variance = good estimate Our approach, based on a Bayesian approach, provides a measure of how well the neural network is performing at the moment

Milestone I: Envelope Tool Basis: Adaptive NN-based controller Lyapunov error bound defines regions of eventual stability Regions where confidence is small might cause instability Informally: a safe envelope is a region where the confidence level is sufficiently high Bayesian approach combined with sensitivity analysis Challenge: methods for efficient determination of safe envelope Can help answer questions like How large is the current safe envelope? How far is the operational point from the edge? Current status: mathematical background formulated, prototypical Matlab/Simulink implementation designed, first simulation experiments

Confidence Envelope Confidence Surface Safety Envelope: area of good confidence airspeed The Envelope tool uses a Bayesian Approach to calculate the current safety envelope 1/confidence good bad altitude

Conclusions & next steps Current work as scheduled toward deliverable (9/2004) prototypical implementation in Matlab/Simulink report on mathematical background and tool Getting Case Study I ready: IFCS Gen-II simulink model Next steps in research: system identification (sysID): estimate confidence of parameters other model representations (e.g., parameter tables with polynomial interpretation) Preparation of Case Study II and III