Pub Date : 2023-04-17DOI: 10.1109/SysCon53073.2023.10131092
Piyush S Bhagdikar, J. Sarlashkar, Stanislav Gankov, S. Rengarajan, Walter Downing, Scott Hotz
Systems incorporating Vehicle to Everything (V2X) and conventional cellular based communication in vehicles can significantly help improve energy consumption via a combination of intelligent powertrain control strategies, smarter routing algorithms and driving in such a way as to minimize fuel economy and the emission of carbon dioxide, known as "eco-driving." In projects led by the Southwest Research Institute (SwRI), large-scale traffic simulations are created to model real-world scenarios with dynamic behavior that is reactive to imposed changes. Coupled with high fidelity powertrain models, the closed loop framework enables research and development of such Connected and Automated Vehicle (CAV) enabled technologies at scale. This paper will discuss a traffic system simulation environment that was built based on the High Street urban corridor in Columbus, Ohio. Eco-driving strategies were tested at scale on a variety of powertrain platforms – internal combustion engines, hybrid electric and fully electric vehicles. The paper will focus on hybrid electric powertrain modeling along with details on how the powertrain model was leveraged to develop a sophisticated clustering scheme to help down-select speed traces from large scale simulation studies for validation on vehicle dynamometer. Nominal energy consumption improvement around 12% was observed with good match between simulation studies and vehicle testing.
将V2X (Vehicle to Everything)技术和传统的基于蜂窝的车辆通信技术结合在一起的系统，可以通过结合智能动力总成控制策略、更智能的路线算法，以及最大限度地降低燃油经济性和二氧化碳排放的驾驶方式，显著提高能耗，被称为“生态驾驶”。在西南研究所(SwRI)领导的项目中，创建了大规模的交通模拟，以模拟现实世界中的动态行为，这些行为是对强加变化的反应。与高保真动力系统模型相结合，闭环框架使此类联网和自动驾驶汽车(CAV)技术的大规模研究和开发成为可能。本文将讨论基于俄亥俄州哥伦布市高街城市走廊的交通系统仿真环境。环保驾驶策略在各种动力系统平台上进行了大规模测试，包括内燃机、混合动力汽车和全电动汽车。本文将重点介绍混合动力系统建模，以及如何利用动力系统模型开发复杂的聚类方案，以帮助从大规模仿真研究中选择速度轨迹，以便在车辆测力计上进行验证。在模拟研究和车辆测试之间，观察到标称能耗改善约12%。
{"title":"Model Based Validation of Intelligent Powertrain Strategies for Connected and Automated Vehicles","authors":"Piyush S Bhagdikar, J. Sarlashkar, Stanislav Gankov, S. Rengarajan, Walter Downing, Scott Hotz","doi":"10.1109/SysCon53073.2023.10131092","DOIUrl":"https://doi.org/10.1109/SysCon53073.2023.10131092","url":null,"abstract":"Systems incorporating Vehicle to Everything (V2X) and conventional cellular based communication in vehicles can significantly help improve energy consumption via a combination of intelligent powertrain control strategies, smarter routing algorithms and driving in such a way as to minimize fuel economy and the emission of carbon dioxide, known as \"eco-driving.\" In projects led by the Southwest Research Institute (SwRI), large-scale traffic simulations are created to model real-world scenarios with dynamic behavior that is reactive to imposed changes. Coupled with high fidelity powertrain models, the closed loop framework enables research and development of such Connected and Automated Vehicle (CAV) enabled technologies at scale. This paper will discuss a traffic system simulation environment that was built based on the High Street urban corridor in Columbus, Ohio. Eco-driving strategies were tested at scale on a variety of powertrain platforms – internal combustion engines, hybrid electric and fully electric vehicles. The paper will focus on hybrid electric powertrain modeling along with details on how the powertrain model was leveraged to develop a sophisticated clustering scheme to help down-select speed traces from large scale simulation studies for validation on vehicle dynamometer. Nominal energy consumption improvement around 12% was observed with good match between simulation studies and vehicle testing.","PeriodicalId":169296,"journal":{"name":"2023 IEEE International Systems Conference (SysCon)","volume":"114 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116440103","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-04-17DOI: 10.1109/SysCon53073.2023.10131088
J. Colombi, Benjamin C. Donohoo
Modern defense systems are often designed to static sets of operational requirements and system specifications. This process has generally worked well in the past but fails to account for the strategic interdependence of design choices made before developing and deploying systems. Capturing this interdependence of design choices may prove beneficial and cost-effective. This paper demonstrates how Game Theory and physics-based simulation can drive design decisions for non-cooperative systems. Game Theory is the mathematical study of strategy and payoffs between rational, self-interested actors. Simulation and Value Focused Thinking (VFT) provide a Normal Form representation that can be analyzed for the Nash Equilibria. A scenario using space domain awareness demonstrates the method, and results show how changing priorities of the value model change the corresponding stable design point.
{"title":"Game-Theoretic System Design for Non-Cooperative Scenarios","authors":"J. Colombi, Benjamin C. Donohoo","doi":"10.1109/SysCon53073.2023.10131088","DOIUrl":"https://doi.org/10.1109/SysCon53073.2023.10131088","url":null,"abstract":"Modern defense systems are often designed to static sets of operational requirements and system specifications. This process has generally worked well in the past but fails to account for the strategic interdependence of design choices made before developing and deploying systems. Capturing this interdependence of design choices may prove beneficial and cost-effective. This paper demonstrates how Game Theory and physics-based simulation can drive design decisions for non-cooperative systems. Game Theory is the mathematical study of strategy and payoffs between rational, self-interested actors. Simulation and Value Focused Thinking (VFT) provide a Normal Form representation that can be analyzed for the Nash Equilibria. A scenario using space domain awareness demonstrates the method, and results show how changing priorities of the value model change the corresponding stable design point.","PeriodicalId":169296,"journal":{"name":"2023 IEEE International Systems Conference (SysCon)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127061647","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-04-17DOI: 10.1109/SysCon53073.2023.10131057
J. Sarlashkar, B. Surampudi, Venkata R. Chundru, W. Downing
Battery energy storage system (BESS) is a key enabler of the modern renewable- and inverter-heavy electric grid. It facilitates integration of variable power generation such as wind and solar, and can provide a host of ancillary grid services and help defer infrastructure upgrades. Longevity and safety of BESS, however, remain unclear when subjected to such duties diverse in timescale, power, and state-of-charge (SOC). Much of the existing deployment of BESS provide a single grid service such as frequency regulation or load shifting. Further, the contemporary operating envelope of the BESS is deliberately conservative. In the work reported here, we extract essential characteristics of existing field operation and systematically develop methods to estimate the effect of extended range of services on BESS performance and safety. This extended range includes a multitude of ancillary services dispatched simultaneously (stacked duty) and sequentially (mixed duty) using wider envelope of timescale, power, and state-of-charge of BESS. The designed laboratory experiments can be used to construct regression models of BESS performance and safety, and to calibrate parameters of physics-inspired electrochemical representations such as the extended SPMeT model presented in a companion paper [1].
{"title":"Statistical Characterization of Battery Energy Storage Systems in Mixed and Stacked Service Electrical Grid Operations","authors":"J. Sarlashkar, B. Surampudi, Venkata R. Chundru, W. Downing","doi":"10.1109/SysCon53073.2023.10131057","DOIUrl":"https://doi.org/10.1109/SysCon53073.2023.10131057","url":null,"abstract":"Battery energy storage system (BESS) is a key enabler of the modern renewable- and inverter-heavy electric grid. It facilitates integration of variable power generation such as wind and solar, and can provide a host of ancillary grid services and help defer infrastructure upgrades. Longevity and safety of BESS, however, remain unclear when subjected to such duties diverse in timescale, power, and state-of-charge (SOC). Much of the existing deployment of BESS provide a single grid service such as frequency regulation or load shifting. Further, the contemporary operating envelope of the BESS is deliberately conservative. In the work reported here, we extract essential characteristics of existing field operation and systematically develop methods to estimate the effect of extended range of services on BESS performance and safety. This extended range includes a multitude of ancillary services dispatched simultaneously (stacked duty) and sequentially (mixed duty) using wider envelope of timescale, power, and state-of-charge of BESS. The designed laboratory experiments can be used to construct regression models of BESS performance and safety, and to calibrate parameters of physics-inspired electrochemical representations such as the extended SPMeT model presented in a companion paper [1].","PeriodicalId":169296,"journal":{"name":"2023 IEEE International Systems Conference (SysCon)","volume":"85 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127132006","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-04-17DOI: 10.1109/SysCon53073.2023.10131180
T. Hendriks, B. Akesson, J. Voeten, M. Hendriks, Javier Coronel Parada, Miguel García-Gordillo, S. Sáez, J. Valls
Market trends show advanced usage of safety-critical systems with novel services based on smart data analytics. Customers require continuous updates to applications and services and seek lower costs, and easy-to-install solutions (maintenance) for safety-critical cyber-physical systems (CPS). Leveraging edge and cloud technologies has the potential to enhance safety-critical CPS, also in regulated environments. This is only possible when safety, performance, cybersecurity, and privacy of data are kept at the same level as in on-device only safety-critical CPS.This paper presents thirteen selected safety and performance concepts for distributed device-edge-cloud CPS solutions. This early result of the TRANSACT project aims to ensure needed end-to-end performance and safety levels from an end-user perspective, to extend edge and cloud benefits of more rapid innovation and inclusion of value-added services, also to safety-critical CPS.
{"title":"Thirteen concepts to play it safe with the cloud","authors":"T. Hendriks, B. Akesson, J. Voeten, M. Hendriks, Javier Coronel Parada, Miguel García-Gordillo, S. Sáez, J. Valls","doi":"10.1109/SysCon53073.2023.10131180","DOIUrl":"https://doi.org/10.1109/SysCon53073.2023.10131180","url":null,"abstract":"Market trends show advanced usage of safety-critical systems with novel services based on smart data analytics. Customers require continuous updates to applications and services and seek lower costs, and easy-to-install solutions (maintenance) for safety-critical cyber-physical systems (CPS). Leveraging edge and cloud technologies has the potential to enhance safety-critical CPS, also in regulated environments. This is only possible when safety, performance, cybersecurity, and privacy of data are kept at the same level as in on-device only safety-critical CPS.This paper presents thirteen selected safety and performance concepts for distributed device-edge-cloud CPS solutions. This early result of the TRANSACT project aims to ensure needed end-to-end performance and safety levels from an end-user perspective, to extend edge and cloud benefits of more rapid innovation and inclusion of value-added services, also to safety-critical CPS.","PeriodicalId":169296,"journal":{"name":"2023 IEEE International Systems Conference (SysCon)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122260971","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-04-17DOI: 10.1109/SysCon53073.2023.10130845
Charles Mathou, Kevin Delmas, J. Chaudemar, Pierre de Saqui-Sannes
Development of unmanned aerial systems (UAS), made of an unmanned aerial vehicle (UAV) and equipment such as a ground station, has increased tremendously in recent years. This has made more pressing the need for new design methodologies that provide a reliable and thorough safety assessment throughout the entire design process. The European specific operations risk assessment (SORA) document provides recommended operational safety objectives (OSO) to achieve. The current paper lays groundwork to comply with OSOs pertaining to UAS flight procedures. Key criteria for modeling such procedures are identified and lead to the choice of the AltaRica DataFlow (ADF) language. The Cecilia Workshop is used to model three real-life UAS emergency flight procedures. Custom components developed for this model are presented while discussing the process of modeling a formal procedure from an informal text source. A safety analysis is performed on the resulting model by computing minimal cut sets on an undesired procedure outcome. The results are then reviewed, providing feedback to increase the procedures’ safety gain.
{"title":"Modeling UAS Flight Procedures for SORA Safety Objectives","authors":"Charles Mathou, Kevin Delmas, J. Chaudemar, Pierre de Saqui-Sannes","doi":"10.1109/SysCon53073.2023.10130845","DOIUrl":"https://doi.org/10.1109/SysCon53073.2023.10130845","url":null,"abstract":"Development of unmanned aerial systems (UAS), made of an unmanned aerial vehicle (UAV) and equipment such as a ground station, has increased tremendously in recent years. This has made more pressing the need for new design methodologies that provide a reliable and thorough safety assessment throughout the entire design process. The European specific operations risk assessment (SORA) document provides recommended operational safety objectives (OSO) to achieve. The current paper lays groundwork to comply with OSOs pertaining to UAS flight procedures. Key criteria for modeling such procedures are identified and lead to the choice of the AltaRica DataFlow (ADF) language. The Cecilia Workshop is used to model three real-life UAS emergency flight procedures. Custom components developed for this model are presented while discussing the process of modeling a formal procedure from an informal text source. A safety analysis is performed on the resulting model by computing minimal cut sets on an undesired procedure outcome. The results are then reviewed, providing feedback to increase the procedures’ safety gain.","PeriodicalId":169296,"journal":{"name":"2023 IEEE International Systems Conference (SysCon)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"113963625","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-04-17DOI: 10.1109/SysCon53073.2023.10131076
Jin Zhang, Lei Zhang
Spiking Neural Network (SNN) is a particular Artificial Neural Networks (ANN) form. An SNN has similar features as an ANN, but an SNN has a different information system that will allow SNN to have higher energy efficiency than an ANN. This paper presents the design and implementation of an SNN on FPGA. The model of the SNN is designed to be lower power consumption than existing SNN models in the aspect of FPGA implementation and lower accuracy loss than the existing training method in the part of the algorithm. The coding scheme of the SNN model proposed in this paper is the rate coding scheme. This paper introduces a conversion method to directly map the trained parameters from ANN to SNN with negligible classification accuracy loss. Also, this paper demonstrates the technique of FPGA implementation for Spiking Exponential Function, Spiking SoftMax Function and Dynamic Adder Tree. This paper also presents the Time Division Component Reuse technic for lower resource utilization in the FPGA implementation of SNN. The proposed model has a power efficiency of 8841.7 frames per watt with negligible accuracy loss. The benchmark SNN model has a power efficiency of 337.6 frames per watt with an accuracy loss of 1.42 percent. The reference accuracy of the ANN model is 90.36 percent. For comparison, the specific model of the SNN has an accuracy of 90.39 percent.
{"title":"Spiking Neural Network Implementation on FPGA for Multiclass Classification","authors":"Jin Zhang, Lei Zhang","doi":"10.1109/SysCon53073.2023.10131076","DOIUrl":"https://doi.org/10.1109/SysCon53073.2023.10131076","url":null,"abstract":"Spiking Neural Network (SNN) is a particular Artificial Neural Networks (ANN) form. An SNN has similar features as an ANN, but an SNN has a different information system that will allow SNN to have higher energy efficiency than an ANN. This paper presents the design and implementation of an SNN on FPGA. The model of the SNN is designed to be lower power consumption than existing SNN models in the aspect of FPGA implementation and lower accuracy loss than the existing training method in the part of the algorithm. The coding scheme of the SNN model proposed in this paper is the rate coding scheme. This paper introduces a conversion method to directly map the trained parameters from ANN to SNN with negligible classification accuracy loss. Also, this paper demonstrates the technique of FPGA implementation for Spiking Exponential Function, Spiking SoftMax Function and Dynamic Adder Tree. This paper also presents the Time Division Component Reuse technic for lower resource utilization in the FPGA implementation of SNN. The proposed model has a power efficiency of 8841.7 frames per watt with negligible accuracy loss. The benchmark SNN model has a power efficiency of 337.6 frames per watt with an accuracy loss of 1.42 percent. The reference accuracy of the ANN model is 90.36 percent. For comparison, the specific model of the SNN has an accuracy of 90.39 percent.","PeriodicalId":169296,"journal":{"name":"2023 IEEE International Systems Conference (SysCon)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128054764","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-04-17DOI: 10.1109/SysCon53073.2023.10131174
Jacqueline Heaton, S. Givigi
Motion planning and control is a necessary aspect of incorporating robots into the real world. There are a variety of different types of control tasks that involve collision avoidance and fine control, that are difficult to program without the use of artificial intelligence (AI), especially in an non-stationary environment. In this paper, one method for applying deep reinforcement learning (RL) to the motion planning of a manipulator robot is described. Using a soft actor-critic (SAC) network, a model is trained to direct the manipulator to various locations so as to avoid colliding either its hand or the object it carries with a game tower. This demonstrates a simple and effective method for training an agent to achieve its goal that generalizes to similar but different environments.
{"title":"A Deep Reinforcement Learning Solution for the Low Level Motion Control of a Robot Manipulator System","authors":"Jacqueline Heaton, S. Givigi","doi":"10.1109/SysCon53073.2023.10131174","DOIUrl":"https://doi.org/10.1109/SysCon53073.2023.10131174","url":null,"abstract":"Motion planning and control is a necessary aspect of incorporating robots into the real world. There are a variety of different types of control tasks that involve collision avoidance and fine control, that are difficult to program without the use of artificial intelligence (AI), especially in an non-stationary environment. In this paper, one method for applying deep reinforcement learning (RL) to the motion planning of a manipulator robot is described. Using a soft actor-critic (SAC) network, a model is trained to direct the manipulator to various locations so as to avoid colliding either its hand or the object it carries with a game tower. This demonstrates a simple and effective method for training an agent to achieve its goal that generalizes to similar but different environments.","PeriodicalId":169296,"journal":{"name":"2023 IEEE International Systems Conference (SysCon)","volume":"236 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114136438","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-04-17DOI: 10.1109/SysCon53073.2023.10131112
Shayan Sheikhrezaei, H. Yeh, S. Kwon
In this paper, we propose the piecewise technique of Rapidly-exploring Random Tree-Star (P-RRT*) algorithm used in low or medium specification agent(s) (rovers) in the two- dimensional (2-D) workspace. The traditional RRT, RRT*, and other path planning algorithms however efficient they have become; all treat a given environment as a whole and attempt to find a feasible path. This may result in higher memory utilization and a significant increase in processing time.We utilize the RRT* algorithm as the base and integrate it with the piecewise approach. Through P-RRT* technique, given an environment with no obstacles, we attempt to minimize the three vital elements used in the RRT* path planning algorithm (memory, power consumption, and time).A 2D simulation is utilized for demonstration purposes. Given a large workspace, we simulate over subregional workspaces where the number of nodes and step size are adjusted properly to minimize the cost. The simulation results show that dividing the entire simulation workspace into subregions and treating each subregion as a new workspace not only reduces memory utilization and processing time but also the power consumption as a result.The simulation results are shown versus the traditional RRT* algorithm; similar constraints are set for both the piecewise RRT* technique and the traditional RRT* algorithm; meaning that the number of nodes and step size is the same for both methods.
{"title":"Piecewise Rapidly-Exploring Random Tree Star","authors":"Shayan Sheikhrezaei, H. Yeh, S. Kwon","doi":"10.1109/SysCon53073.2023.10131112","DOIUrl":"https://doi.org/10.1109/SysCon53073.2023.10131112","url":null,"abstract":"In this paper, we propose the piecewise technique of Rapidly-exploring Random Tree-Star (P-RRT*) algorithm used in low or medium specification agent(s) (rovers) in the two- dimensional (2-D) workspace. The traditional RRT, RRT*, and other path planning algorithms however efficient they have become; all treat a given environment as a whole and attempt to find a feasible path. This may result in higher memory utilization and a significant increase in processing time.We utilize the RRT* algorithm as the base and integrate it with the piecewise approach. Through P-RRT* technique, given an environment with no obstacles, we attempt to minimize the three vital elements used in the RRT* path planning algorithm (memory, power consumption, and time).A 2D simulation is utilized for demonstration purposes. Given a large workspace, we simulate over subregional workspaces where the number of nodes and step size are adjusted properly to minimize the cost. The simulation results show that dividing the entire simulation workspace into subregions and treating each subregion as a new workspace not only reduces memory utilization and processing time but also the power consumption as a result.The simulation results are shown versus the traditional RRT* algorithm; similar constraints are set for both the piecewise RRT* technique and the traditional RRT* algorithm; meaning that the number of nodes and step size is the same for both methods.","PeriodicalId":169296,"journal":{"name":"2023 IEEE International Systems Conference (SysCon)","volume":"66 7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125869268","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-04-17DOI: 10.1109/SysCon53073.2023.10131227
Nandith Narayan, Parth Ganeriwala, Randolph M. Jones, M. Matessa, S. Bhattacharyya, Jennifer Davis, Hemant Purohit, Simone Fulvio Rollini
Autonomous agents are expected to intelligently handle emerging situations with appropriate interaction with humans, while executing the operations. This is possible today with the integration of advanced technologies, such as machine learning, but these complex algorithms pose a challenge to verification and thus the eventual certification of the autonomous agent. In the discussed approach, we illustrate how safety properties for a learning-enabled increasingly autonomous agent can be formally verified early in the design phase. We demonstrate this methodology by designing a learning-enabled increasingly autonomous agent in a cognitive architecture, Soar. The agent includes symbolic decision logic with numeric decision preferences that are tuned by reinforcement learning to produce post-learning decision knowledge. The agent is then automatically translated into nuXmv, and properties are verified over the agent.
{"title":"Assuring Learning-Enabled Increasingly Autonomous Systems*","authors":"Nandith Narayan, Parth Ganeriwala, Randolph M. Jones, M. Matessa, S. Bhattacharyya, Jennifer Davis, Hemant Purohit, Simone Fulvio Rollini","doi":"10.1109/SysCon53073.2023.10131227","DOIUrl":"https://doi.org/10.1109/SysCon53073.2023.10131227","url":null,"abstract":"Autonomous agents are expected to intelligently handle emerging situations with appropriate interaction with humans, while executing the operations. This is possible today with the integration of advanced technologies, such as machine learning, but these complex algorithms pose a challenge to verification and thus the eventual certification of the autonomous agent. In the discussed approach, we illustrate how safety properties for a learning-enabled increasingly autonomous agent can be formally verified early in the design phase. We demonstrate this methodology by designing a learning-enabled increasingly autonomous agent in a cognitive architecture, Soar. The agent includes symbolic decision logic with numeric decision preferences that are tuned by reinforcement learning to produce post-learning decision knowledge. The agent is then automatically translated into nuXmv, and properties are verified over the agent.","PeriodicalId":169296,"journal":{"name":"2023 IEEE International Systems Conference (SysCon)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130020094","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-04-17DOI: 10.1109/SysCon53073.2023.10131078
R. Chen, Guoxin Wang, Shouxuan Wu, Jinzhi Lu, Yan Yan
When using Model-Based Systems Engineering (MBSE) to develop complex systems, models using different syntax and semantics are typically implemented in a heterogeneous environment which leads to difficulties to realize data integrations across the entire lifecycle. Specifically, seamless exchanges between models of different modeling tools are needed to support system lifecycle activities such as requirement analysis, function analysis, verification and validation. This article illustrates a service-oriented approach to support model integration for model-based systems engineering, especially between architecture design and system verification. First, a set of semantic mapping rules between architecture models and simulation models based on Open Service of Lifecycle Collaboration (OSLC) are proposed to support the formalization of technical resources (models, data, APIs). Then OSLC adapters are developed to transform models, data and APIs into web-based services. The services are deployed by a service discovering plug-in within a specific modeling tool for model information exchange. The approach is illustrated by a case study on KARMA architecture model and Modelica simulation model for a six-degree-of-freedom robot (RobotR3) system. We evaluate the availability and efficiency of this method from both qualitative and quantitative perspectives. The results show that our approach is effective in model and data integration.
当使用基于模型的系统工程(MBSE)开发复杂系统时，使用不同语法和语义的模型通常在异构环境中实现，这会导致难以实现跨整个生命周期的数据集成。具体来说，需要在不同建模工具的模型之间进行无缝交换，以支持系统生命周期活动，例如需求分析、功能分析、验证和确认。本文说明了一种面向服务的方法来支持基于模型的系统工程的模型集成，特别是在体系结构设计和系统验证之间。首先，提出了一套基于生命周期协作开放服务(Open Service of Lifecycle Collaboration, OSLC)的体系结构模型和仿真模型之间的语义映射规则，以支持技术资源(模型、数据、api)的形式化。然后开发OSLC适配器来将模型、数据和api转换为基于web的服务。服务由特定建模工具中的服务发现插件部署，用于模型信息交换。以六自由度机器人(RobotR3)系统的KARMA体系结构模型和Modelica仿真模型为例说明了该方法的可行性。我们从定性和定量两方面评价了该方法的有效性和效率。结果表明，该方法在模型和数据集成方面是有效的。
{"title":"A Service-oriented Approach Supporting Model Integration in Model-based Systems Engineering","authors":"R. Chen, Guoxin Wang, Shouxuan Wu, Jinzhi Lu, Yan Yan","doi":"10.1109/SysCon53073.2023.10131078","DOIUrl":"https://doi.org/10.1109/SysCon53073.2023.10131078","url":null,"abstract":"When using Model-Based Systems Engineering (MBSE) to develop complex systems, models using different syntax and semantics are typically implemented in a heterogeneous environment which leads to difficulties to realize data integrations across the entire lifecycle. Specifically, seamless exchanges between models of different modeling tools are needed to support system lifecycle activities such as requirement analysis, function analysis, verification and validation. This article illustrates a service-oriented approach to support model integration for model-based systems engineering, especially between architecture design and system verification. First, a set of semantic mapping rules between architecture models and simulation models based on Open Service of Lifecycle Collaboration (OSLC) are proposed to support the formalization of technical resources (models, data, APIs). Then OSLC adapters are developed to transform models, data and APIs into web-based services. The services are deployed by a service discovering plug-in within a specific modeling tool for model information exchange. The approach is illustrated by a case study on KARMA architecture model and Modelica simulation model for a six-degree-of-freedom robot (RobotR3) system. We evaluate the availability and efficiency of this method from both qualitative and quantitative perspectives. The results show that our approach is effective in model and data integration.","PeriodicalId":169296,"journal":{"name":"2023 IEEE International Systems Conference (SysCon)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130136509","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：