PROJECT A
Title
Quantifying Uncertainty in the Battlespace Environment, NRL/UCAR, Award: $1,139,000 Duration: 06/01/14-05/31/17, Role: UConn PI.
Team
Naval Research Laboratory, Marine Meteorology Division, Naval Research Laboratory, Acoustics Division; Naval Research Laboratory, Marine Geosciences Division; Army Research Laboratory; Army Engineer Research and Development Center; University of Southern California, University of Connecticut (Krishna R. Pattipati & Omer Khan)
Project Summary
The goals of this project are to develop capabilities that are applicable across a broad range of DoD missions areas that are impacted by the environment either directly (e.g., ship routing) or indirectly (e.g., sensor performance-based asset allocation). At project completion, we intend to have:
Methods and frameworks to efficiently (as measured by scaling) combine and propagate heterogeneous sources of uncertainty through multiple linked models from data to decisions;
Methods of supervision and work load balancing, and concurrency controls for Dynamic Programming and Branch-and-Bound formulations with very large search spaces motivated by the need for COA guidance to decision makers in dynamic and uncertain mission environments
Methods and frameworks to efficiently and accurately produce optimal COAs for classes of missions that rely on linked model outputs utilizing strategies from dynamic stochastic optimization and game theory under Dempster-Shafer (DS) uncertainties and that provide estimates of the sensitivity of results to relevant inputs.
PROJECT B
Title
CyberSEES: Type 2: Fault Detection, Diagnosis and Prognosis of HVAC Systems, National Science Foundation, $1,090,000, 09/01/13-08/31/16, Award # CCF-1331850, Role: PI.
Team
Krishna R. Pattipati, Peter B. Luh, Bing Wang, Robert Gao (UConn); George Kuchel (UCHC)
Project Summary
The goal of this multi-disciplinary project is to develop a simple, robust, generic, and scalable model-based and data-driven Fault Detection, Diagnosis and Prognosis (FDDP) process and the associated detection, inference and predictive analytics that are applicable to a variety of buildings. The research is motivated by the observation that buildings account for more than 40% of US energy consumption. Heating, Ventilation and Air Conditioning (HVAC) constitutes 57% of energy used in commercial and residential buildings, valued at $223B in 2009. About 20% of the energy consumed by HVAC is wasted due to abrupt faults (e.g., stuck dampers), performance degradations (e.g., air filter clogging), poor controls (e.g., biases in set points), and improper commissioning (e.g., poorly balanced parallel chillers). This project will develop FDDP methodologies for HVACs to improve equipment availability, lower energy and operating costs, extend equipment life, and enhance occupants’ comfort. The FDDP process will be validated and evaluated by applying it to UConn’s Tech Park Building, Duncaster, a life-care retirement community, located in Bloomfield, CT, and to the ENN Langfang Eco-city in China. The project contributes to the vision of green and sustainable buildings equipped with cyber-physical substrata consisting of HVAC modules, networked sensors providing information on spatial and temporal distribution of occupants, smart building management systems providing situation awareness and decision support to human operators, and improved tenant comfort.
PROJECT C
Title
Agile Information and Decision Support Concepts for Dynamic Planning/Re-planning in C2 of Unmanned and Undersea Systems, Office of Naval Research, Amount: $1,134,000; Duration: 1/01/2012-12/31/2015, Award # N00014-12-1-0238, FRS# 561436, Role: PI
Team
Krishna R. Pattipati and David L. Kleinman
Project Summary
Rapid mission planning, re-planning and execution is a major operational problem in highly dynamic, asymmetric, and unpredictable mission environments. The proposed research seeks to develop and test agile C2 coordination and decision support concepts for improving resource management in dynamic mission planning. Our research focus will be on the dynamic planning/re-planning processes associated with maritime operations involving unmanned and undersea systems. UConn’s research approach will consist of developing: 1) Models to quantify the value of uncertain information; 2) Formalisms for directing flow of information among decision makers based on context and task requirements; 3) An analytical framework for representing dynamic multi-level mission environments; 4) Embedded multi-objective agent-based resource management algorithms that enable vertical (multi-level) and horizontal (within level) coordination and applying them to airspace management in a heterogeneous mission area, e.g., Unmanned Aerial Vehicles (UAVs) and Manned aircraft; 5) mixed-initiative decision and information support software for collaborative mission planning, execution and monitoring; and 6) testing the decision and information support concepts via team-in-the-loop experiments at the Naval Postgraduate School. An additional area of proposed research involves identification, learning and tracking social and ad hoc networks via graph matching algorithms based on uncertain data.
PROJECT D
Title
Goali: Diagnosis and Prognosis of Automotive Chassis Systems, National Science Foundation, Amount: $349,995, Duration: 09/01/10-8/31/15, Award #NECCS-1001445, FRS#525701, Role: PI.
Team
Krishna R. Pattipati and GM R&D
Project Summary
This project has two major goals. The first goal involves the basic research problem of investigating a hybrid model-based/data-driven/knowledge-based prognostic framework, and the associated fault models, prediction and inference algorithms, to detect and isolate incipient component degradations in coupled systems. The second goal is to demonstrate our diagnostic and prognostic framework on automotive chassis systems, an application deemed highly-relevant and salient to GM. The specific objectives of this project are to explore: (i) a hybrid model-based, data-driven and knowledge-based prognostic framework applicable to coupled systems, (ii) statistical hypothesis testing and machine learning-based classification techniques for anomaly detection (test outcomes), (iii) prediction and inference algorithms based on fault-test dependency models to isolate incipient component degradations, and (iv) demonstration of the models, algorithms and software on test fleet data. Our focus will be on chassis health determination to replaceable components. GM R&D center will provide the list of subsystems of interest, their dynamics, failure (degradation) modes and their dynamics (if known), and on-board and off-board information accessible via the CAN bus, the OnStar system and dealer repair logs.