Despite the non-contact underwater explosion phenomena(UNDEX)have been studied for decades and several numerical methods have been proposed in literature,its effects on military structures,especially composite ones,ar...Despite the non-contact underwater explosion phenomena(UNDEX)have been studied for decades and several numerical methods have been proposed in literature,its effects on military structures,especially composite ones,are even nowadays matter of research.In early design phases,it is not always possible to verify the shock resistance of hull structures modelling the whole phenomenon,in which fluid,gas and solid properties must be properly set in a fully coupled fluid-structure interaction(FSI)numerical model.These ones are extremely complex to set,computationally demanding and certainly not suitable for everyday design practice.In this paper,a simplified finite element(FE)model,easy to use in an early design phase,is proposed.Both,the structure and the fluid are simulated.In this approximation,the fluid behaviour is simplified,using special finite elements,available in a commercial software environment.This choice reduces the computational time and numerical efforts avoiding the problem of combining computational fluid dynamics(CFD)and FE domains and equations in a fully coupled fluid-structure interaction model.A typical parallel body block of a minesweeper is modelled,using two-dimensional multi-layered shell elements to properly account for the composite materials behaviour.For the fluid instead,three dimensional volumetric elements,directly coupled to the structural elements,are placed.In addition,the same calculation is performed,modelling separately fluid in the CFD environment and structures in the finite element one.Thus,realizing a fully coupled fluid-structure interaction model.The results obtained by applying both numerical models are compared with the structural response measured on board of a composite ship during a full-scale shock test.The simplified proposed procedure provides results in satisfactory agreement with experiments,allowing the validation of the model.Approximations are discussed and differences with the real phenomenon and fully coupled CFD+FE method are shown,providing a better understanding of the phenomena.Eventually,the modelling strategy has been considered a valuable and cost-effective tool for the concept and preliminary design of composite structures subject to underwater explosions.展开更多
Satellite communication networks have been evolving from standalone networks with ad-hoc infrastructures to possibly interconnected portions of a wider Future Internet architecture. Experts belonging to the fifth-gene...Satellite communication networks have been evolving from standalone networks with ad-hoc infrastructures to possibly interconnected portions of a wider Future Internet architecture. Experts belonging to the fifth-generation(5 G) standardization committees are considering satellites as a technology to integrate in the 5 G environment. Software Defined Networking(SDN) is one of the paradigms of the next generation of mobile and fixed communications. It can be employed to perform different control functionalities, such as routing, because it allows traffic flow identification based on different parameters and traffic flow management in a centralized way. A centralized set of controllers makes the decisions and sends the corresponding forwarding rules for each traffic flow to the involved intermediate nodes that practically forward data up to the destination. The time to perform this process in integrated terrestrial-satellite networks could be not negligible due to satellite link delays. The aim of this paper is to introduce an SDN-based terrestrial satellite network architecture and to estimate the mean time to deliver the data of a new traffic flow from the source to the destination including the time required to transfer SDN control actions. The practical effect is to identify the maximum performance than can be expected.展开更多
Correct evaluation of rudder performance is a key issue in assessing ship maneuverability.This paper presents a simplified approach based on a viscous flow solver to address propeller and rudder interactions.Viscous f...Correct evaluation of rudder performance is a key issue in assessing ship maneuverability.This paper presents a simplified approach based on a viscous flow solver to address propeller and rudder interactions.Viscous flow solvers have been applied to this type of problems,but the large computational requests limit(or even prevent)their application at a preliminary ship design stage.Based on this idea,a simplified approach to include the propeller effect in front of the rudder is considered to speed up the solution.Based on the concept of body forces,this approach enables sufficiently fast computation for a preliminary ship design stage,therebymaintaining its reliability.To define the limitations of the proposed procedure,an extensive analysis of the simplified method is performed and the results are compared with experimental data presented in the literature.Initially,the reported results show the capability of the body-force approach to represent the inflow field to the rudder without the full description of the propeller,also with regard to the complex bollard pull condition.Consequently,the rudder forces are satisfactorily predicted at least with regard to the lift force.However,the drag force evaluation ismore problematic and causes higher discrepancies.Nevertheless,these discrepancies may be accepted due to their lower influence on the overall ship maneuverability performance.展开更多
The current design philosophy for submarine hulls,in the preliminary design stage,generally considers as governing limit states material yielding along with various buckling modes.It is common belief that,beyond the d...The current design philosophy for submarine hulls,in the preliminary design stage,generally considers as governing limit states material yielding along with various buckling modes.It is common belief that,beyond the design pressure,material yielding of the shell plating should occur first,eventually followed by local buckling,while global buckling currently retains the highest safety factor.On the other hand,in the aeronautical field,in some cases structural components are designed in such a way that local instability may occur within the design loads,being the phenomena inside the material elastic range and not leading to a significant drop in term of stiffness.This paper is aimed at investigating the structural response beyond a set of selected limit states,using nonlinear FE method adopting different initial imperfection models,to provide the designers with new information useful for calibrating safety factors.It was found that both local and global buckling can be considered as ultimate limit states,with a significant sensitivity towards initial imperfection,while material yielding and tripping buckling of frames show a residual structural capacity.In conclusion,it was found that the occurrence of local buckling leads to similar sudden catastrophic consequences as global buckling,with the ultimate strength capacity highly affected by the initial imperfection shape and amplitude.展开更多
The hydrodynamic analysis of a new semi-small waterplane area twin hull （SWATH） suitable for various applications such as small and medium size passenger ferries is presented. This may be an attractive crossover con...The hydrodynamic analysis of a new semi-small waterplane area twin hull （SWATH） suitable for various applications such as small and medium size passenger ferries is presented. This may be an attractive crossover configuration resulting from the merging of two classical shapes： a conventional SWATH and a fast catamaran. The final hull design exhibits a wedge-like waterline shape with the maximum beam at the stem; the hull ends with a very narrow entrance angle, has a prominent bulbous bow typical of SWATH vessels, and features full stern to arrange waterjet propellers. Our analysis aims to perform a preliminary assessment of the hydrodynamic performance of a hull with such a complex shape both in terms of resistance of the hull in calm water and seakeeping capability in regular head waves and compare the performance with that of a conventional SWATH. The analysis is performed using a boundary element method that was preliminarily validated on a conventional SWATH vessel.展开更多
Comfort levels on modern superyachts have recently been the object of specific attention of the most important Classification Societies, which issued new rules and regulations for evaluating noise and vibration maximu...Comfort levels on modern superyachts have recently been the object of specific attention of the most important Classification Societies, which issued new rules and regulations for evaluating noise and vibration maximum levels. These rules are named "Comfort Class Rules" and set the general criteria for noise and vibration measurements in different vessels' areas, as well as the maximum noise and vibration limit values. As far as the vibration assessment is concerned, the Comfort Class Rules follow either the ISO 6954:1984 standard or the ISO 6954:2000. After an introduction to these relevant standards, the authors herein present a procedure developed to predict the vibration levels on ships. This procedure builds on finite element linear dynamic analysis and is applied to predict the vibration levels on a 60 m superyacht considered as a case study. The results of the numerical simulations are then benchmarked against experimental data acquired during the sea trial of the vessel. This analysis also allows the authors to evaluate the global damping ratio to be used by designers in the vibration analysis of superyachts.展开更多
Goal based and limit state design is nowadays a well-established approach in many engineering fields.Ship construction rules started introducing such concepts since early 2000.However,classification societies’rules d...Goal based and limit state design is nowadays a well-established approach in many engineering fields.Ship construction rules started introducing such concepts since early 2000.However,classification societies’rules do not provide hints on how to verify limit states and to determine the structural layout of submerged thin-walled stiffened cylinders,whose most prominent examples are submarines.Rather,they generally offer guidance and prescriptive formulations to assess shell plating and stiffening members.Such marine structures are studied,designed and built up to carry payloads below the sea surface.In the concept-design stage,the maximum operating depth is the governing hull scantling parameter.Main dimensions are determined based on the analysis of operational requirements.This study proposes a practical conceptdesign approach for conceptual submarine design,aimed at obtaining hull structures that maximize the payload capacity in terms of available internal volume by suitably adjusting structural layout and stiffening members’scantling,duly accounting for robustness and construction constraints as well as practical fabrication issues.The proposed scantling process highlights that there is no need of complex algorithms if sound engineering judgment is applied in setting down rationally the hull scantling problem.A systematic approach based on a computer-coded procedure developed on purpose was effectively implemented and satisfactorily applied in design practice.展开更多
In this paper, we report our study on a numerical fluid-structure interaction problem originally presented by Mok et al.(2001) in two dimensions and later studied in three dimensions by Valdés Vazquez(2007), Lomb...In this paper, we report our study on a numerical fluid-structure interaction problem originally presented by Mok et al.(2001) in two dimensions and later studied in three dimensions by Valdés Vazquez(2007), Lombardi(2012), and Trimarchi(2012). We focus on a 3D test case in which we evaluated the sensitivity of several input parameters on the fluid and structural results. In particular, this analysis provides a starting point from which we can look deeper into specific aspects of these simulations and analyze more realistic cases, e.g., in sails design. In this study, using the commercial software ADINATM, we addressed a well-known unsteadiness problem comprising a square box representing the fluid domain with a flexible bottom modeled with structural shell elements. We compared data from previously published work whose authors used the same numerical approach, i.e., a partitioned approach coupling a finite volume solver(for the fluid domain) and a finite element solver(for the solid domain). Specifically, we established several benchmarks and made comparisons with respect to fluid and solid meshes, structural element types, and structural damping, as well as solution algorithms. Moreover, we compared our method with a monolithic finite element solution method. Our comparisons of new and old results provide an outline of best practices for such simulations.展开更多
To further exploit the potential of marine composites applications in building ship hulls,offshore structures,and marine equipment and components,design approaches should be improved,facing the challenge of a more com...To further exploit the potential of marine composites applications in building ship hulls,offshore structures,and marine equipment and components,design approaches should be improved,facing the challenge of a more comprehensive and explicit assessment of appropriately defined limit states.The structure ultimate/limit conditions shall be verified in principle within the whole structural domain and throughout the ship service life.What above calls for extended and reliable materials characterization on the one hand and for accurate and wide-ranging procedures in structural analyses.This paper presents an overview of recent industrial developments of marine composites limit states assessments and design approaches,as available in open literature,focusing on pleasure crafts and yachts as well as navy ships and thus showing a starting point to fill the gap in this respect.After a general introduction about composites characterization techniques,current design practice and rule requirements are briefly summarized.Both inter-ply and intra-ply failure modes and corresponding limit states are then presented along with recently proposed assessment approaches.Three-dimensional aspects in failure modes and manufacturing methods have been identified as the main factors influencing marine composite robustness.Literature review highlighted also fire resistance and hybrid joining techniques as significant issues in the use of marine composites.展开更多
Industrial Control Systems(ICS)in Operational Technology(OT)environments face unique cybersecurity challenges due to legacy systems,critical operational needs,and incompatibility with standard IT security practices.To...Industrial Control Systems(ICS)in Operational Technology(OT)environments face unique cybersecurity challenges due to legacy systems,critical operational needs,and incompatibility with standard IT security practices.To address these challenges,this paper presents the Security Operation and Event Management(SOEM)platform,a software designed to support Security Operations Centers(SOCs)in reaching full visibility of OT environments.SOEM integrates diverse log sources and intrusion detection systems,including logs generated by the control system itself and additional on-the-shelf products,to enhance situational awareness and enable rapid incident response.The pilot project was carried out within the funded project SOC-OT-IGE from the“Centro di Competenza Start 4.0”and is being developed in partnership with Ansaldo Energia and HWG Sababa.The validation has been conducted in a real-world pilot project.Thanks to the mapping to requirements for compliance with IEC 62443,the platform demonstrates its effectiveness through defined key performance indicators(KPIs).This work bridges the gap between IT-centric SOC methodologies and the specialized needs of industrial cybersecurity.展开更多
This study provides a systematic analysis of the resource-consuming training of deep reinforcement-learning (DRL) agents for simulated low-speed automated driving (AD). In Unity, this study established two case studie...This study provides a systematic analysis of the resource-consuming training of deep reinforcement-learning (DRL) agents for simulated low-speed automated driving (AD). In Unity, this study established two case studies: garage parking and navigating an obstacle-dense area. Our analysis involves training a path-planning agent with real-time-only sensor information. This study addresses research questions insufficiently covered in the literature, exploring curriculum learning (CL), agent generalization (knowledge transfer), computation distribution (CPU vs. GPU), and mapless navigation. CL proved necessary for the garage scenario and beneficial for obstacle avoidance. It involved adjustments at different stages, including terminal conditions, environment complexity, and reward function hyperparameters, guided by their evolution in multiple training attempts. Fine-tuning the simulation tick and decision period parameters was crucial for effective training. The abstraction of high-level concepts (e.g., obstacle avoidance) necessitates training the agent in sufficiently complex environments in terms of the number of obstacles. While blogs and forums discuss training machine learning models in Unity, a lack of scientific articles on DRL agents for AD persists. However, since agent development requires considerable training time and difficult procedures, there is a growing need to support such research through scientific means. In addition to our findings, we contribute to the R&D community by providing our environment with open sources.展开更多
In this study, we propose a hybrid AI optimal method to improve the efficiency of energy managementin a smart grid such as Renewable Energy Community. This method adopts a Time Delay Neural Networkto forecast the futu...In this study, we propose a hybrid AI optimal method to improve the efficiency of energy managementin a smart grid such as Renewable Energy Community. This method adopts a Time Delay Neural Networkto forecast the future values of the energy features in the community. Then, these forecasts are used by astochastic Model Predictive Control to optimize the community operations with a proper control strategy ofBattery Energy Storage System. The results of the predictions performed on a public dataset with a predictionhorizon of 24 h return a Mean Absolute Error of 1.60 kW, 2.15 kW, and 0.30 kW for photovoltaic generation,total energy consumption, and common services, respectively. The model predictive control fed with suchpredictions generates maximum income compared to the competitors. The total income is increased by 18.72%compared to utilizing the same management system without exploiting predictions from a forecasting method.展开更多
文摘Despite the non-contact underwater explosion phenomena(UNDEX)have been studied for decades and several numerical methods have been proposed in literature,its effects on military structures,especially composite ones,are even nowadays matter of research.In early design phases,it is not always possible to verify the shock resistance of hull structures modelling the whole phenomenon,in which fluid,gas and solid properties must be properly set in a fully coupled fluid-structure interaction(FSI)numerical model.These ones are extremely complex to set,computationally demanding and certainly not suitable for everyday design practice.In this paper,a simplified finite element(FE)model,easy to use in an early design phase,is proposed.Both,the structure and the fluid are simulated.In this approximation,the fluid behaviour is simplified,using special finite elements,available in a commercial software environment.This choice reduces the computational time and numerical efforts avoiding the problem of combining computational fluid dynamics(CFD)and FE domains and equations in a fully coupled fluid-structure interaction model.A typical parallel body block of a minesweeper is modelled,using two-dimensional multi-layered shell elements to properly account for the composite materials behaviour.For the fluid instead,three dimensional volumetric elements,directly coupled to the structural elements,are placed.In addition,the same calculation is performed,modelling separately fluid in the CFD environment and structures in the finite element one.Thus,realizing a fully coupled fluid-structure interaction model.The results obtained by applying both numerical models are compared with the structural response measured on board of a composite ship during a full-scale shock test.The simplified proposed procedure provides results in satisfactory agreement with experiments,allowing the validation of the model.Approximations are discussed and differences with the real phenomenon and fully coupled CFD+FE method are shown,providing a better understanding of the phenomena.Eventually,the modelling strategy has been considered a valuable and cost-effective tool for the concept and preliminary design of composite structures subject to underwater explosions.
文摘Satellite communication networks have been evolving from standalone networks with ad-hoc infrastructures to possibly interconnected portions of a wider Future Internet architecture. Experts belonging to the fifth-generation(5 G) standardization committees are considering satellites as a technology to integrate in the 5 G environment. Software Defined Networking(SDN) is one of the paradigms of the next generation of mobile and fixed communications. It can be employed to perform different control functionalities, such as routing, because it allows traffic flow identification based on different parameters and traffic flow management in a centralized way. A centralized set of controllers makes the decisions and sends the corresponding forwarding rules for each traffic flow to the involved intermediate nodes that practically forward data up to the destination. The time to perform this process in integrated terrestrial-satellite networks could be not negligible due to satellite link delays. The aim of this paper is to introduce an SDN-based terrestrial satellite network architecture and to estimate the mean time to deliver the data of a new traffic flow from the source to the destination including the time required to transfer SDN control actions. The practical effect is to identify the maximum performance than can be expected.
文摘Correct evaluation of rudder performance is a key issue in assessing ship maneuverability.This paper presents a simplified approach based on a viscous flow solver to address propeller and rudder interactions.Viscous flow solvers have been applied to this type of problems,but the large computational requests limit(or even prevent)their application at a preliminary ship design stage.Based on this idea,a simplified approach to include the propeller effect in front of the rudder is considered to speed up the solution.Based on the concept of body forces,this approach enables sufficiently fast computation for a preliminary ship design stage,therebymaintaining its reliability.To define the limitations of the proposed procedure,an extensive analysis of the simplified method is performed and the results are compared with experimental data presented in the literature.Initially,the reported results show the capability of the body-force approach to represent the inflow field to the rudder without the full description of the propeller,also with regard to the complex bollard pull condition.Consequently,the rudder forces are satisfactorily predicted at least with regard to the lift force.However,the drag force evaluation ismore problematic and causes higher discrepancies.Nevertheless,these discrepancies may be accepted due to their lower influence on the overall ship maneuverability performance.
基金The research activity on this topic is still under development in the frame of the ASAMS(Aspetti specialistici e approccio metodologico per progettazione di sottomarini di ultima generazione)project(2019-2022)which has been funded by the Italian MoD–Segredifesa,in collaboration with Fincantieri.
文摘The current design philosophy for submarine hulls,in the preliminary design stage,generally considers as governing limit states material yielding along with various buckling modes.It is common belief that,beyond the design pressure,material yielding of the shell plating should occur first,eventually followed by local buckling,while global buckling currently retains the highest safety factor.On the other hand,in the aeronautical field,in some cases structural components are designed in such a way that local instability may occur within the design loads,being the phenomena inside the material elastic range and not leading to a significant drop in term of stiffness.This paper is aimed at investigating the structural response beyond a set of selected limit states,using nonlinear FE method adopting different initial imperfection models,to provide the designers with new information useful for calibrating safety factors.It was found that both local and global buckling can be considered as ultimate limit states,with a significant sensitivity towards initial imperfection,while material yielding and tripping buckling of frames show a residual structural capacity.In conclusion,it was found that the occurrence of local buckling leads to similar sudden catastrophic consequences as global buckling,with the ultimate strength capacity highly affected by the initial imperfection shape and amplitude.
文摘The hydrodynamic analysis of a new semi-small waterplane area twin hull （SWATH） suitable for various applications such as small and medium size passenger ferries is presented. This may be an attractive crossover configuration resulting from the merging of two classical shapes： a conventional SWATH and a fast catamaran. The final hull design exhibits a wedge-like waterline shape with the maximum beam at the stem; the hull ends with a very narrow entrance angle, has a prominent bulbous bow typical of SWATH vessels, and features full stern to arrange waterjet propellers. Our analysis aims to perform a preliminary assessment of the hydrodynamic performance of a hull with such a complex shape both in terms of resistance of the hull in calm water and seakeeping capability in regular head waves and compare the performance with that of a conventional SWATH. The analysis is performed using a boundary element method that was preliminarily validated on a conventional SWATH vessel.
文摘Comfort levels on modern superyachts have recently been the object of specific attention of the most important Classification Societies, which issued new rules and regulations for evaluating noise and vibration maximum levels. These rules are named "Comfort Class Rules" and set the general criteria for noise and vibration measurements in different vessels' areas, as well as the maximum noise and vibration limit values. As far as the vibration assessment is concerned, the Comfort Class Rules follow either the ISO 6954:1984 standard or the ISO 6954:2000. After an introduction to these relevant standards, the authors herein present a procedure developed to predict the vibration levels on ships. This procedure builds on finite element linear dynamic analysis and is applied to predict the vibration levels on a 60 m superyacht considered as a case study. The results of the numerical simulations are then benchmarked against experimental data acquired during the sea trial of the vessel. This analysis also allows the authors to evaluate the global damping ratio to be used by designers in the vibration analysis of superyachts.
基金Supported by the Italian Ministry of Defense-Segredifesa,in collaboration with Fincantieri under Grant of the ASAMS(Aspetti specialistici e approccio metodologico per progettazione di sottomarini di ultima generazione)project(2019-2022).
文摘Goal based and limit state design is nowadays a well-established approach in many engineering fields.Ship construction rules started introducing such concepts since early 2000.However,classification societies’rules do not provide hints on how to verify limit states and to determine the structural layout of submerged thin-walled stiffened cylinders,whose most prominent examples are submarines.Rather,they generally offer guidance and prescriptive formulations to assess shell plating and stiffening members.Such marine structures are studied,designed and built up to carry payloads below the sea surface.In the concept-design stage,the maximum operating depth is the governing hull scantling parameter.Main dimensions are determined based on the analysis of operational requirements.This study proposes a practical conceptdesign approach for conceptual submarine design,aimed at obtaining hull structures that maximize the payload capacity in terms of available internal volume by suitably adjusting structural layout and stiffening members’scantling,duly accounting for robustness and construction constraints as well as practical fabrication issues.The proposed scantling process highlights that there is no need of complex algorithms if sound engineering judgment is applied in setting down rationally the hull scantling problem.A systematic approach based on a computer-coded procedure developed on purpose was effectively implemented and satisfactorily applied in design practice.
文摘In this paper, we report our study on a numerical fluid-structure interaction problem originally presented by Mok et al.(2001) in two dimensions and later studied in three dimensions by Valdés Vazquez(2007), Lombardi(2012), and Trimarchi(2012). We focus on a 3D test case in which we evaluated the sensitivity of several input parameters on the fluid and structural results. In particular, this analysis provides a starting point from which we can look deeper into specific aspects of these simulations and analyze more realistic cases, e.g., in sails design. In this study, using the commercial software ADINATM, we addressed a well-known unsteadiness problem comprising a square box representing the fluid domain with a flexible bottom modeled with structural shell elements. We compared data from previously published work whose authors used the same numerical approach, i.e., a partitioned approach coupling a finite volume solver(for the fluid domain) and a finite element solver(for the solid domain). Specifically, we established several benchmarks and made comparisons with respect to fluid and solid meshes, structural element types, and structural damping, as well as solution algorithms. Moreover, we compared our method with a monolithic finite element solution method. Our comparisons of new and old results provide an outline of best practices for such simulations.
文摘To further exploit the potential of marine composites applications in building ship hulls,offshore structures,and marine equipment and components,design approaches should be improved,facing the challenge of a more comprehensive and explicit assessment of appropriately defined limit states.The structure ultimate/limit conditions shall be verified in principle within the whole structural domain and throughout the ship service life.What above calls for extended and reliable materials characterization on the one hand and for accurate and wide-ranging procedures in structural analyses.This paper presents an overview of recent industrial developments of marine composites limit states assessments and design approaches,as available in open literature,focusing on pleasure crafts and yachts as well as navy ships and thus showing a starting point to fill the gap in this respect.After a general introduction about composites characterization techniques,current design practice and rule requirements are briefly summarized.Both inter-ply and intra-ply failure modes and corresponding limit states are then presented along with recently proposed assessment approaches.Three-dimensional aspects in failure modes and manufacturing methods have been identified as the main factors influencing marine composite robustness.Literature review highlighted also fire resistance and hybrid joining techniques as significant issues in the use of marine composites.
基金supported by the project“Airfield”under the PoC Launchpad initiative funded by the Fondazione Compagnia di San Paolo.
文摘Industrial Control Systems(ICS)in Operational Technology(OT)environments face unique cybersecurity challenges due to legacy systems,critical operational needs,and incompatibility with standard IT security practices.To address these challenges,this paper presents the Security Operation and Event Management(SOEM)platform,a software designed to support Security Operations Centers(SOCs)in reaching full visibility of OT environments.SOEM integrates diverse log sources and intrusion detection systems,including logs generated by the control system itself and additional on-the-shelf products,to enhance situational awareness and enable rapid incident response.The pilot project was carried out within the funded project SOC-OT-IGE from the“Centro di Competenza Start 4.0”and is being developed in partnership with Ansaldo Energia and HWG Sababa.The validation has been conducted in a real-world pilot project.Thanks to the mapping to requirements for compliance with IEC 62443,the platform demonstrates its effectiveness through defined key performance indicators(KPIs).This work bridges the gap between IT-centric SOC methodologies and the specialized needs of industrial cybersecurity.
文摘This study provides a systematic analysis of the resource-consuming training of deep reinforcement-learning (DRL) agents for simulated low-speed automated driving (AD). In Unity, this study established two case studies: garage parking and navigating an obstacle-dense area. Our analysis involves training a path-planning agent with real-time-only sensor information. This study addresses research questions insufficiently covered in the literature, exploring curriculum learning (CL), agent generalization (knowledge transfer), computation distribution (CPU vs. GPU), and mapless navigation. CL proved necessary for the garage scenario and beneficial for obstacle avoidance. It involved adjustments at different stages, including terminal conditions, environment complexity, and reward function hyperparameters, guided by their evolution in multiple training attempts. Fine-tuning the simulation tick and decision period parameters was crucial for effective training. The abstraction of high-level concepts (e.g., obstacle avoidance) necessitates training the agent in sufficiently complex environments in terms of the number of obstacles. While blogs and forums discuss training machine learning models in Unity, a lack of scientific articles on DRL agents for AD persists. However, since agent development requires considerable training time and difficult procedures, there is a growing need to support such research through scientific means. In addition to our findings, we contribute to the R&D community by providing our environment with open sources.
文摘In this study, we propose a hybrid AI optimal method to improve the efficiency of energy managementin a smart grid such as Renewable Energy Community. This method adopts a Time Delay Neural Networkto forecast the future values of the energy features in the community. Then, these forecasts are used by astochastic Model Predictive Control to optimize the community operations with a proper control strategy ofBattery Energy Storage System. The results of the predictions performed on a public dataset with a predictionhorizon of 24 h return a Mean Absolute Error of 1.60 kW, 2.15 kW, and 0.30 kW for photovoltaic generation,total energy consumption, and common services, respectively. The model predictive control fed with suchpredictions generates maximum income compared to the competitors. The total income is increased by 18.72%compared to utilizing the same management system without exploiting predictions from a forecasting method.
