一种结合均值和互信息的线上AP选取新算法

探讨了基于RSS(receivedsignalstrength)的WiFi室内定位技术的AP(accesspoints)选取算法,并分析了两种现有AP选取算法对WiFi室内定位性能的影响。基于均值最大的AP选取算法存在无法顾及AP相互干扰的局限,而基于线上互信息最小的AP选取算法计算过程复杂,当AP个数较多时,其时耗较大,同时难以完全摒弃观测质量过差的AP。考虑到上述两种方法的局限,提出了一种线上AP选取新算法,首先利用均值最大选取算法对AP进行预处理,从而加快AP选取的速度,并剔除部分观测质量较差的AP,然后利用互信息进行AP的精确选取。实验分析表明,新算法能够有效地提高位置估计的速度,并且改进了位置估计精度。     

雄安新区建设数字孪生城市的逻辑与创新

川藏铁路广域范围内实体要素粒度差异大、语义繁杂,多学科多专业空间信息数据多源异构,时空基准与数据格式不统一,难以高效集成管理与融合分析。为构建川藏铁路时空信息统一表达的认知共享概念体系,提出地理、地质、设施、气象、生态、灾害、人员7大实体要素类型,并对其几何、尺度、拓扑、属性、行为特征及相互作用关系等进行刻画,将要素类型进行实体化分类,制定了川藏铁路实体要素的统一编码规则以及扩充方法,对高位分类（门类、亚门类、大类和中类）以及低位分类（一、二、三级小类）进行扩充与删减,形成共包含2个门类、6个亚门类、27个大类和87个中类的高维分类体系。在此基础上,建立川藏铁路实体要素分类体系,并对每一个实体要素进行编码,赋予其唯一的、无歧义的身份标识,实现川藏铁路多维动态时空信息与实体要素之间的精准映射,为数字孪生川藏铁路建设奠定坚实基础。     

利用C_SURF配准的空基视频运动目标检测

为解决空中多目标检测中目标在图像中所占像素数少、目标相互毗邻和相互遮挡的难题,提出了联合时空信息和轨迹关联的空中多目标检测算法。该算法首先利用基于像素的背景建模算法得到目标空间信息,结合邻近帧差算法提取的目标时间信息,联合得到目标的时空信息图;然后,利用卡尔曼预测器对目标位置进行预测,利用匈牙利匹配算法关联目标得到目标轨迹,根据目标轨迹补充漏检目标, 提高目标的查全率;最后,分段提取目标轨迹特征,根据轨迹特征进行判断,滤除虚警目标, 提高目标的查准率。采用空中多目标视频进行实验验证,结果表明,本文算法检测性能优良,查全率高于96%,查准率高于98%,F测度高于97%。     

应急测绘保障体系若干关键问题研究

  Objectives   China is one of the countries with the most severe natural disasters in the world. There are many types of natural disasters and high frequency of disasters. It often causes huge economic losses and casualties in the disaster area. Combining the key technologies of the internet of things(IOT), the Internet, and geographic information for surveying and mapping, so it is possible to grasp the disaster situation in a timely and efficient manner. Thereby we could scientifically and effectively formulate disaster prevention and mitigation strategies, and implement emergency rescue measures, which are of great significance to the safety of people’s lives and property.  Methods   The development of big data and disaster reduction technologies has brought new possibilities for disaster management and emergency response. Geographic information big data provides basic support for emergency rapid response, location and navigation big data provides location information of personnel and materials for emergency rescue. Two- and three-dimensional visualization technology more realistically restores the disaster environment, and disaster models scientifically simulate the process of disaster occurrence, internet public opinion big data in disaster assessment also plays an important role. Big data is profoundly changing emergency rescue and disaster management with its ability to visualize, analyze and predict disasters. The combination of disaster big data and surveying and mapping geographic information technology has become an important means to improve disaster prevention and mitigation capabilities.  Results   We analyze and summary the integrated and comprehensive disaster reduction technology in the context of big data. The characteristics of integrated and comprehensive disaster reduction technology are elaborated, including comprehensive and thorough perception, broadband ubiquitous interconnection, precise and fast integration, ubiquitous positioning, efficient and intelligent processing, and intelligent collaborative services. The progress of integrated comprehensive disaster reduction technology is reviewed, which includes indoor and outdoor high-precision integrated positioning, rapid integration of multi-source data, scene fusion and enhanced visualization, typical disaster model construction and management, integrated disaster reduction comprehensive services, and China’s disaster reduction service system platform. In addition, we introduced the integrated comprehensive disaster reduction intelligent service system platform, APP and its application services, and summarized and prospected the integrated comprehensive disaster reduction technology and services.  Conclusions   The continuous development of big data technology provides new ideas and methods for people to analyze and solve emergency response problems, and provides a very convenient way for the sharing, integration, analysis, mining, and decision-making of comprehensive disaster reduction services. However, with the in-depth application of new concepts, new technologies, and new methods such as      

地质导向技术在L型煤层气水平井 T-P05井中的应用

通过分析山西L型煤层气水平井的技术难点,着重建立造斜着陆与水平段导向两大导向模型,形成了以标志层划分与对比、地层倾角推算、地质建模和实时轨迹控制四个方面为主的L型煤层气水平井地质导向技术流程,该技术在T-P05井的成功应用,不仅丰富了煤层气水平井地质导向技术的内容与方法,更进一步验证了该技术在现场具有较强的可操作性和良好的应用前景,值得大力推广。     

大直径加重管绳索取心技术在页岩气勘探中的应用研究

根据页岩气、煤层气勘探对钻探技术的特殊要求,进行了大直径绳索取心技术的探索研究。文中重点介绍了加重管组合式绳索取心钻具的结构、作用原理及特点,以及施工试验效果。通过施工实践的启示,对大直径绳索取心技术的发展提出了几点认识。     

页岩气水平井TOC测井评价新方法

由于存在地层缺失、井眼穿层不断变化、重点层段测井曲线较少、井眼环境复杂、测井响应值与储层参数关系较为复杂等特点,页岩气储层水平井测井解释难度较大.针对该问题,本文采用“水平井钻遇地层模型—水平井间及水平井与直井响应差异分析—水平井曲线校正—基于直井的参数评价模型”的评价技术流程,进行页岩气水平井TOC测井评价.在测井评价中,针对响应规律复杂的特点,采用直方图法进行测井响应规律分析,同时考虑到水平井测井曲线较少,为了综合利用测井曲线信息,结合曲线重叠方法提出双重信息融合评价模型.结果 表明,本方法在充分分析测井响应差异的基础上对水平井测并曲线进行校正,且提出的双重信息融合模型具有能充分利用测井信息、评价精度高以及操作简单等优点.     

一种基于机器学习算法的岩性填图方法

通过野外地质调查与机器学习方法的有机融合,提出了一种基于梯度提升决策树算法的岩性单元填图方法。研究以多龙矿集区为模型试验区,选择1∶5万勘查地球化学数据为基础预测数据,以1∶5万区域地质图为参考,进行基于梯度提升决策树算法的岩性预测填图模型试验。首先选择研究区内小范围空白区开展野外填图,建立原始数据集并初步构建岩性单元与预测数据对应关系;其次利用机器学习方法对预测数据进行多分类任务,进而开展目标填图区预测填图工作;最后通过概率选区选定概率较低目标区,开展进一步的小范围野外地质调查填图,对原始数据和知识库进行补充,迭代循环以上流程,直至预测填图达到要求。试验显示,随着迭代次数的增加,模型精度不断提高,并在7次迭代后模型准确率达到87%。该方法强调在实际应用中野外地质调查与基于机器学习预测填图的深度融合,以及野外实地工作在整个流程中的重要性和不可或缺性;同时能够充分挖掘已有数据资料的有用信息,用于辅助修正已有岩性填图内容,或根据已勘探区资料对邻近的未勘探区进行岩性分类,有效减少野外填图工作量,是对岩性填图方法、地质单元定量预测识别的有益探索,为区域地质填图工作提供了新的参考思路和辅助手段。