Accurate test effectiveness estimation for analogue and mixed-signal Systems on a Chip (SoCs) is currently prohibitive in the design environment. One of the factors that sky rockets fault simulation costs is the numbe...Accurate test effectiveness estimation for analogue and mixed-signal Systems on a Chip (SoCs) is currently prohibitive in the design environment. One of the factors that sky rockets fault simulation costs is the number of structural faults which need to be simulated at circuit-level. The purpose of this paper is to propose a novel fault list compression technique by defining a stratified fault list, build with a set of “representative” faults, one per stratum. Criteria to partition the fault list in strata, and to identify representative faults are presented and discussed. A fault representativeness metric is proposed, based on an error probability. The proposed methodology allows different tradeoffs between fault list compression and fault representation accuracy. These tradeoffs may be optimized for each test preparation phase. The fault representativeness vs. fault list compression tradeoff is evaluated with an industrial case study—a DC-DC (switched buck converter). Although the methodology is presented in this paper using a very simple fault model, it may be easily extended to be used with more elaborate fault models. The proposed technique is a significant contribution to make mixed-signal fault simulation cost-effective as part of the production test preparation.展开更多
Objective: Individual differences in the sensitivity to pain and the factors that may contribute to these differences are well studied. Nevertheless, there is no single test that can reliably classify subjects as bein...Objective: Individual differences in the sensitivity to pain and the factors that may contribute to these differences are well studied. Nevertheless, there is no single test that can reliably classify subjects as being sensitive or insensitive to pain. Methods: In the present study, hierarchical clustering and K-means cluster analysis was used to identify subgroups among 191 healthy subjects (105 females, 86 males) according to their sensitivity to pain. Group determination was based on the subjects’ response to experimental noxious stimuli of heat (pain intensity), cold (cold pain threshold, tolerance, and intensity), and conditioned pain modulation (CPM, tested by co-administering repeated short painful heat stimuli and a conditioning tonic cold pain stimulation). In addition, in order to determine if the subjects in these subgroups differed on personality traits scores on Cloninger’s Tridimensional Personality Questionnaire (TPQ, outcome measure) for the three dimensions of personality: Novelty Seeking (NS);Harm Avoidance (HA);and Reward Dependence (RD) were calculated. Results: Based on pain scores, subjects were grouped as low pain (57%) with a low level of sensitivity in pain parameters, or high pain (43%) cluster members. The high pain had significant higher scores of HA (p = 0.05) and RD (p = 0.05) than the low pain group. Conclusions: This method of sub-grouping may be useful for identifying the mechanisms underlying individual variability in the sensitivity to pain and may point to groups at risk for experiencing high levels of clinical pain.展开更多
A design for an air-cooled test tube, with a series of modular and movable rings (cylindrical “crystallization comb”), installed in a laboratory crucible furnace is presented. The setup allows easy regulation simult...A design for an air-cooled test tube, with a series of modular and movable rings (cylindrical “crystallization comb”), installed in a laboratory crucible furnace is presented. The setup allows easy regulation simultaneous crystallization tests of a series of different crystallization rates in several columns (matrix) of test tubes, enabling fast studies of obtaining crystals. This low-budget, portable device (i.e. adjustable airstream with more simple control options), can also be applied in tube and chamber furnaces. The relations between the crystallization rate and parameters of air-cooled test tube are given and numerically analyzed.展开更多
为构建科学合理的城市道路自动驾驶测试场景,基于美国加利福尼亚州机动车管理局(California Department of Motor Vehicles, DMV)2021—2023年公开的280起自动驾驶汽车(Autonomous Vehicle, AV)碰撞事故报告,挖掘典型危险场景并完成测...为构建科学合理的城市道路自动驾驶测试场景,基于美国加利福尼亚州机动车管理局(California Department of Motor Vehicles, DMV)2021—2023年公开的280起自动驾驶汽车(Autonomous Vehicle, AV)碰撞事故报告,挖掘典型危险场景并完成测试场景转化。首先,通过多元Logistic回归分析提取人员受伤情况的显著影响因素。其次,引入独热编码(One-Hot Encoding)对分类变量进行二进制向量转换,消除传统标签编码的数值顺序偏差。然后,采用二阶聚类算法挖掘典型危险场景组,并进一步通过交叉表分析场景组与事故结果变量、道路环境变量间的关联性。最后,将危险场景转化设计为自动驾驶测试场景。结果显示,独热编码处理后的变量,聚类质量较传统方法提升50%;聚类分析共识别出12类典型危险场景,且交叉表分析表明场景组与事故结果及道路环境变量显著相关;进一步结合事故机理与测试需求,将这12类危险场景归纳为6类代表性测试场景,其中“AV停止或减速状态下被后方直行车辆追尾”的场景最为典型,在全部场景中占比46.1%。研究表明,独热编码方法显著提升了聚类分析的准确性,基于真实事故数据的场景聚类方法能识别AV在城市道路的事故模式,并为自动驾驶测试场景库的优先级划分与标准化设计提供数据驱动支撑。展开更多
文摘Accurate test effectiveness estimation for analogue and mixed-signal Systems on a Chip (SoCs) is currently prohibitive in the design environment. One of the factors that sky rockets fault simulation costs is the number of structural faults which need to be simulated at circuit-level. The purpose of this paper is to propose a novel fault list compression technique by defining a stratified fault list, build with a set of “representative” faults, one per stratum. Criteria to partition the fault list in strata, and to identify representative faults are presented and discussed. A fault representativeness metric is proposed, based on an error probability. The proposed methodology allows different tradeoffs between fault list compression and fault representation accuracy. These tradeoffs may be optimized for each test preparation phase. The fault representativeness vs. fault list compression tradeoff is evaluated with an industrial case study—a DC-DC (switched buck converter). Although the methodology is presented in this paper using a very simple fault model, it may be easily extended to be used with more elaborate fault models. The proposed technique is a significant contribution to make mixed-signal fault simulation cost-effective as part of the production test preparation.
文摘Objective: Individual differences in the sensitivity to pain and the factors that may contribute to these differences are well studied. Nevertheless, there is no single test that can reliably classify subjects as being sensitive or insensitive to pain. Methods: In the present study, hierarchical clustering and K-means cluster analysis was used to identify subgroups among 191 healthy subjects (105 females, 86 males) according to their sensitivity to pain. Group determination was based on the subjects’ response to experimental noxious stimuli of heat (pain intensity), cold (cold pain threshold, tolerance, and intensity), and conditioned pain modulation (CPM, tested by co-administering repeated short painful heat stimuli and a conditioning tonic cold pain stimulation). In addition, in order to determine if the subjects in these subgroups differed on personality traits scores on Cloninger’s Tridimensional Personality Questionnaire (TPQ, outcome measure) for the three dimensions of personality: Novelty Seeking (NS);Harm Avoidance (HA);and Reward Dependence (RD) were calculated. Results: Based on pain scores, subjects were grouped as low pain (57%) with a low level of sensitivity in pain parameters, or high pain (43%) cluster members. The high pain had significant higher scores of HA (p = 0.05) and RD (p = 0.05) than the low pain group. Conclusions: This method of sub-grouping may be useful for identifying the mechanisms underlying individual variability in the sensitivity to pain and may point to groups at risk for experiencing high levels of clinical pain.
文摘A design for an air-cooled test tube, with a series of modular and movable rings (cylindrical “crystallization comb”), installed in a laboratory crucible furnace is presented. The setup allows easy regulation simultaneous crystallization tests of a series of different crystallization rates in several columns (matrix) of test tubes, enabling fast studies of obtaining crystals. This low-budget, portable device (i.e. adjustable airstream with more simple control options), can also be applied in tube and chamber furnaces. The relations between the crystallization rate and parameters of air-cooled test tube are given and numerically analyzed.
文摘为构建科学合理的城市道路自动驾驶测试场景,基于美国加利福尼亚州机动车管理局(California Department of Motor Vehicles, DMV)2021—2023年公开的280起自动驾驶汽车(Autonomous Vehicle, AV)碰撞事故报告,挖掘典型危险场景并完成测试场景转化。首先,通过多元Logistic回归分析提取人员受伤情况的显著影响因素。其次,引入独热编码(One-Hot Encoding)对分类变量进行二进制向量转换,消除传统标签编码的数值顺序偏差。然后,采用二阶聚类算法挖掘典型危险场景组,并进一步通过交叉表分析场景组与事故结果变量、道路环境变量间的关联性。最后,将危险场景转化设计为自动驾驶测试场景。结果显示,独热编码处理后的变量,聚类质量较传统方法提升50%;聚类分析共识别出12类典型危险场景,且交叉表分析表明场景组与事故结果及道路环境变量显著相关;进一步结合事故机理与测试需求,将这12类危险场景归纳为6类代表性测试场景,其中“AV停止或减速状态下被后方直行车辆追尾”的场景最为典型,在全部场景中占比46.1%。研究表明,独热编码方法显著提升了聚类分析的准确性,基于真实事故数据的场景聚类方法能识别AV在城市道路的事故模式,并为自动驾驶测试场景库的优先级划分与标准化设计提供数据驱动支撑。
