Pub Date : 2020-01-01DOI: 10.1109/RAMS48030.2020.9153718
Maskura Nafreen, Saikath Bhattacharya, L. Fiondella
With the increased interest to incorporate machine learning into software and systems, methods to characterize the impact of the reliability of machine learning are needed to ensure the reliability of the software and systems in which these algorithms reside. Towards this end, we build upon the architecture-based approach to software reliability modeling, which represents application reliability in terms of the component reliabilities and the probabilistic transitions between the components. Traditional architecture-based software reliability models consider all components to be deterministic software. We therefore extend this modeling approach to the case, where some components represent learning enabled components. Here, the reliability of a machine learning component is interpreted as the accuracy of its decisions, which is a common measure of classification algorithms. Moreover, we allow these machine learning components to be fault-tolerant in the sense that multiple diverse classifier algorithms are trained to guide decisions and the majority decision taken. We demonstrate the utility of the approach to assess the impact of machine learning on software reliability as well as illustrate the concept of reliability growth in machine learning. Finally, we validate past analytical results for a fault tolerant system composed of correlated components with real machine learning algorithms and data, demonstrating the analytical expression’s ability to accurately estimate the reliability of the fault tolerant machine learning component and subsequently the architecture-based software within which it resides.
{"title":"Architecture-based Software Reliability Incorporating Fault Tolerant Machine Learning","authors":"Maskura Nafreen, Saikath Bhattacharya, L. Fiondella","doi":"10.1109/RAMS48030.2020.9153718","DOIUrl":"https://doi.org/10.1109/RAMS48030.2020.9153718","url":null,"abstract":"With the increased interest to incorporate machine learning into software and systems, methods to characterize the impact of the reliability of machine learning are needed to ensure the reliability of the software and systems in which these algorithms reside. Towards this end, we build upon the architecture-based approach to software reliability modeling, which represents application reliability in terms of the component reliabilities and the probabilistic transitions between the components. Traditional architecture-based software reliability models consider all components to be deterministic software. We therefore extend this modeling approach to the case, where some components represent learning enabled components. Here, the reliability of a machine learning component is interpreted as the accuracy of its decisions, which is a common measure of classification algorithms. Moreover, we allow these machine learning components to be fault-tolerant in the sense that multiple diverse classifier algorithms are trained to guide decisions and the majority decision taken. We demonstrate the utility of the approach to assess the impact of machine learning on software reliability as well as illustrate the concept of reliability growth in machine learning. Finally, we validate past analytical results for a fault tolerant system composed of correlated components with real machine learning algorithms and data, demonstrating the analytical expression’s ability to accurately estimate the reliability of the fault tolerant machine learning component and subsequently the architecture-based software within which it resides.","PeriodicalId":360096,"journal":{"name":"2020 Annual Reliability and Maintainability Symposium (RAMS)","volume":"184 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116202533","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 : 2020-01-01DOI: 10.1109/RAMS48030.2020.9153606
Zekun Song, Yichen Wang, P. Zong, Lin Wang, G. Feng, Wenqian Kang
How to evaluate software reliability based on historical data of embedded software projects is one of the problems we have to face in practical engineering. Therefore, we establish a software reliability evaluation model based on code metrics. The model uses code metrics to score software reliability. This evaluation technique requires the aggregation of software code metrics into project metrics. What are the differences among different aggregation methods in the software reliability evaluation process, and which methods can improve the accuracy of the reliability evaluation model we have established are our concerns. In view of the above problems, we conduct an empirical study on the application of software code metric aggregation methods based on actual projects.
{"title":"An Empirical Study of Comparison of Code Metric Aggregation Methods and Software Reliability Evaluation","authors":"Zekun Song, Yichen Wang, P. Zong, Lin Wang, G. Feng, Wenqian Kang","doi":"10.1109/RAMS48030.2020.9153606","DOIUrl":"https://doi.org/10.1109/RAMS48030.2020.9153606","url":null,"abstract":"How to evaluate software reliability based on historical data of embedded software projects is one of the problems we have to face in practical engineering. Therefore, we establish a software reliability evaluation model based on code metrics. The model uses code metrics to score software reliability. This evaluation technique requires the aggregation of software code metrics into project metrics. What are the differences among different aggregation methods in the software reliability evaluation process, and which methods can improve the accuracy of the reliability evaluation model we have established are our concerns. In view of the above problems, we conduct an empirical study on the application of software code metric aggregation methods based on actual projects.","PeriodicalId":360096,"journal":{"name":"2020 Annual Reliability and Maintainability Symposium (RAMS)","volume":"526 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116220366","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}
Diesel engine is the power source of warship and the core part of power system. Diesel engine not only has complex fuselage structure and many moving parts, but also is in a worse operating environment than other parts.
{"title":"Fault Diagnosis and Prediction Method for Valve Clearance of Diesel Engine Based on Linear Regression","authors":"Yinglai Liu, Wenbing Chang, Siyue Zhang, Shenghan Zhou","doi":"10.1109/RAMS48030.2020.9153697","DOIUrl":"https://doi.org/10.1109/RAMS48030.2020.9153697","url":null,"abstract":"Diesel engine is the power source of warship and the core part of power system. Diesel engine not only has complex fuselage structure and many moving parts, but also is in a worse operating environment than other parts.","PeriodicalId":360096,"journal":{"name":"2020 Annual Reliability and Maintainability Symposium (RAMS)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123720413","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 : 2020-01-01DOI: 10.1109/RAMS48030.2020.9153625
J. Weglian, J. Riley, F. Ferrante
SUMMARY & CONCLUSIONSCommercial nuclear power plants use probabilistic risk assessment (PRA) models to gain insights into the risks associated with operating the plants. PRA models can be used to assess a variety of hazards such as internal events (transients and loss of coolant accidents), internal flooding, fire, seismic, and other hazards. Each model can provide risk insights, identify vulnerabilities, and identify significant equipment or operator actions. This information can be used to improve plant performance and safety via equipment or operational changes. It is often convenient, and for some risk-informed regulations it may be required, to combine all hazard PRA models into a single calculational model. This “one-top” model provides a single PRA fault tree that can be solved to generate the risk for all hazards. Combining these models can be a challenge, if they were built with different revisions to the internal events model at their core. The one-top model provides a convenient platform for assessing the risk from all of the modeled hazards in a single quantification. Certain software tools, such as the EPRI FRANX software, simplify the process of creating a one-top model. However, quantifying a one-top model has challenges, because each hazard model is built with different assumptions, data, biases, and uncertainty. Furthermore, when one hazard generates a risk value much larger than the risk value from another hazard, combining the results runs the risk of masking risk insights from the hazard with the smaller risk value. In addition to quantifying the one-top model for risk-informed applications that require or benefit from it, hazard models should still be quantified separately to get the risk insights the individual models provide.
{"title":"Combining Hazards into a Single-Top Fault Tree","authors":"J. Weglian, J. Riley, F. Ferrante","doi":"10.1109/RAMS48030.2020.9153625","DOIUrl":"https://doi.org/10.1109/RAMS48030.2020.9153625","url":null,"abstract":"SUMMARY & CONCLUSIONSCommercial nuclear power plants use probabilistic risk assessment (PRA) models to gain insights into the risks associated with operating the plants. PRA models can be used to assess a variety of hazards such as internal events (transients and loss of coolant accidents), internal flooding, fire, seismic, and other hazards. Each model can provide risk insights, identify vulnerabilities, and identify significant equipment or operator actions. This information can be used to improve plant performance and safety via equipment or operational changes. It is often convenient, and for some risk-informed regulations it may be required, to combine all hazard PRA models into a single calculational model. This “one-top” model provides a single PRA fault tree that can be solved to generate the risk for all hazards. Combining these models can be a challenge, if they were built with different revisions to the internal events model at their core. The one-top model provides a convenient platform for assessing the risk from all of the modeled hazards in a single quantification. Certain software tools, such as the EPRI FRANX software, simplify the process of creating a one-top model. However, quantifying a one-top model has challenges, because each hazard model is built with different assumptions, data, biases, and uncertainty. Furthermore, when one hazard generates a risk value much larger than the risk value from another hazard, combining the results runs the risk of masking risk insights from the hazard with the smaller risk value. In addition to quantifying the one-top model for risk-informed applications that require or benefit from it, hazard models should still be quantified separately to get the risk insights the individual models provide.","PeriodicalId":360096,"journal":{"name":"2020 Annual Reliability and Maintainability Symposium (RAMS)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121660015","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 : 2020-01-01DOI: 10.1109/RAMS48030.2020.9153664
K. Gopikrishna, T. D. Gunneswara Rao, C. Putcha
The location of critical section in any civil infrastructure viz., bridges, trusses etc., that experience maximum value of design parameters for the subjected moving loads is imperative in order to design a safe structure. Traditionally these locations are evaluated by constructing the influence line diagrams (ILD) for the structures subjected to moving loads. In general the variations of these critical parameters viz., deflections, bending moment, shear force etc are evaluated at different locations while designing the configurations of the members of bridges, trusses etc. The location of critical parameters for instance the location of absolute maximum bending moment in a structure, may not coincide with the location of maximum deflection for a given position of load. Hence, in this paper an alternative probabilistic approach based on first order reliability principles is proposed for evaluating the location of critical section for the bridge structures subjected to moving loads, subsequently critical parameters can be evaluated at that location. In order to illustrate the approach, the following cases are considered
{"title":"Comparison of Critical Bending Moment in a Bridge Based on Influence Line Diagram (ILD), FOSM Method and Optimization","authors":"K. Gopikrishna, T. D. Gunneswara Rao, C. Putcha","doi":"10.1109/RAMS48030.2020.9153664","DOIUrl":"https://doi.org/10.1109/RAMS48030.2020.9153664","url":null,"abstract":"The location of critical section in any civil infrastructure viz., bridges, trusses etc., that experience maximum value of design parameters for the subjected moving loads is imperative in order to design a safe structure. Traditionally these locations are evaluated by constructing the influence line diagrams (ILD) for the structures subjected to moving loads. In general the variations of these critical parameters viz., deflections, bending moment, shear force etc are evaluated at different locations while designing the configurations of the members of bridges, trusses etc. The location of critical parameters for instance the location of absolute maximum bending moment in a structure, may not coincide with the location of maximum deflection for a given position of load. Hence, in this paper an alternative probabilistic approach based on first order reliability principles is proposed for evaluating the location of critical section for the bridge structures subjected to moving loads, subsequently critical parameters can be evaluated at that location. In order to illustrate the approach, the following cases are considered","PeriodicalId":360096,"journal":{"name":"2020 Annual Reliability and Maintainability Symposium (RAMS)","volume":"82 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122184810","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 : 2020-01-01DOI: 10.1109/RAMS48030.2020.9153710
Marie Ireland, J. Gonzales
Mission reliability for launch vehicles is the probability of successfully placing a payload into its delivery orbit within the required accuracy constraints accounting for design and process reliability. Launch vehicle design reliability provides an estimate of reliability by accounting for potential failure modes that originate within the system hardware and software while process reliability includes consideration of failure modes introduced by manufacturing, infrastructure, assembly, ground processing, and system integration activities. In order to account for the failure modes described above, the design reliability predictions are updated with historical flight successes and failures from similar launch vehicles to get the overall mission reliability.
{"title":"Bayesian Reliability Analysis with Slice Sampling in Launch Vehicle Applications","authors":"Marie Ireland, J. Gonzales","doi":"10.1109/RAMS48030.2020.9153710","DOIUrl":"https://doi.org/10.1109/RAMS48030.2020.9153710","url":null,"abstract":"Mission reliability for launch vehicles is the probability of successfully placing a payload into its delivery orbit within the required accuracy constraints accounting for design and process reliability. Launch vehicle design reliability provides an estimate of reliability by accounting for potential failure modes that originate within the system hardware and software while process reliability includes consideration of failure modes introduced by manufacturing, infrastructure, assembly, ground processing, and system integration activities. In order to account for the failure modes described above, the design reliability predictions are updated with historical flight successes and failures from similar launch vehicles to get the overall mission reliability.","PeriodicalId":360096,"journal":{"name":"2020 Annual Reliability and Maintainability Symposium (RAMS)","volume":"258 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122665424","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 : 2020-01-01DOI: 10.1109/RAMS48030.2020.9153583
Justin Brown, Ian Campbell
Summary and ConclusionsThermal Environmental Stress Screening (ESS) is a proven method used to detect manufacturing defects in production hardware. Numerous multi-hour cycles are performed to properly screen systems with thousands of solder connections, complex mechanical configurations, and intricate electrical designs. The current industry standard thermal ESS process is to perform a survey on the system, define the profile, and establish a set quantity of cycles to perform per system. It is also well-understood that machine learning has the capability to improve manufacturing processes [1]. In an effort to reduce test times and unnecessary stress, a Machine Learning (ML) model, based on the amount of production rework performed on the system prior to ESS, can be generated to predict the optimal amount of cycles to perform on the system. This approach improves both cost and schedule of the system under test.
{"title":"Dynamic Environmental Stress Screening Using Machine Learning","authors":"Justin Brown, Ian Campbell","doi":"10.1109/RAMS48030.2020.9153583","DOIUrl":"https://doi.org/10.1109/RAMS48030.2020.9153583","url":null,"abstract":"Summary and ConclusionsThermal Environmental Stress Screening (ESS) is a proven method used to detect manufacturing defects in production hardware. Numerous multi-hour cycles are performed to properly screen systems with thousands of solder connections, complex mechanical configurations, and intricate electrical designs. The current industry standard thermal ESS process is to perform a survey on the system, define the profile, and establish a set quantity of cycles to perform per system. It is also well-understood that machine learning has the capability to improve manufacturing processes [1]. In an effort to reduce test times and unnecessary stress, a Machine Learning (ML) model, based on the amount of production rework performed on the system prior to ESS, can be generated to predict the optimal amount of cycles to perform on the system. This approach improves both cost and schedule of the system under test.","PeriodicalId":360096,"journal":{"name":"2020 Annual Reliability and Maintainability Symposium (RAMS)","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122720566","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 : 2020-01-01DOI: 10.1109/RAMS48030.2020.9153590
Zhenggeng Ye, Zhiqiang Cai, F. Zhou, P. Zhang
The machine fault prognosis method by monitoring the manufacturing system dynamical performance has been widely studied recently, which is also the most common and significant problem faced in manufacturing industries. In a manufacturing system, machine is the key component. The dynamic and precise identification of the healthy state of the machine can support the decision making of production operation. In this paper, since the propagation of unqualified products will lead to the deterioration of machine’s processing accuracy, quality of imported products is considered to be an important factor affecting machine’s performance. Considering this practical scenario, a non-homogeneous Poisson process is applied to model the number of quality failures in a manufacturing system, and the log-normal distribution is used to depict the impact strength of unqualified products to a machine. At last, the applicability of the proposed model is discussed for the serial manufacturing system, and an analysis procedure of machine’s accuracy degradation is provided to illustrate its actionability.
{"title":"Degradation Analysis of Machine Processing Accuracy for Manufacturing Systems with Effect of Unqualified Products","authors":"Zhenggeng Ye, Zhiqiang Cai, F. Zhou, P. Zhang","doi":"10.1109/RAMS48030.2020.9153590","DOIUrl":"https://doi.org/10.1109/RAMS48030.2020.9153590","url":null,"abstract":"The machine fault prognosis method by monitoring the manufacturing system dynamical performance has been widely studied recently, which is also the most common and significant problem faced in manufacturing industries. In a manufacturing system, machine is the key component. The dynamic and precise identification of the healthy state of the machine can support the decision making of production operation. In this paper, since the propagation of unqualified products will lead to the deterioration of machine’s processing accuracy, quality of imported products is considered to be an important factor affecting machine’s performance. Considering this practical scenario, a non-homogeneous Poisson process is applied to model the number of quality failures in a manufacturing system, and the log-normal distribution is used to depict the impact strength of unqualified products to a machine. At last, the applicability of the proposed model is discussed for the serial manufacturing system, and an analysis procedure of machine’s accuracy degradation is provided to illustrate its actionability.","PeriodicalId":360096,"journal":{"name":"2020 Annual Reliability and Maintainability Symposium (RAMS)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131234897","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 : 2020-01-01DOI: 10.1109/rams48030.2020.9153695
A. Bahret
SUMMARY & CONCLUSIONSCharacterizing use and environment is critical to setting and satisfying reliability goals. The incorrect assessment and definitions of use case and environment is one of the most common reasons reliability targets are not met in the customer’s hands. This paper will discuss a technique for improving the process of characterizing how products are used for the purpose of deriving use cases and environmental conditions.
{"title":"Use Case7, Identifying the Boundaries of Use and Environment in Your Product","authors":"A. Bahret","doi":"10.1109/rams48030.2020.9153695","DOIUrl":"https://doi.org/10.1109/rams48030.2020.9153695","url":null,"abstract":"SUMMARY & CONCLUSIONSCharacterizing use and environment is critical to setting and satisfying reliability goals. The incorrect assessment and definitions of use case and environment is one of the most common reasons reliability targets are not met in the customer’s hands. This paper will discuss a technique for improving the process of characterizing how products are used for the purpose of deriving use cases and environmental conditions.","PeriodicalId":360096,"journal":{"name":"2020 Annual Reliability and Maintainability Symposium (RAMS)","volume":"273 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122927723","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：