基于EOF-DBSCAN-GRU的分布式光伏集群出力预测方法研究

提出一种基于EOF-DBSCAN划分集群的改进统计升尺度的光伏集群出力预测建模方法。针对传统统计升尺度方法子集群中光伏电站出力特性不一致问题,通过皮尔逊相关系数和经验正交函数(EOF)优化特征空间,再根据密度聚类模型(DBSCAN)对区域内光伏电站划分集群,从而增强光伏电站聚类后集群出力特性的一致性。针对待预测日权重系数时间序列动态特性的提取、预测问题,提出一种基于动态时间规整(DTW)的相似日选取算法。最后利用循环神经网络(GRU)模型进行光伏电站出力预测。实验表明该集群预测方法的平均误差百分数(MAPE)约为6.33%,均方根误差(RMSE)约为13.93 kW,均方误差(MSE)为194.25 kW,通过实际光伏电站数据证明了所提方法的准确性和有效性。

Spatial Distribution Pattern and Influencing Factors of Bed-and-breakfasts(B&Bs)from the Perspective of Urban-rural Differences:A Case Study of Jiaodong Peninsula,China

There are significant differences between urban and rural bed-and-breakfasts(B&Bs)in terms of customer positioning,economic strength and spatial carrier.Accurately identifying the differences in spatial characteristics and influencing factors of each type,is essential for creating urban and rural B&B agglomeration areas.This study used density-based spatial clustering of applications with noise(DBSCAN)and the multi-scale geographically weighted regression(MGWR)model to explore similarities and differences in the spatial distribution patterns and influencing factors for urban and rural B&Bs on the Jiaodong Peninsula of China from 2010 to 2022.The results showed that:1)both urban and rural B&Bs in Jiaodong Peninsula went through three stages:a slow start from 2010 to 2015,rapid development from 2015 to 2019,and hindered development from 2019 to 2022.However,urban B&Bs demonstrated a higher development speed and agglomeration intensity,leading to an increasingly evident trend of uneven development between the two sectors.2)The clustering scale of both urban and rural B&Bs continued to expand in terms of quantity and volume.Urban B&B clusters characterized by a limited number,but a higher likelihood of transitioning from low-level to high-level clusters.While the number of rural B&B clusters steadily increased over time,their clustering scale was comparatively lower than that of urban B&Bs,and they lacked the presence of high-level clustering.3)In terms of development direction,urban B&B clusters exhibited a relatively stable pattern and evolved into high-level clustering centers within the main urban areas.Conversely,rural B&Bs exhibited a more pronounced spatial diffusion effect,with clusters showing a trend of multi-center development along the coastline.4)Transport emerged as a common influencing factor for both urban and rural B&Bs,with the density of road network having the strongest explanatory power for their spatial distribution.In terms of differences,population agglomeration had a positive impact on the distribution of urban B&Bs and a negative effect on the distribution of rural B&Bs.Rural B&Bs clustering was more influenced by tourism resources compared with urban B&Bs,but increasing tourist stay duration remains an urgent issue to be addressed.The findings of this study could provide a more precise basis for government planning and management of urban and rural B&B agglomeration areas.

内河航道事故黑点识别自适应参数DBSCAN聚类算法研究

内河水上交通事故时有发生,对水路运输安全、高效发展带来威胁。研究提出一种基于自适应参数的DBSCAN(Density-Based Spatial Clustering of Applications with Noise)方法,用于识别内河事故黑点水域。该方法支持对邻域半径ε和邻域中数据对象数目阈值P_(min)参数的自动选取,可提高聚类分析的精度和效率。基于2010—2019年长江干线下游散货船舶事故数据开展案例研究,对各典型事故黑点段的事故特征和事故原因进行分析,得到8个事故黑点。此外,采用Getis-Ord General G聚类识别事故黑点中的高等级事故区域,得到事故黑点及高等级事故主要分布于江心洲、桥区、港口码头区域。研究结果与实际情况基本吻合,一定程度上表明了该方法在内河水上交通事故分布特征分析上的科学性和实用性。

基于改进DBSCAN的船舶会遇识别模型

为解决大数据下船舶会遇识别算法效率不高且存在误判等问题,提出一种融合国际海上避碰规则(International Regulations for Preventing Collisions at Sea,COLREGs)的带噪声的基于密度的空间聚类(density-based spatial clustering of applications with noise,DBSCAN)算法,建立船舶会遇识别模型。在DBSCAN算法对邻域内的船舶数量进行统计时,计算船舶间的最近会遇距离(distance to closest point of approach,DCPA)和最近会遇时间(time to closest point of approach,TCPA),初步筛选邻域内的噪声点;基于模糊综合评价模型计算船舶会遇风险,对邻域内的船舶进行二次筛选,实现船舶会遇态势的提取。结果表明:改进后的DBSCAN算法过滤掉传统DBSCAN算法识别到的非会遇局面,并且在同一会遇局面下的船舶数量均保持在4艘以内;输出的会遇船舶风险演变趋势对实际水域内高风险船舶的监控适用性较好,能有效辅助船舶避碰。所提识别模型对保障航行安全和提高海事监管效率具有重要意义。

Spatial Distribution Pattern and Influencing Factors of Physical Bookstores of Large Cities:A Case Study of Three National Central Cities in Western China

As cultural facilities,physical bookstore is an important part of urban infrastructure.Influenced by the development of social economy and the internet,physical bookstores also have become a combination of cultural space and tourism experience.In this case,it is necessary to explore the spatial characteristics and influencing factors of physical bookstores.This study uses Density-Based Spatial Clustering of Applications with Noise(DBSCAN),spatial analysis and geographical detectors to calculate the spatial distribution pattern and factors influencing physical bookstores in national central cities/municipality(hereafter using cities)in western China.Based on spatial data,population density,road density and other data,this study constructed a data set of the influencing factors of physical bookstores,consisting of 11 factors along 6 dimensions for 3 national central cities in western China.The results are as follows:first,the spatial distribution pattern of physical bookstores in Xi’an,Chengdu,and Chongqing is unbalanced.The spatial distribution of physical bookstores in Xi’an and Chongqing is from southwest to northeast and are relatively clustered,while those in Chengdu are relatively discrete.Second,the spatial distribution pattern of physical bookstores has been formed under the influence of different factors.The intensity and significance of influencing factors differ in the case cities.However,in general,the social factor,business factor,the density of research facilities,tourism factor and road density are the main driving factors in the three cities.There is a synergistic relationship between public libraries and physical bookstores.Third,the explanatory power becomes stronger after the interaction between various factors.In Xi’an and Chengdu,the density of communities and the density of research facilities have stronger explanatory power for the dependent variable after interacting with other factors.However,in Chongqing,the traffic factors have stronger explanatory power for the dependent variable after interacting with other factors.The results could provide a practical reference for the sustainable development of physical bookstores and encourage a love of reading among the public.

基于改进DBSCAN和距离共识评估的分段点云去噪方法

针对点云数据中噪声点的剔除问题,提出了一种基于改进DBSCAN(density-based spatial clustering of applications with noise)算法的多尺度点云去噪方法。应用统计滤波对孤立离群点进行预筛选,去除点云中的大尺度噪声;对DBSCAN算法进行优化,减少算法时间复杂度和实现参数的自适应调整,以此将点云分为正常簇、疑似簇及异常簇,并立即去除异常簇;利用距离共识评估法对疑似簇进行精细判定,通过计算疑似点与其最近的正常点拟合表面之间的距离,判定其是否为异常,有效保持了数据的关键特征和模型敏感度。利用该方法对两个船体分段点云进行去噪,并与其他去噪算法进行对比,结果表明,该方法在去噪效率和特征保持方面具有优势,精确地保留了点云数据的几何特性。

基于DBSCAN算法的树木分割与应用

为快速准确地提取地面三维激光扫描仪获取林分点云中的单株树木点云,提出一种基于密度的抗噪空间聚类(Density-Based Spatial Clustering of Application with Noise,DBSCAN)的树木分割算法。首先采用高斯滤波对林分点云去噪,在林分点云归一化的基础上对林分点云垂直分段,然后采用DBSCAN算法垂直分段聚类,再计算每个垂直分段中每个簇的中心点,根据簇中心点间的距离判定簇间的相邻关系,并由此匹配树干段点云,最后采用RANSAC(Random Sample Consensus)算法对树干段点云拟合直线,并根据点与拟合直线间的距离判定点的归属以实现树木分割。在郁闭度分别为中与高的林分中,所提算法的调和值F范围分别为0.88~0.99与0.72~0.74,基于距离判别的树木分割算法的F范围分别为0.84~0.90与0.73~0.79。所提算法在不同郁闭度的林分点云中均能有效分割单株树木点云,特别是在郁闭度为中的林分中有较好表现,可实现对林分点云的精确树木分割。

Three-dimensional(3D)parametric measurements of individual gravels in the Gobi region using point cloud technique

Gobi spans a large area of China,surpassing the combined expanse of mobile dunes and semi-fixed dunes.Its presence significantly influences the movement of sand and dust.However,the complex origins and diverse materials constituting the Gobi result in notable differences in saltation processes across various Gobi surfaces.It is challenging to describe these processes according to a uniform morphology.Therefore,it becomes imperative to articulate surface characteristics through parameters such as the three-dimensional(3D)size and shape of gravel.Collecting morphology information for Gobi gravels is essential for studying its genesis and sand saltation.To enhance the efficiency and information yield of gravel parameter measurements,this study conducted field experiments in the Gobi region across Dunhuang City,Guazhou County,and Yumen City(administrated by Jiuquan City),Gansu Province,China in March 2023.A research framework and methodology for measuring 3D parameters of gravel using point cloud were developed,alongside improved calculation formulas for 3D parameters including gravel grain size,volume,flatness,roundness,sphericity,and equivalent grain size.Leveraging multi-view geometry technology for 3D reconstruction allowed for establishing an optimal data acquisition scheme characterized by high point cloud reconstruction efficiency and clear quality.Additionally,the proposed methodology incorporated point cloud clustering,segmentation,and filtering techniques to isolate individual gravel point clouds.Advanced point cloud algorithms,including the Oriented Bounding Box(OBB),point cloud slicing method,and point cloud triangulation,were then deployed to calculate the 3D parameters of individual gravels.These systematic processes allow precise and detailed characterization of individual gravels.For gravel grain size and volume,the correlation coefficients between point cloud and manual measurements all exceeded 0.9000,confirming the feasibility of the proposed methodology for measuring 3D parameters of individual gravels.The proposed workflow yields accurate calculations of relevant parameters for Gobi gravels,providing essential data support for subsequent studies on Gobi environments.

基于DBSCAN聚类算法的毕星团成员星辨认

疏散星团[1]在天文学和天文物理学的研究中至关重要,对其成员星的正确判断是所有研究工作的基础。成员星识别的可靠性随着天体测量技术的不断发展,而变得愈发重要。本文运用DBSCAN聚类算法,实现毕星团成员星认定问题的分析与研究。通过色指数与温度公式求出各个侯选星成员星的温度。以绝对星等为纵坐标,将其成员星的赫罗图作为恒星表面温度的横坐标,对筛选出的毕星团成员星体绘制了一幅赫罗图。

基于DBSCAN-GRU算法的终端区4D航迹预测

终端区空域环境复杂、航班密集,精确的航迹预测能极大地提高空中交通服务水平,保障航班飞行安全。针对终端区的高精度多航班4D航迹预测问题,本文提出了一种基于密度的带噪声空间聚类算法(Density-Based Spatial Clustering of Applications with Noise,DBSCAN)和门控循环单元(Gated Recurrent Unit,GRU)相结合的航迹预测方法,通过DBSCAN聚类,将终端区中航迹相近的航班聚类到一簇中,对每一簇航班建立基于GRU神经网络的航迹预测模型,对终端区航班进行预测时,先判断该航班属于哪一簇,然后采用与该簇对应的航迹预测模型,进行4D航迹预测。与仅研究单一航班的传统预测方法相比,本算法有效地利用了终端区的航迹数据,所建模型可以针对多架航班进行航迹预测,扩大了模型的适用范围,提高了航迹预测的预测精度。

K-means Find Density Peaks in Molecular Conformation Clustering

Performing cluster analysis on molecular conformation is an important way to find the representative conformation in the molecular dynamics trajectories.Usually,it is a critical step for interpreting complex conformational changes or interaction mechanisms.As one of the density-based clustering algorithms,find density peaks(FDP)is an accurate and reasonable candidate for the molecular conformation clustering.However,facing the rapidly increasing simulation length due to the increase in computing power,the low computing efficiency of FDP limits its application potential.Here we propose a marginal extension to FDP named K-means find density peaks(KFDP)to solve the mass source consuming problem.In KFDP,the points are initially clustered by a high efficiency clustering algorithm,such as K-means.Cluster centers are defined as typical points with a weight which represents the cluster size.Then,the weighted typical points are clustered again by FDP,and then are refined as core,boundary,and redefined halo points.In this way,KFDP has comparable accuracy as FDP but its computational complexity is reduced from O(n^(2))to O(n).We apply and test our KFDP method to the trajectory data of multiple small proteins in terms of torsion angle,secondary structure or contact map.The comparing results with K-means and density-based spatial clustering of applications with noise show the validation of the proposed KFDP.

基于RAPIDS的无参DBSCAN算法

具有噪声的基于密度的空间聚类(Density‑based spatial clustering of applications with noise,DBSCAN)能够发现不同密度和大小的类簇,对噪声也有很好的鲁棒性,被广泛地应用到数据挖掘的任务中。DBSCAN通常需要调整参数MinPts和Eps以达到更优的聚类效果,但往往在搜索最优参数的过程中会影响DBSCAN的性能。本文从两个方面优化DBSCAN,一方面,提出一种无参的方法优化DBSCAN全局参数选择。无参方法利用自然最近邻获得数据集的自然特征值,并将自然特征值作为参数MinPts值。然后,根据自然特征值计算自然特征集合,利用自然特征集合中的数据分布特性,分别采取统计最小值、平均值和最大值3种方式得到Eps值。另一方面,采用集成数据科学实时加速平台(Real‑time acceleration platform for integrated data science,RAPIDS)的图形处理器(Graphics processing unit,GPU)计算加快DBSCAN算法的收敛速度。实验结果表明,本文提出的方法在优化DBSCAN参数选择的同时,取得了与密度峰值聚类(Density peaks clustering,DPC)相当的聚类结果。

A heuristic clustering algorithm based on high density-connected partitions

Clustering data with varying densities and complicated structures is important,while many existing clustering algorithms face difficulties for this problem. The reason is that varying densities and complicated structure make single algorithms perform badly for different parts of data. More intensive parts are assumed to have more information probably,an algorithm clustering from high density part is proposed,which begins from a tiny distance to find the highest density-connected partition and form corresponding super cores,then distance is iteratively increased by a global heuristic method to cluster parts with different densities. Mean of silhouette coefficient indicates the cluster performance. Denoising function is implemented to eliminate influence of noise and outliers. Many challenging experiments indicate that the algorithm has good performance on data with widely varying densities and extremely complex structures. It decides the optimal number of clusters automatically.Background knowledge is not needed and parameters tuning is easy. It is robust against noise and outliers.

利用DBSCAN和概率密度估计的欠定盲源分离混合矩阵估计

针对欠定盲源分离中混合矩阵估计精度不佳的问题,本文提出了一种结合带噪声的基于密度的空间聚类(combining density-based spatial clustering of application with noise,DBSCAN)和概率密度估计的混合矩阵估计算法。首先,通过向量转换方式获得单声源时频点检测准则,并基于此准则从混合信号中检测出单声源点。其次,利用基于密度的空间聚类算法对单声源点进行聚类,由此估计出声源个数以及各类别所属的单声源点。再次,利用概率密度估计获得各类别的聚类中心,并构成混合矩阵。所提混合矩阵估计方法不需要提前设定声源个数,并且避免了由于数据分布不均所造成的聚类效果差的问题。最后,采用压缩感知技术实现源信号恢复,从而从混合信号中分离出各个声源信号。实验结果表明,本文所提的混合矩阵估计方法在声源个数未知的情况下,能够准确估计出混合矩阵;并且分离出的信号具有较高的质量。

由VMD与DBSCAN在线检测锂离子电池热失控

热失控影响锂离子电池系统的推广和应用。为预测锂离子电池系统的热失控,提出基于变分模态分解(VMD)与密度的噪声空间聚类(DBSCAN)算法的热失控在线检测方法。针对实际热失控案例,结合VMD与滑动窗口,在线分解窗口内各电池的电压信号,得到电压稳态分量;之后,提取各电池稳态分量与稳态分量均值的标准化皮尔逊相关系数,以及余弦相似度,并构建二维特征矩阵;最后,由DBSCAN自动辨识电池组中的故障电池,最早可在热失控前111 s检测出电压异常。

基于DBSCAN聚类的热能发电大数据异常检测模型

为了解决热能发电大数据异常检测时,存在误报率高、检测率低和漏报率高的问题,提出了基于DBSCAN聚类的热能发电大数据异常检测模型。首先通过时间序列模型对原始发电大数据进行异常值修正,然后将修正后的数据归一化处理,最后基于Spark Streaming设计Streaming DBSCAN算法,结合历史数据和相似发电厂数据的聚类特征,完成热能发电异常数据的检测。实验结果表明,所提方法可以有效降低误报率和漏报率、提高检测率并准确地获取异常值。

改进DBSCAN算法下的轨迹点到充电站位置的探测方法

在当前新能源汽车快速发展的背景下,针对相关充电设施位置信息更新缓慢等问题,如何通过第三方数据得到工作状态正常的充电站点分布具有重要意义。本文分析了新能源汽车充电停留轨迹数据的典型特征,并基于这些特征构建了时空关联静动(Stop/Move)模型。利用新能源汽车轨迹数据作为数据源,采用具有噪声的基于密度的聚类(DBSCAN)算法来检测满足充电停留点的点簇,并进一步探测充电站的位置。同时,针对DBSCAN算法具有高时间复杂度的问题,通过构建K维空间树(KD树)数据结构提高了算法执行效率;针对不同参数会影响DBSCAN算法聚类结果的问题,运用邻域参数自适应优化方法提升了轨迹点的聚类效果。利用深圳市的新能源车轨迹数据进行实验分析,结果表明,相比原始DBSCAN算法和k均值聚类(K-MEANS)算法,改进DBSCAN算法提高了算法执行效率,对真实充电站点探测成功率较高。

融合DBSCAN多测点算法的磁方位校正

单测点校正法计算复杂、不稳定、误差较大,无法满足井下地磁方位角的精度要求。基于间接单测点分析法和Brooks多测点分析法,提出一种新方法,通过DBSCAN(Density-Based Spatial Clustering of Applications with Noise)算法识别并剔除噪声,采用椭圆校正法校正径向干扰,轴向干扰则由单轴多测点分析法校正。实验证明:改进多测点法不仅可以进一步提高椭圆校正的拟合效果,还能降低噪声对参考点计算值的影响,计算得到的方差曲线收敛性更强、更稳定,校正后方位角误差进一步降低。

基于DBSCAN算法的海战场环境下目标分群

针对海战场环境下态势评估中目标数量多、类型复杂多样的问题,首先引入数据聚类对态势评估的目标分群环节进行聚类分群,提出基于DBSCAN(density-based spatial clustering of applications with noise)算法的密度聚类,可聚类任意形状的数据簇,遍历性好,能够对战场环境下目标进行全面合理的分群;然后,给出了算法计算的基本步骤,并利用算例对已知战场态势的目标群进行正确性验证;最后,将该算法与基于划分的K-means算法、基于层次的AGNES(AGglomerative NESting)算法进行了对比分析,证明了该算法的有效性和合理性。

基于混合优化算法的电磁监测裂缝参数识别

压裂实时监测是水力压裂效果评价和工程参数优化的重要保障措施之一。传统电磁监测裂缝参数识别方法的准确性和精度难以保证,影响了监测效果。为了提高裂缝参数识别能力,提出了一种基于海洋捕食者密度聚类混合优化算法的电磁监测裂缝参数识别方法。利用海洋捕食者算法(MPA)进行多次寻优,以每次寻优结果作为初始数据集,然后,利用密度聚类算法(DBSCAN)进行聚类,构建中间样本数据集,最后,抽取该样本数据中值作为最终输出结果。采用Rastrigin函数进行测试,分析混合优化算法寻优能力。测试结果表明,相对粒子群优化算法(PSO),MPA算法单次寻优效果较佳。但两种算法寻优结果均具有较强随机性,其中,PSO和MPA算法50次寻优精度分别为10^(-7)~10^(2)和10^(-10)~10^(-2),而改进的混合优化算法寻优效果更稳定,寻优精度达10^(-7)。构建缝长、方位压裂模型并进行了数值模拟实验,结果表明,在噪声低于15%时,缝长和方位识别平均绝对误差分别小于1 m和1°。利用改进的算法对四川盆地某井页岩气压裂电磁监测实测数据进行分析,确定了裂缝改造的长度(缝长)与方位。实例分析结果验证了改进算法的可行性和有效性。