Richard Kolacinski - Academia.edu (original) (raw)

Papers by Richard Kolacinski

Cornell University - arXiv, Jan 10, 2020

This paper summarizes the operational challenges imposed by integration of renewable electricity ... more This paper summarizes the operational challenges imposed by integration of renewable electricity generation units in transmission level where the most common renewable generation units are solar and wind farms at the scale of 100s to 1000s MW. Such units, because of their stochastic nature, introduce new complexity and uncertainty to the grid. Throughout this paper, some results from the recent planning study of integration of 1,000 MW offshore wind farm into the U.S. Eastern Interconnection transmission system are shown.

2020 IEEE 14th International Conference on Semantic Computing (ICSC), Feb 1, 2020

Click-through data has proven to be a valuable resource for improving search-ranking quality. Sea... more Click-through data has proven to be a valuable resource for improving search-ranking quality. Search engines can easily collect click data, but biases introduced in the data can make it difficult to use the data effectively. In order to measure the effects of biases, many click models have been proposed in the literature. However, none of the models can explain the observation that users with different search intent (e.g., informational, navigational, etc.) have different click behaviors. In this paper, we study how differences in user search intent can influence click activities and determined that there exists a bias between user search intent and the relevance of the document relevance. Based on this observation, we propose a search intent bias hypothesis that can be applied to most existing click models to improve their ability to learn unbiased relevance. Experimental results demonstrate that after adopting the search intent hypothesis, click models can better interpret user clicks and substantially improve retrieval performance.

2013 IEEE Energytech, 2013

Today's power systems are undergoing an evolutionary transformation where novel sensors, organize... more Today's power systems are undergoing an evolutionary transformation where novel sensors, organized as sensor networks, are providing tremendous amounts of real-time data that is useful for improving situational awareness. It is becoming increasing evident in large-scale power systems, as in many other applications, that the associated big data is both an opportunity and a challenge. In this paper we describe an approach that merges important concepts from information theory, applied to cyber-physical systems (CPS), with a computational method for modeling large-scale cyber-physical systems, referred to as System Structure. An important aspect of this approach is the interpretation of a large-scale cyber-physical system as an information processing network, where System Structure provides an approach for quantifying interactions between nodes in the network through the computation of system measures.

AIAA Guidance, Navigation, and Control Conference, 2009

Small unmanned aerial vehicles (UAVs), equipped with navigation systems and video capability, are... more Small unmanned aerial vehicles (UAVs), equipped with navigation systems and video capability, are currently being deployed for intelligence, reconnaissance and surveillance missions. One particular mission of interest involves computing location estimates for targets detected by onboard sensors. Combining UAV state estimates with information gathered by the imaging sensors leads to bearing measurements of the target that can be used to determine the target's location. This 3-D bearings-only estimation problem is nonlinear and traditional filtering methods produce biased and uncertain estimates, occasionally leading to filter instabilities. Careful selection of the measurement locations greatly enhances filter performance, motivating the development of UAV trajectories that minimize target location estimation error and improve filter convergence. The objective of this work is to develop guidance algorithms that enable the UAV to fly trajectories that increase the amount of information provided by the measurements and improve overall estimation observability, resulting in proper target tracking and an accurate target location estimate. The performance of the target estimation is dependent upon the positions from which measurements are taken relative to the target and to previous measurements. Past research has provided methods to quantify the information content of a set of measurements using the Fisher Information Matrix (FIM). Forming objective functions based on the FIM and using numerical optimization methods produce UAV trajectories that locally maximize the information content for a given number of measurements. In this project, trajectory optimization leads to the development of UAV flight paths that provide the highest amount of information about the target, while considering sensor restrictions, vehicle dynamics and operation constraints. The UAV trajectory optimization is performed for stationary targets, dynamic targets and multiple targets, for many different scenarios of vehicle motion constraints. The resulting trajectories show spiral paths taken by the UAV, which focus on increasing the angular separation between measurements and reducing the relative range to the target, thus maximizing the information provided by each measurement and improving the performance of the estimation. First of all I would like to thank my advisors Rich Kolacinski and Emilio Frazzoli. This thesis would really not have been possible without your endless support, advice and inspiration. Thank you so much for the frequent meetings and discussions, for always inspiring and guiding me, for giving me new ideas, and for showing me new ways to think. You have really helped me grow as an engineer and I am very grateful to have had the privilege of working with you both. To Brent Appleby, George Schmidt, and Linda Fuhrman, thank you for providing me with the opportunity of being a Draper Fellow. It has been a truly great experience and I really appreciate all your advice and support, with the thesis, career development, and life in general. To my MIT professors, Jon How, Nick Roy and John Deyst, I very much appreciate your guidance and support throughout this thesis. Thank you for always taking the time to advise me and point me in new directions. To Barbara Lechner and Marie Stuppard thanks for always being there to help me out and for all your encouragement and support.

International Journal of Control, 2000

ABSTRACT We investigate the problems of asymptotic stabilization and output regulation of a biomi... more ABSTRACT We investigate the problems of asymptotic stabilization and output regulation of a biomimetic actuation system which independently modulates position and net stiffness. We show how the passivity formalism and the technique of input saturation developed by Lin (1995 and 1996) can be used to asymptotically stabilize this mechanical system over its feasible domain via bounded feedback. Two output regulators, a linear PID controller and a feedback linearizing controller, are developed. The performance of the proposed controllers is illustrated through numerical simulation.

2020 IEEE 14th International Conference on Semantic Computing (ICSC), Feb 1, 2020

Millions of online discussions are generated everyday on social media platforms. Topic modelling ... more Millions of online discussions are generated everyday on social media platforms. Topic modelling is an efficient way of better understanding large text datasets at scale. Conventional topic models have had limited success in online discussions, and to overcome their limitations, we use the discussion thread tree structure and propose a "popularity" metric to quantify the number of replies to a comment to extend the frequency of word occurrences, and the "transitivity" concept to characterize topic dependency among nodes in a nested discussion thread. We build a Conversational Structure Aware Topic Model (CSATM) based on popularity and transitivity to infer topics and their assignments to comments. Experiments on real forum datasets are used to demonstrate improved performance for topic extraction with six different measurements of coherence and impressive accuracy for topic assignments.

2020 IEEE 14th International Conference on Semantic Computing (ICSC)

Millions of online discussions are generated everyday on social media platforms. Topic modelling ... more Millions of online discussions are generated everyday on social media platforms. Topic modelling is an efficient way of better understanding large text datasets at scale. Conventional topic models have had limited success in online discussions, and to overcome their limitations, we use the discussion thread tree structure and propose a "popularity" metric to quantify the number of replies to a comment to extend the frequency of word occurrences, and the "transitivity" concept to characterize topic dependency among nodes in a nested discussion thread. We build a Conversational Structure Aware Topic Model (CSATM) based on popularity and transitivity to infer topics and their assignments to comments. Experiments on real forum datasets are used to demonstrate improved performance for topic extraction with six different measurements of coherence and impressive accuracy for topic assignments.

International Journal of Semantic Computing

With the explosive growth of online discussions published everyday on social media platforms, com... more With the explosive growth of online discussions published everyday on social media platforms, comprehension and discovery of the most popular topics have become a challenging problem. Conventional topic models have had limited success in online discussions because the corpus is extremely sparse and noisy. To overcome their limitations, we use the discussion thread tree structure and propose a “popularity” metric to quantify the number of replies to a comment to extend the frequency of word occurrences, and the “transitivity” concept to characterize topic dependency among nodes in a nested discussion thread. We build a Conversational Structure Aware Topic Model (CSATM) based on popularity and transitivity to infer topics and their assignments to comments. Experiments on real forum datasets are used to demonstrate improved performance for topic extraction with six different measurements of coherence and impressive accuracy for topic assignments.

IEEE Transactions on Industry Applications

Swarm and Evolutionary Computation

2015 Swarm/Human Blended Intelligence Workshop (SHBI), 2015

Proceedings. 1998 IEEE International Conference on Robotics and Automation (Cat. No.98CH36146), 1998

... Recent work by many researchers has demonstrated the advantages of using biomimetic methodolo... more ... Recent work by many researchers has demonstrated the advantages of using biomimetic methodologies for ... A hexapod walking machine [4] using active, variable joint stiffnesses has been ... Examination of (8) reveals that increasing defor-mation, 3, will increase the stiffness as ...

Proceedings of the 18th IFAC World Congress, 2011

This paper focuses upon the development of a methodology for data-driven construction of mesoscop... more This paper focuses upon the development of a methodology for data-driven construction of mesoscopic models of the T&D system for use in real-time monitoring and control. The system dynamics are lifted to a discrete covering space which provides an encoding of the system dynamics within symbol strings. These symbol strings are treated as Bernoulli shifts and are characterized, via the machinery of information theory and formal language theory, as probabilistic automata. As these automata are fundamentally pattern recognizers, they provide a fundamental basis for event/anomaly detection and thus a basis for critical grid monitoring functions such as security state identification.

2012 Future of Instrumentation International Workshop (FIIW) Proceedings, 2012

This paper presents an enterprise architecture that supports the development and deployment of ad... more This paper presents an enterprise architecture that supports the development and deployment of advanced control and condition monitoring algorithms in power generating plants. The architecture is based on information-theoretic concepts that are used to transform multi-modal data streams into actionable information.

Under the boundedness and observability conditions, we present a globally convergent observer for... more Under the boundedness and observability conditions, we present a globally convergent observer for a class of multi-output nonlinear systems which covers the blocktriangular observer forms studied previously in the literature. The result presented in this short note incorporates and generalizes the earlier work on the observer design for singleoutput observable systems.

Typescript. Department of Mechanical and Aerospace Engineering. Thesis (Ph. D.)--Case Western Res... more Typescript. Department of Mechanical and Aerospace Engineering. Thesis (Ph. D.)--Case Western Reserve University, 1997. Includes bibliographical references (leaves 166-169).

Cornell University - arXiv, Jan 10, 2020

This paper summarizes the operational challenges imposed by integration of renewable electricity ... more This paper summarizes the operational challenges imposed by integration of renewable electricity generation units in transmission level where the most common renewable generation units are solar and wind farms at the scale of 100s to 1000s MW. Such units, because of their stochastic nature, introduce new complexity and uncertainty to the grid. Throughout this paper, some results from the recent planning study of integration of 1,000 MW offshore wind farm into the U.S. Eastern Interconnection transmission system are shown.

2020 IEEE 14th International Conference on Semantic Computing (ICSC), Feb 1, 2020

Click-through data has proven to be a valuable resource for improving search-ranking quality. Sea... more Click-through data has proven to be a valuable resource for improving search-ranking quality. Search engines can easily collect click data, but biases introduced in the data can make it difficult to use the data effectively. In order to measure the effects of biases, many click models have been proposed in the literature. However, none of the models can explain the observation that users with different search intent (e.g., informational, navigational, etc.) have different click behaviors. In this paper, we study how differences in user search intent can influence click activities and determined that there exists a bias between user search intent and the relevance of the document relevance. Based on this observation, we propose a search intent bias hypothesis that can be applied to most existing click models to improve their ability to learn unbiased relevance. Experimental results demonstrate that after adopting the search intent hypothesis, click models can better interpret user clicks and substantially improve retrieval performance.

2013 IEEE Energytech, 2013

Today's power systems are undergoing an evolutionary transformation where novel sensors, organize... more Today's power systems are undergoing an evolutionary transformation where novel sensors, organized as sensor networks, are providing tremendous amounts of real-time data that is useful for improving situational awareness. It is becoming increasing evident in large-scale power systems, as in many other applications, that the associated big data is both an opportunity and a challenge. In this paper we describe an approach that merges important concepts from information theory, applied to cyber-physical systems (CPS), with a computational method for modeling large-scale cyber-physical systems, referred to as System Structure. An important aspect of this approach is the interpretation of a large-scale cyber-physical system as an information processing network, where System Structure provides an approach for quantifying interactions between nodes in the network through the computation of system measures.

AIAA Guidance, Navigation, and Control Conference, 2009

Small unmanned aerial vehicles (UAVs), equipped with navigation systems and video capability, are... more Small unmanned aerial vehicles (UAVs), equipped with navigation systems and video capability, are currently being deployed for intelligence, reconnaissance and surveillance missions. One particular mission of interest involves computing location estimates for targets detected by onboard sensors. Combining UAV state estimates with information gathered by the imaging sensors leads to bearing measurements of the target that can be used to determine the target's location. This 3-D bearings-only estimation problem is nonlinear and traditional filtering methods produce biased and uncertain estimates, occasionally leading to filter instabilities. Careful selection of the measurement locations greatly enhances filter performance, motivating the development of UAV trajectories that minimize target location estimation error and improve filter convergence. The objective of this work is to develop guidance algorithms that enable the UAV to fly trajectories that increase the amount of information provided by the measurements and improve overall estimation observability, resulting in proper target tracking and an accurate target location estimate. The performance of the target estimation is dependent upon the positions from which measurements are taken relative to the target and to previous measurements. Past research has provided methods to quantify the information content of a set of measurements using the Fisher Information Matrix (FIM). Forming objective functions based on the FIM and using numerical optimization methods produce UAV trajectories that locally maximize the information content for a given number of measurements. In this project, trajectory optimization leads to the development of UAV flight paths that provide the highest amount of information about the target, while considering sensor restrictions, vehicle dynamics and operation constraints. The UAV trajectory optimization is performed for stationary targets, dynamic targets and multiple targets, for many different scenarios of vehicle motion constraints. The resulting trajectories show spiral paths taken by the UAV, which focus on increasing the angular separation between measurements and reducing the relative range to the target, thus maximizing the information provided by each measurement and improving the performance of the estimation. First of all I would like to thank my advisors Rich Kolacinski and Emilio Frazzoli. This thesis would really not have been possible without your endless support, advice and inspiration. Thank you so much for the frequent meetings and discussions, for always inspiring and guiding me, for giving me new ideas, and for showing me new ways to think. You have really helped me grow as an engineer and I am very grateful to have had the privilege of working with you both. To Brent Appleby, George Schmidt, and Linda Fuhrman, thank you for providing me with the opportunity of being a Draper Fellow. It has been a truly great experience and I really appreciate all your advice and support, with the thesis, career development, and life in general. To my MIT professors, Jon How, Nick Roy and John Deyst, I very much appreciate your guidance and support throughout this thesis. Thank you for always taking the time to advise me and point me in new directions. To Barbara Lechner and Marie Stuppard thanks for always being there to help me out and for all your encouragement and support.

International Journal of Control, 2000

ABSTRACT We investigate the problems of asymptotic stabilization and output regulation of a biomi... more ABSTRACT We investigate the problems of asymptotic stabilization and output regulation of a biomimetic actuation system which independently modulates position and net stiffness. We show how the passivity formalism and the technique of input saturation developed by Lin (1995 and 1996) can be used to asymptotically stabilize this mechanical system over its feasible domain via bounded feedback. Two output regulators, a linear PID controller and a feedback linearizing controller, are developed. The performance of the proposed controllers is illustrated through numerical simulation.

2020 IEEE 14th International Conference on Semantic Computing (ICSC), Feb 1, 2020

Millions of online discussions are generated everyday on social media platforms. Topic modelling ... more Millions of online discussions are generated everyday on social media platforms. Topic modelling is an efficient way of better understanding large text datasets at scale. Conventional topic models have had limited success in online discussions, and to overcome their limitations, we use the discussion thread tree structure and propose a "popularity" metric to quantify the number of replies to a comment to extend the frequency of word occurrences, and the "transitivity" concept to characterize topic dependency among nodes in a nested discussion thread. We build a Conversational Structure Aware Topic Model (CSATM) based on popularity and transitivity to infer topics and their assignments to comments. Experiments on real forum datasets are used to demonstrate improved performance for topic extraction with six different measurements of coherence and impressive accuracy for topic assignments.

2020 IEEE 14th International Conference on Semantic Computing (ICSC)

Millions of online discussions are generated everyday on social media platforms. Topic modelling ... more Millions of online discussions are generated everyday on social media platforms. Topic modelling is an efficient way of better understanding large text datasets at scale. Conventional topic models have had limited success in online discussions, and to overcome their limitations, we use the discussion thread tree structure and propose a "popularity" metric to quantify the number of replies to a comment to extend the frequency of word occurrences, and the "transitivity" concept to characterize topic dependency among nodes in a nested discussion thread. We build a Conversational Structure Aware Topic Model (CSATM) based on popularity and transitivity to infer topics and their assignments to comments. Experiments on real forum datasets are used to demonstrate improved performance for topic extraction with six different measurements of coherence and impressive accuracy for topic assignments.

International Journal of Semantic Computing

With the explosive growth of online discussions published everyday on social media platforms, com... more With the explosive growth of online discussions published everyday on social media platforms, comprehension and discovery of the most popular topics have become a challenging problem. Conventional topic models have had limited success in online discussions because the corpus is extremely sparse and noisy. To overcome their limitations, we use the discussion thread tree structure and propose a “popularity” metric to quantify the number of replies to a comment to extend the frequency of word occurrences, and the “transitivity” concept to characterize topic dependency among nodes in a nested discussion thread. We build a Conversational Structure Aware Topic Model (CSATM) based on popularity and transitivity to infer topics and their assignments to comments. Experiments on real forum datasets are used to demonstrate improved performance for topic extraction with six different measurements of coherence and impressive accuracy for topic assignments.

IEEE Transactions on Industry Applications

Swarm and Evolutionary Computation

2015 Swarm/Human Blended Intelligence Workshop (SHBI), 2015

Proceedings. 1998 IEEE International Conference on Robotics and Automation (Cat. No.98CH36146), 1998

... Recent work by many researchers has demonstrated the advantages of using biomimetic methodolo... more ... Recent work by many researchers has demonstrated the advantages of using biomimetic methodologies for ... A hexapod walking machine [4] using active, variable joint stiffnesses has been ... Examination of (8) reveals that increasing defor-mation, 3, will increase the stiffness as ...

Proceedings of the 18th IFAC World Congress, 2011

This paper focuses upon the development of a methodology for data-driven construction of mesoscop... more This paper focuses upon the development of a methodology for data-driven construction of mesoscopic models of the T&D system for use in real-time monitoring and control. The system dynamics are lifted to a discrete covering space which provides an encoding of the system dynamics within symbol strings. These symbol strings are treated as Bernoulli shifts and are characterized, via the machinery of information theory and formal language theory, as probabilistic automata. As these automata are fundamentally pattern recognizers, they provide a fundamental basis for event/anomaly detection and thus a basis for critical grid monitoring functions such as security state identification.

2012 Future of Instrumentation International Workshop (FIIW) Proceedings, 2012

This paper presents an enterprise architecture that supports the development and deployment of ad... more This paper presents an enterprise architecture that supports the development and deployment of advanced control and condition monitoring algorithms in power generating plants. The architecture is based on information-theoretic concepts that are used to transform multi-modal data streams into actionable information.

Under the boundedness and observability conditions, we present a globally convergent observer for... more Under the boundedness and observability conditions, we present a globally convergent observer for a class of multi-output nonlinear systems which covers the blocktriangular observer forms studied previously in the literature. The result presented in this short note incorporates and generalizes the earlier work on the observer design for singleoutput observable systems.

Typescript. Department of Mechanical and Aerospace Engineering. Thesis (Ph. D.)--Case Western Res... more Typescript. Department of Mechanical and Aerospace Engineering. Thesis (Ph. D.)--Case Western Reserve University, 1997. Includes bibliographical references (leaves 166-169).