高密度存量城区具有人口密集、功能叠合、空间受限等特点,其绿地建设受到诸多制约.如何解决居民绿地需求和绿地空间不足之间的矛盾,已成为高密度存量城区发展的关注要点.通过对高密度存量城区绿地规划实施困境进行分析,结合武汉市江汉区绿地规划实践,从系统性、精准度、综合性方面提出规划策略,并提出高密度存量城区绿地规划的相关思考与建议:(1)挖掘存量空间进行绿地建设;(2)充分发挥绿地的综合作用;(3)建立可持续的绿地实施模式.

In view of the serious turbulence near the pier and the smooth and difficult adsorption of the pier surface,a ROV for pier defect detection was designed by using the composite control technology.Firstly,the ROV body structure adopted the eight-vector thruster layout scheme,which realized the omnidirectional motion of the robot floating state and improved the ROV immunity.Secondly,the holding arm clamping mechanism adopted the rope transmission method,which provided a reliable platform for the stable detection in the ROV crawling state with simple mechanical linkage.Finally,the YOLOv5s network based on attention mechanism was used to identify pier defects,and the real-time and accuracy of defect detection were improved by introducing ASPP,IAAM and other structures.The results of real water test showed that the designed ROV had good applicability and effectiveness,and achieved a good application effect of pier defect detection in complex water environment,which could replace divers to complete pier defect collection.

针对在变工况中采集到的信号通常为非平稳信号,并且采集到的振动数据分布不一致等情况,提出了基于角域重采样的宽度迁移学习(ADR-BTL)算法,用于轴承故障诊断.首先将非平稳的时域振动信号转换为角域平稳信号,将平稳化处理后的信号进行特征提取,构建多域特征数据集,然后将不同工况下的源域数据和目标域数据通过平衡分布自适应(BDA)方法进行领域适配来减小域间的分布差异,最后构建宽度迁移学习模型.实验首先验证了角域重采样方法可以将振动信号进行平稳化处理,然后通过仿真样本分析得出BDA方法能够解决数据分布不一致问题,最后通过实验结果得出,提出的ADR-BTL识别率达到了 98.9％,识别效果是最好的,证明了所提的方法在轴承故障诊断方面是有效的.

城市绿地系统和公园体系是城市公共服务体系的重要内容,也是城市人居环境中"最公平的公共产品和最普惠的民生福祉"[1-2].但在存量城区,由于受到人口密集、空间有限等因素的制约,"人多地少""房多绿少"的问题非常突出.作为典型的存量城区,湖北省武汉市江汉区的人口密度位列全市第一,绿地总量和人均指标都低于中心城区平均水平,区内建筑密集、空间局促,已无新增建设用地可用于绿化建设.如何在空间资源有限的条件下,因地制宜开展存量城区的绿地规划建设,已成为新时期江汉区高质量发展亟待解决的问题.

针对情感分类研究中广泛存在的数据不平衡问题,提出了一种基于边界度的过采样方法(BD-SMOTE).首先,根据少数类样本的多数类最近邻和少数类最近邻确定其边界度;其次,根据边界度计算少数类样本的采样权重;最后,根据采样权重自适应确定每一个少数类样本需要生成新样本的数量.实验结果表明,将该算法应用于不平衡情感数据集并结合SVM分类器训练分类模型,实现了准确分类.

Roller bearings are commonly used components in rotating machinery and are pruned to be failure, which may cause huge economic losses and even casualties. Vibration signals of rolling bearings are signs of rotating machinery. Therefore, they can be used to describe the process of performance degradation. Unfortunately, the signals exhibit non-stationary and non-linear as well as highly non-monotonic behaviors as the bearing condition degrades especially under complex working conditions. In this paper, a method to construct a constitution health index based on vibration signals is presented. Firstly, 40 features in time domain, frequency domain, time-frequency domain and entropy domain are extracted individually. Then, the combined evaluation index is constructed based on the above 40 features in terms of monotonicity, robustness, predictability and correlation. Finally, by comparing the degradation fusion indexes of rolling bearings under different operating conditions with the traditional characteristics, the remaining useful life prediction method based on grey correlation analysis is proposed. The results suggest that the proposed method can effectively predict the RUL of rolling bearings with an acceptable degree of accuracy up to 90%, which outperforms the traditional features.

针对目标快速移动和遮挡情况导致的客流统计存在误差的问题,设计目标的检测、跟踪、进门行为判断等策略,提出基于深度学习的餐饮业客流统计方法.首先,通过多数据集对YOLOv3-tiny模型进行训练,实现对于小目标的准确检测;进而设计多通道特征融合的目标跟踪算法,完成目标快速移动情况下的稳定跟踪;最后设计目标进门行为的判断方法,通过重叠率对目标的进门行为进行判断,实现对进门客流量的准确统计.最终通过实验验证,客流量统计的平均准确率达到93.5％.

为了提高复杂背景噪声环境下的车型识别准确性,该文基于近似熵理论,对机动车行驶中辐射的声信号进行了研究.近似熵具有抗干扰能力强的特点,可用于提取动态背景噪声下机动车声信号的车型特征信息.首先,对声信号进行3层小波包分解;然后,利用近似熵量化第3层上各子频带信号的不规则性,描述各频带之间不同的变化趋势并作为目标车辆的声特征.为了提高分类有效性,将分解后的8个子频带信号的近似熵邻比值作为信号的特征向量,并基于支持向量机分类器实现了车型识别.分别在正常和有风两种气候条件下进行了实验,基于小波包近似熵的车型特征均获得了较为理想的分类精度.实验结果显示,小波包近似熵特征能有效地应用于机动车的声识别且对气候的影响具有一定的鲁棒性.

谱聚类算法中,当样本的簇边缘分布不均匀或不同簇边缘分布密度相近时,会导致错分现象.通过对相似度矩阵的改进,提出基于流形距离核的自适应迁移谱聚类算法.使用流形距离作为构造相似度矩阵的度量方法,共享近邻方法对相似度矩阵进行自适应调整,且使用加权距离自适应调节核参数,提高谱聚类对复杂数据集的处理能力.针对样本匮乏或受到污染时聚类效果不佳问题,引入迁移学习,利用源域知识指导目标域进行聚类.经实验验证,该算法性能优于传统谱聚类算法.

User intent analysis is a continuous research hotspot in the field of query expansion. However, the big amount of irrelevant feedbacks in search log has negatively impacted the precision of user intent model. By observing the log, it can be found that tentative click is a major source of irrelevant feedback. It is also observed that a kind of new feedback information can be extracted from the log to recognize the characteristics of tentative clicks. With this new feedback information, this paper proposes an advanced user intent model and applies it into query expansion. Experiment results show that the model can effectively decrease the negative impact of irrelevant feedbacks that belong to tentative clicks and increase the precision of query expansion, especially for those informational queries.

国际社会普遍认为一个有吸引力、活跃且完善的公共空间有助于建立社区意识和市民认同,同时也可以推动经济和文化的发展.武汉近年来一直在积极践行"人居Ⅲ"和《新城市议程》关于公共空间议题的有关倡议,与联合国人居署合作开展了若干"中国改善城市公共空间"试点项目.文章以武汉江汉区为例,尝试构建有中国本土特色的城市公共空间评估指标体系,探索运用新型技术手段解决公共空间评估过程中遇到的核心问题,同时根据评估结论提出合理化的规划建议,制定公共空间设计导则,以期促进高品质城市公共空间的回归,推动城市可持续发展.

Transfer learning is a method of helping target domain to cluster or classify using source domain knowledge. It can solve the problem of low amount of data available or having noise in the target domain when clustering or classifying. But the distance measure of the common transfer clustering usually uses the traditional European distance. In some cases, the method of similarity measurement based on traditional European distance cannot reflact the complex spatial structures between different data clusters. In order to solve this problem, the transfer spectral clustering based on manifold distance (TSC-MD) algorithm is proposed, which utilizes the manifold distance as a measure of similarity. TSC-MD algorithm can improve the transfer clustering performance of the original algorithm by taking into account the global consistency and complex spatial distribution characteristics of the sample. Compared with the clustering algorithm of similar transfer learning methods, the experiments of synthetic and real-life scenarios demonstrate that the algorithm can achieve better clustering results.

In recent years, indoor and outdoor positioning technology for the blind has attracted much more attention. The outdoor positioning technology has relatively matured. However, three-dimensional indoor positioning technology is facing many problems. Due to low positioning accuracy as well as slow positioning speed, there is no ideal implementation system at present. Light Fidelity (LiFi) technology is a new wireless transmission technology that uses the visible light spectrum, such as the light emitted by a light bulb, to transmit data. It transmits signals through the changing light of LED lights. When LED lights are properly laid out in the building, not only the function of lighting but also that of transmitting information are met at the same time. It is a good way to solve indoor positioning problems. This article studies an indoor positioning glasses for blind based on the LiFi technology. Different from other indoor positioning technologies such as ultrasound and wireless fidelity (WiFi), the design of indoor positioning glasses for the blind will propose new ideas of three-dimensional indoor positioning problems for the blind.

In recent years, the outbreak of natural disasters is becoming more and more frequent, which makes the damage to people's property and life getting worse and worse.Therefore, how to distribute emergency relief supplies after the disaster becomes very important.In this paper, taking the urgent needs of the victims after the disaster into account, based on the shortest delivery time algorithm, a mathematical model of route optimization is established and the model is solved by improved genetic algorithm.The proposed method is applied on a given example with Matlab7.0, and two optimal schemes are obtained, which verified the rationality of the proposed mathematical model and the feasibility of the improved algorithm.

针对动态噪声环境下行进中的机动车辐射出的声信号具有强非平稳性、多尺度性及低信噪比的问题,提出一种基于局部均值分解(LMD)和局部投影能量计算的车型声特征提取方法.首先,利用LMD方法对采集的声信号进行自适应分解,得到各尺度上的乘积函数(PF)分量,从强背景噪声中分离出包含车型特征频率成分的PF分量;其次,对LMD分解结果进行加权优化,重构特征PF分量,滤除虚假成分及弱相关分量,以增强特征信息;最后,将特征PF分量的能量等距离投影到能量聚集区内,基于能量尺度构造声信号的低维特征向量,并通过人工神经网络的学习对特征向量进行识别.在试验数据集上,采用LMD局部投影能量特征对目标车辆进行车型识别,并对试验数据集添加不同强度的噪声,进行LMD分解及局部投影能量计算,将计算结果与其他特征提取方法计算结果进行对比分析.结果表明:该方法对于车型信息十分敏感,识别率达到93.4％;可以有效抑制动态环境下的背景噪声干扰,获取目标敏感的窄带信号,具有很好的抗噪能力;选择在重构窄带信号的能量聚集区内进行投影计算,可以有效去除冗余特征,同时提高算法的实时性.

Roller bearings are commonly used components in rotating machinery and are pruned to be failure, which may cause the system break down and result in economic loss. Therefore, performance degradation prediction of rolling bearings is important to prevent any unexpected roller bearings failure. In this paper, time - frequency image fusion technology as well as the BP neural networks are utilized for fault feature extraction and prediction based on vibration signals. BP neural networks are used to learn the fault prediction features of vibration signals. Finally, the test data is used to testing the whole neural networks to establish the bearing condition monitoring model. Experimental results show that the proposed method achieves is about 80% accuracy to the bearing state recognition, and the recognition rate of the degraded performance stage is the highest, which can meet the engineering requirements.

Shenyang City was the political, economic and cultural centre of the northeast China and was also a heavily polluted industrial city. The understanding of the distribution characteristics of air pollutants concentration and changes was still lacked. To reduce the impact of disturbing factors such as firecrackers in the traditional Chinese festivals, the observation period of the monitoring data was selected from 17 th May to 21 st July in 2016. The data sources were picked from eight national monitoring stations and the daily average concentration of the main air pollutants that included PM2.5, PM10, SO2, NO2 and O3. The overall analysis of distribution characteristics of the air pollutant was shown that the principal pollutants with highest frequency were O3 and PM10, the average proportion was 74.1% and 20.8% respectively.

Shenyang was a representative heavily polluted industrial city in northeast China. Two types of urban area with significant spatial position difference were divided by arterial traffic as the surrounding and the core urban area. To reduce the impact of disturbing factors such as firecrackers in the traditional Chinese festivals, the observation period of the monitoring data was selected from November 1, 2015 to January 1, 2016. The data sources were picked from eight national monitoring stations and the daily average concentration of the main air pollutants that included PM2.5, PM10, SO2, NO2 and O-3. The overall analyzes of distribution characteristics of the air pollutant was shown that the principal pollutants with highest frequency were PM10, SO2 and PM2.5 , the average proportion was 70.36%, 14.11% and 11.89% respectively. In the contrastive analysis of the core urban area and the surrounding urban area, the principal pollutants in core urban areas was PM10, SO2 and PM2.5, but in surrounding areas PM10, PM2.5 and SO2, which were quite different. So the conclusion that the emission of SO2 in the core urban areas was greatly higher than that in the surrounding urban areas was confirmed.

In signal processing, the estimation of instantaneous frequency is a kind of very important issue for non-stationary multi-component signal. Firstly, this paper adopts EMD decomposition to decompose non-stationary multi-component signal and convert into a series of single component signal. Then using the method of hilber transform and teager energy operator respectively, the instantaneous frequency of single component can be obtained. Finally, the simulation results show that adopting hilber transform can make root-mean-square error of instantaneous frequency and amplitude of fluctuation more smaller. So this method presented in the paper is feasible and effective.