为了解决局部阴影下传统最大功率点追踪(maximum power point tracking, MPPT)算法容易陷入局部最优从而降低光伏系统发电效率的问题,本研究提出融合正弦余弦算法和自适应策略的布谷鸟优化算法(cuckoo search algorithm fusing sine cos...为了解决局部阴影下传统最大功率点追踪(maximum power point tracking, MPPT)算法容易陷入局部最优从而降低光伏系统发电效率的问题,本研究提出融合正弦余弦算法和自适应策略的布谷鸟优化算法(cuckoo search algorithm fusing sine cosine algorithm and adaptive strategy, AFCS),并应用于光伏全局MPPT控制中,以改善其收敛速度与追踪精度.设置多种光照情况,并与扰动观察法、花朵授粉算法和粒子群算法进行对比.经过Matlab/Simulink仿真验证,表明本算法拥有较快的收敛速度和较高的追踪精度,在各个光照条件下均能快速追踪到光伏阵列最大功率点,可以有效提高光伏系统的发电效率.展开更多
针对目标检测算法在交通标志检测中存在的不足,文中提出了一种融合感受野增强模块和注意力机制的交通标志检测算法。该算法在YOLOv5(You Only Look Once version 5)算法的基础上改进,选用感受野模块(Receptive Field Block,RFB)替换原...针对目标检测算法在交通标志检测中存在的不足,文中提出了一种融合感受野增强模块和注意力机制的交通标志检测算法。该算法在YOLOv5(You Only Look Once version 5)算法的基础上改进,选用感受野模块(Receptive Field Block,RFB)替换原骨干网络中的空间金字塔池化(Spatial Pyramid Pooling,SPP)模块,在特征融合网络中嵌入高效通道注意模块(Efficient Channel Attention Module,ECAM)和卷积块注意模块(Convolutional Block Attention Module,CBAM),选用矩阵非极大值抑制(Matrix Non-Maximum Suppression,Matrix NMS)筛选候选框以提升算法的检测精度和检测速度。实验结果表明,在模型参数量与原网络相比未变化的前提下,该算法的均值平均精度达到了82.31%,与原算法相比提升了8.59%,检测速度达到了51.89 frame·s^(-1),且该算法在各个测试场景中未出现错检漏检现象,证明其泛化能力优于原算法,可以实时检测交通标志。展开更多
In this paper, we exploit the concept of data fusion in hybrid localization systems by combining different TOA (Time of Arrival) observables coming from different RATs (Radio Access Technology) and characterized by di...In this paper, we exploit the concept of data fusion in hybrid localization systems by combining different TOA (Time of Arrival) observables coming from different RATs (Radio Access Technology) and characterized by different precisions in order to enhance the positioning accuracy. A new Maximum Likelihood estimator is developed to fuse different measured ranges with different variances. In order to evaluate this estimator, Monte Carlo simulations are carried out in a generic environment and Cramer Rao Lower Bounds (CRLB) are investigated. This algorithm shows enhanced positioning accuracy at reasonable noise levels comparing to the typical Weighted Least Square estimator. The CRLB reveals that the choice of the number, and the configuration of Anchor nodes, and the type of RAT may enhance positioning accuracy.展开更多
In inertial confinement fusion （ICF）, X-ray coded imaging is considered as the most potential means to diagnose the compressed core. The traditional Richardson-Lucy （RL） method has a strong ability to deblur the i...In inertial confinement fusion （ICF）, X-ray coded imaging is considered as the most potential means to diagnose the compressed core. The traditional Richardson-Lucy （RL） method has a strong ability to deblur the image where the noise follows the Poisson distribution. However, it always suffers from over-fitting and noise amplification, especially when the signal-to-noise ratio of image is relatively low. In this paper, we propose an improved deconvolution method for X-ray coded imaging. We model the image data as a set of independent Gaussian distributions and derive the iterative solution with a maximum-likelihood scheme. The experimental results on X-ray coded imaging data demonstrate that this method is superior to the RL method in terms of anti-overfitting and noise suppression.展开更多
针对人体关键点检测存在检测精确度低的不足,在KAPAO(keypoints and pose as objects)网络的基础上进行改进。使用PoseTrans(pose transformation)进行数据增强,提高网络的泛化性;针对特征融合能力的不足,设计融合注意力机制的BiFPN(Bi-...针对人体关键点检测存在检测精确度低的不足,在KAPAO(keypoints and pose as objects)网络的基础上进行改进。使用PoseTrans(pose transformation)进行数据增强,提高网络的泛化性;针对特征融合能力的不足,设计融合注意力机制的BiFPN(Bi-directional feature network)模块充分融合不同语义特征,提高网络对深层语义信息和浅层语义信息的融合能力;在网络输出阶段设计自适应扩张卷积模块,将不同扩张率的输出分支进行自适应融合,有效获得图像的全局信息;在网络的后处理部分设计SDR-NMS(soft DIOU relocation non-maximum suppression)替代传统的NMS,保留最优的关键点预测框。实验结果表明,网络的AP分数提高了4.8%,AP为68.6%,检测速度为19.1 ms。网络精确度和检测速度均具有较好的表现性。展开更多
文摘为了解决局部阴影下传统最大功率点追踪(maximum power point tracking, MPPT)算法容易陷入局部最优从而降低光伏系统发电效率的问题,本研究提出融合正弦余弦算法和自适应策略的布谷鸟优化算法(cuckoo search algorithm fusing sine cosine algorithm and adaptive strategy, AFCS),并应用于光伏全局MPPT控制中,以改善其收敛速度与追踪精度.设置多种光照情况,并与扰动观察法、花朵授粉算法和粒子群算法进行对比.经过Matlab/Simulink仿真验证,表明本算法拥有较快的收敛速度和较高的追踪精度,在各个光照条件下均能快速追踪到光伏阵列最大功率点,可以有效提高光伏系统的发电效率.
文摘针对目标检测算法在交通标志检测中存在的不足,文中提出了一种融合感受野增强模块和注意力机制的交通标志检测算法。该算法在YOLOv5(You Only Look Once version 5)算法的基础上改进,选用感受野模块(Receptive Field Block,RFB)替换原骨干网络中的空间金字塔池化(Spatial Pyramid Pooling,SPP)模块,在特征融合网络中嵌入高效通道注意模块(Efficient Channel Attention Module,ECAM)和卷积块注意模块(Convolutional Block Attention Module,CBAM),选用矩阵非极大值抑制(Matrix Non-Maximum Suppression,Matrix NMS)筛选候选框以提升算法的检测精度和检测速度。实验结果表明,在模型参数量与原网络相比未变化的前提下,该算法的均值平均精度达到了82.31%,与原算法相比提升了8.59%,检测速度达到了51.89 frame·s^(-1),且该算法在各个测试场景中未出现错检漏检现象,证明其泛化能力优于原算法,可以实时检测交通标志。
文摘In this paper, we exploit the concept of data fusion in hybrid localization systems by combining different TOA (Time of Arrival) observables coming from different RATs (Radio Access Technology) and characterized by different precisions in order to enhance the positioning accuracy. A new Maximum Likelihood estimator is developed to fuse different measured ranges with different variances. In order to evaluate this estimator, Monte Carlo simulations are carried out in a generic environment and Cramer Rao Lower Bounds (CRLB) are investigated. This algorithm shows enhanced positioning accuracy at reasonable noise levels comparing to the typical Weighted Least Square estimator. The CRLB reveals that the choice of the number, and the configuration of Anchor nodes, and the type of RAT may enhance positioning accuracy.
基金Project supported by the National High-Tech ICF Committee of China,Foundation of China Academy of Engineering Physics(Grant Nos.2009A0102003 and 2011B0102021)the National Natural Science Foundation of China(Grant No.10905051)
文摘In inertial confinement fusion （ICF）, X-ray coded imaging is considered as the most potential means to diagnose the compressed core. The traditional Richardson-Lucy （RL） method has a strong ability to deblur the image where the noise follows the Poisson distribution. However, it always suffers from over-fitting and noise amplification, especially when the signal-to-noise ratio of image is relatively low. In this paper, we propose an improved deconvolution method for X-ray coded imaging. We model the image data as a set of independent Gaussian distributions and derive the iterative solution with a maximum-likelihood scheme. The experimental results on X-ray coded imaging data demonstrate that this method is superior to the RL method in terms of anti-overfitting and noise suppression.
