To study the diagnostic problem of Wire-OR （W-O） interconnect fault of PCB （Printed Circuit Board）, five modified boundary scan adaptive algorithms for interconnect test are put forward. These algorithms apply Glo...To study the diagnostic problem of Wire-OR （W-O） interconnect fault of PCB （Printed Circuit Board）, five modified boundary scan adaptive algorithms for interconnect test are put forward. These algorithms apply Global-diagnosis sequence algorithm to replace the equal weight algorithm of primary test, and the test time is shortened without changing the fault diagnostic capability. The descriptions of five modified adaptive test algorithms are presented, and the capability comparison between the modified algorithm and the original algorithm is made to prove the validity of these algorithms.展开更多
针对基于深度强化学习的自主超声扫描方法存在训练扫描精度低、训练时间长、扫描任务成功率较低的问题,提出了一种基于改进型多模态信息融合深度强化学习的自主超声扫描方法。首先,该方法融合了超声图像、双视角探头操作图像和6D触觉反...针对基于深度强化学习的自主超声扫描方法存在训练扫描精度低、训练时间长、扫描任务成功率较低的问题,提出了一种基于改进型多模态信息融合深度强化学习的自主超声扫描方法。首先,该方法融合了超声图像、双视角探头操作图像和6D触觉反馈提供全面的多模态感知信息。为精准捕捉多模态中的时空信息和实现多模态特征的高效融合,设计了一个基于自注意力机制(self-attention mechanism,SA)的多模态特征提取与融合模块。其次,将机器人的6D位姿动作决策任务建模为深度强化学习问题。为贴近专业超声从业医生的操作,设计了混合奖励函数。最后,为解决深度强化学习训练中出现的局部最优和收敛速度慢的问题,提出了DSAC-PERDP(discrete soft actor-critic with prioritized experience replay based on dynamic priority)算法。在真实环境中的测试表明,该方法在扫描精度、任务成功率和训练速度方面较基线模型分别提升了49.8%、13.4%和260.0%,在干扰条件下仍保持良好性能。实验证明,该方法显著提升了扫描精度、任务成功率和训练速度,并具有一定的抗干扰能力。展开更多
This paper presents a new hybrid genetic algorithm for the vertex cover problems in which scan-repair and local improvement techniques are used for local optimization. With the hybrid approach, genetic algorithms are ...This paper presents a new hybrid genetic algorithm for the vertex cover problems in which scan-repair and local improvement techniques are used for local optimization. With the hybrid approach, genetic algorithms are used to perform global exploration in a population, while neighborhood search methods are used to perform local exploitation around the chromosomes. The experimental results indicate that hybrid genetic algorithms can obtain solutions of excellent quality to the problem instances with different sizes. The pure genetic algorithms are outperformed by the neighborhood search heuristics procedures combined with genetic algorithms.展开更多
文摘To study the diagnostic problem of Wire-OR （W-O） interconnect fault of PCB （Printed Circuit Board）, five modified boundary scan adaptive algorithms for interconnect test are put forward. These algorithms apply Global-diagnosis sequence algorithm to replace the equal weight algorithm of primary test, and the test time is shortened without changing the fault diagnostic capability. The descriptions of five modified adaptive test algorithms are presented, and the capability comparison between the modified algorithm and the original algorithm is made to prove the validity of these algorithms.
文摘针对基于深度强化学习的自主超声扫描方法存在训练扫描精度低、训练时间长、扫描任务成功率较低的问题,提出了一种基于改进型多模态信息融合深度强化学习的自主超声扫描方法。首先,该方法融合了超声图像、双视角探头操作图像和6D触觉反馈提供全面的多模态感知信息。为精准捕捉多模态中的时空信息和实现多模态特征的高效融合,设计了一个基于自注意力机制(self-attention mechanism,SA)的多模态特征提取与融合模块。其次,将机器人的6D位姿动作决策任务建模为深度强化学习问题。为贴近专业超声从业医生的操作,设计了混合奖励函数。最后,为解决深度强化学习训练中出现的局部最优和收敛速度慢的问题,提出了DSAC-PERDP(discrete soft actor-critic with prioritized experience replay based on dynamic priority)算法。在真实环境中的测试表明,该方法在扫描精度、任务成功率和训练速度方面较基线模型分别提升了49.8%、13.4%和260.0%,在干扰条件下仍保持良好性能。实验证明,该方法显著提升了扫描精度、任务成功率和训练速度,并具有一定的抗干扰能力。
基金This project was supported by the National Natural Science Foundation of China the Open Project Foundation of Comput-er Software New Technique National Key Laboratory of Nanjing University.
文摘This paper presents a new hybrid genetic algorithm for the vertex cover problems in which scan-repair and local improvement techniques are used for local optimization. With the hybrid approach, genetic algorithms are used to perform global exploration in a population, while neighborhood search methods are used to perform local exploitation around the chromosomes. The experimental results indicate that hybrid genetic algorithms can obtain solutions of excellent quality to the problem instances with different sizes. The pure genetic algorithms are outperformed by the neighborhood search heuristics procedures combined with genetic algorithms.
