Michael Balchanos

Strategies for Integrating Models of Interdependent Subsystems of Complex System-Of-Systems Products

The Office of Naval Research has established a need for improved design and analysis methods for the next generation of naval surface combatants. The Aerospace Systems Design Laboratory (ASDL) has initiated the integrated reconfigurable intelligent systems project to address design issues associated with the future systems. A goal of this program is to define preliminary approaches for developing an integrated modeling and simulation environment for complex systems. Since such systems are heterogeneous, dynamical and interdependent we suggest that a system-of-systems multidisciplinary approach is most appropriate for investigating and executing solutions. An integration methodology employing innovative techniques and a framework of tools that can be used to couple disparate models and simulations is presented. Methods for validating the final product to justify the selected approach and demonstrate a proof of concept for the integrated model are also discussed

A framework for the quantitative assessment of performance-based system resilience

Increasing system complexity and threat uncertainty require the consideration of resilience in the design and analysis of engineered systems. While the resilience engineering community has begun to converge on a definition and set of characteristics for resilience, methods for quantifying the concept are still limited in their applicability to system designers. This paper proposes a framework for assessing resilience that focuses on the ability of a system to absorb disruptions, recover from them, and adapt over time. The framework extends current approaches by explicitly considering temporal aspects of system responses to disruptions, volatility in system performance data, and the possibility of multiple disruption events. Notional system performance data is generated using the logistic function, providing an experimental platform for a parametric comparison of the proposed resilience metric with an integration-based metric. An information exchange network model is used to demonstrate the applicability of the framework towards system design tradeoff studies using stochastic simulations. The presented framework is domain-agnostic and flexible, such that it can be applied to a variety of systems and adjusted to focus on specific aspects of resilience.

Towards a method for assessing resilience of complex dynamical systems

System survivability is one of the key requirements for the conceptual design of an Integrated Reconfigurable Intelligent (IRIS) system. Current approaches in survivability engineering may not effectively address the challenges in designing revolutionary, large scale complex and multi-capable systems. The main objective of this study is to investigate the concept of resilience in the context of system safety and survivability and suggest a technique for assessing resilience in systems engineering. Resilience is expected to be the enabler for integrating safety and survivability in the early conceptual design. For this purpose, a small scale cooling network system architecture has been utilized to demonstrate the technique, both for a 32-valve baseline, as well as for six other configurations. The application of the technique allowed for the comparative assessment and tradeoff investigation of resilience function capacities, as well for the identification of solution feasibility, under adaptability and robustness constraints.

Towards a Data Calibrated, Simulation-Based Campus Energy Analysis Environment for Situational Awareness and Future Energy System Planning

Researchers at Georgia Tech (GT) have recently begun the GT Smart Energy Campus initiative, which combines campus energy metering data with physics-based modeling and simulation to create an integrated analysis environment for campus energy. The environment consists of a digital representation of campus, which supports situational awareness, as well as a virtual test bed for analyzing emerging energy technologies and future scenarios. The first year of the initiative has focused on evaluating campus energy metering data using visual analytics and statistical analysis techniques. Data analysis is presented as having value for two main uses: (1) as attention-directing information to help system operators diagnose anomalies and (2) as a precursor to modeling and simulation (M&S) in future phases of the Smart Energy Campus initiative. The environment is explained using the initial study scoping of the campus thermal energy generation and distribution systems. Furthermore, a modeling and simulation approach leveraging the Modelica M&S language is described, and preliminary results in using it to represent the campus chilled water system are presented.Copyright

Modeling and simulation of integrated intelligent systems

Complex systems consist of a large number of entities with their independent local rules and goals, along with their interactions. The effect of these properties tends to produce complex behaviors that are required to be understood in order to analyze and design the systems. However, these behaviors are difficult to be predicted a priori, and can only be studied through simulation. The study presented in this paper proposes a process for developing an integrated dynamic modeling and simulation environment designed for understanding the behavior of the next generation naval ship which is envisioned to be self-sensing, self-assessing and self-reacting. Various models, including power model, fluid model and control model, are developed to investigate the functionalities of the naval ship systems. An object oriented approach is employed to validate the architectural design of the integrated simulation environment and a surrogate modeling technique is utilized to accelerate the simulation speed.

A Data Mining and Visual Analytics Perspective on Sustainability-Oriented Infrastructure Planning

The research presented in this paper focuses on developing a multi-scale, integrated environment that supports situational awareness, optimization, as well as forecasting and virtual experimentation at the campus level. One of the key features of this research is its ability to extend beyond the common data-driven load-forecasting exercise and integrate System-of-Systems (SoS) level predictive capabilities to enable the aforementioned functionalities. FORESIGHT, an interactive campus data browser designed to handle any visual analytics tasks on real-time data streams is presented.

Metrics-based analysis and evaluation framework for engineering resilient systems

The DoDs ERS initiative calls for affordable, effective, and adaptable systems development. In support of this, a metrics-based analysis framework is introduced to address certain challenges for the design of future C2 military System-of-Systems (SoS). The interpretation of the concept of resilience, as well as a supporting threat analysis procedure for military SoS applications, have been the key driver for the evaluation of a systems ability to maintain its mission capability and health, when under attack due to given threats. An agent-based C2 UAV communication network application has been developed for the demonstration of the framework. Scenario-based case studies that involved communication jamming by the adversary forces are introduced for the evaluation the C2 systems response to a threat, including both degradation and recovery periods.

Mobility Analysis of Electric Autonomous Vehicle Networks Driven by Energy-Efficient Rerouting

Increasing demand for urban mobility has contributed to heavy traffic in metro areas, rising transportation costs and travel times. The concept of driverless centrally-controlled electric vehicles has been attracting attention for its potential of minimizing traffic congestion. This study is introducing the tools and methods needed for evaluating the benefits of the concept. An intelligent route planning algorithm, along with a traffic simulation environment, which is coupled to a vehicle performance model have been developed. An interactive graphical tool has been created to showcase the resulting benefits of the concept and to support the conceptual design of energy efficient electric vehicles for centralized large scale operations. The study has identified correlations between a driving environments traffic density and the potential of relieving the effects of congestion, with considerable travel time, energy and vehicle wear savings attainable.

Incipient failure detection: A particle filtering approach with application to actuator systems

The background, simulation and experimental evaluation of an anomaly detector for Brushless DC motor winding faults is described in this paper in the context of an aircraft Electro-Mechanical Actuator (EMA) application. Results acquired from an internal Failure Modes and Effects Analysis (FMEA) study identified turn-to-turn winding faults as the primary mechanism, or mode, of failure. Physics-of-failure mechanisms used to develop a model for the identified fault are provided. Then, an experimental test procedure was devised and executed to validate the model. Additionally, a diagnostic feature, identified by the fault model and derived using Hilbert transform theory, was validated using the system model and experimental data for several fault dimensions. Next, a feature extraction routine preprocesses monitoring parameters and passes the resulting features to a particle filter. The particle filter, based on Bayesian estimation theory, allows for representation and management of uncertainty in a computationally efficient manner. The resulting anomaly detection routine declares a fault only when a specified confidence level is reached at a given false alarm rate. Finally, the real-time performance of the anomaly detector is evaluated using LabVIEW.

Modeling and Simulation-based Analysis for Large Scale Campus Chilled Water Networks

As part of the GT Smart Energy Campus research initiative, a chilled water district cooling network model has been created in Modelica and analyzed via simulations for different system configuration settings, both from chiller plant and campus building side, and under different heat loading conditions. The model allows for exploration of the design space formed by different chiller staging options and operational settings. A parametric study was performed to determine the most energy efficient chiller plant staging and chiller setpoint temperatures for a range of different weather test point conditions. The parametric study yields a map that, if calibrated, can be used by operators to choose the best chiller settings for weather conditions that might be encountered. The sensitivity of best settings to building-side ΔT fluctuations are also observed. The basic model that has been established can in the future be calibrated to actual metering data to provide insight relevant to an operating campus chilled water system. Ultimately, such a modeling baseline can be a starting point for studies on how new technology additions can improve chilled water system performance, as well as foundation for more advanced energy system virtual test-bed capabilities.