Sensor Validation using Bayesian Networks

One of NASA’s key mission requirements is robust state estimation. Sensing, using a wide range of sensors and sensor fusion approaches, plays a central role in robust state estimation, and there is a need to diagnose sensor failure as well as component failure. Sensor validation techniques address this problem: given a vector of sensor readings, decide whether sensors have failed, therefore producing bad data. We take in this paper a probabilistic approach, using Bayesian networks, to diagnosis and sensor validation, and investigate several relevant but slightly different Bayesian network queries. We emphasize that onboard inference can be performed on a compiled model, giving fast and predictable execution times. Our results are illustrated using an electrical power system, and we show that a Bayesian network with over 400 nodes can be compiled into an arithmetic circuit that can correctly answer queries in less than 500 microseconds on average.

[1]  Du Zhang,et al.  Diagnosis of component failures in the Space Shuttle main engines using Bayesian belief networks: a feasibility study , 2002, 14th IEEE International Conference on Tools with Artificial Intelligence, 2002. (ICTAI 2002). Proceedings..

[2]  Adnan Darwiche,et al.  Model-Based Diagnosis under Real-World Constraints , 2000, AI Mag..

[3]  Adnan Darwiche,et al.  Compiling Bayesian Networks Using Variable Elimination , 2007, IJCAI.

[4]  R. L. Bickford,et al.  REAL-TIME SENSOR VALIDATION FOR AUTONOMOUS FLIGHT CONTROL , 1997 .

[5]  David J. Spiegelhalter,et al.  Local computations with probabilities on graphical structures and their application to expert systems , 1990 .

[6]  James M. Rehg,et al.  Vision-based speaker detection using Bayesian networks , 1999, Proceedings. 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No PR00149).

[7]  William A. Maul,et al.  Sensor Data Qualification for Autonomous Operation of Space Systems , 2006, AAAI Fall Symposium: Spacecraft Autonomy.

[8]  Ole J. Mengshoel,et al.  Advanced Diagnostics and Prognostics Testbed , 2007 .

[9]  Ole J. Mengshoel.  Macroscopic Models of Clique Tree Growth for Bayesian Networks , 2007, AAAI.

[10]  A. Darwiche,et al.  Complexity Results and Approximation Strategies for MAP Explanations , 2011, J. Artif. Intell. Res..

[11]  Timothy W. Bickmore.  A probabilistic approach to sensor data validation , 1992 .

[12]  Judea Pearl,et al.  Probabilistic reasoning in intelligent systems - networks of plausible inference , 1991, Morgan Kaufmann series in representation and reasoning.

[13]  Fernando Figueroa,et al.  Rocket Testing and Integrated System Health Management , 2006 .

[14]  Xavier Boyen,et al.  Exploiting the Architecture of Dynamic Systems , 1999, AAAI/IAAI.

[15]  Henrik Schiøler,et al.  Sensor Fusion Using Dynamic Bayesian Networks in Livestock Production Buildings , 2006, 2006 International Conference on Computational Inteligence for Modelling Control and Automation and International Conference on Intelligent Agents Web Technologies and International Commerce (CIMCA'06).

[16]  Jeffrey S. Lavell,et al.  Report on the Loss of the Mars Polar Lander and Deep Space 2 Missions , 2000 .

[17]  David C. Wilkins,et al.  Controlled generation of hard and easy Bayesian networks: Impact on maximal clique size in tree clustering , 2006, Artif. Intell..

[18]  Ann E. Nicholson,et al.  Dynamic Belief Networks for Discrete Monitoring , 1994, IEEE Trans. Syst. Man Cybern. Syst..

[19]  Ole J. Mengshoel,et al.  Designing Resource-Bounded Reasoners using Bayesian Networks: System Health Monitoring and Diagnosis , 2007 .

[20]  Gautam Biswas,et al.  Bayesian Fault Detection and Diagnosis in Dynamic Systems , 2000, AAAI/IAAI.

[21]  Alan S. Willsky,et al.  A survey of design methods for failure detection in dynamic systems , 1976, Autom..

[22]  Ron J. Patton,et al.  Fault detection and diagnosis in aerospace systems using analytical redundancy , 1991 .

[23]  S. Ferrari,et al.  Demining sensor modeling and feature-level fusion by Bayesian networks , 2006, IEEE Sensors Journal.