【作者】 刘鹏；

【导师】 韩启龙； 邹乃峰；

【作者基本信息】 哈尔滨工程大学 ， 软件工程（专业学位）， 2016， 硕士

【摘要】 地理信息系统(Geographic Information System)和水声传感器网络(Underwater Acoustic Sensor Networks)的快速发展使得水下移动对象的跟踪和监测成为现实。然而在水下未知目标识别的研究中,传统方法针对恶劣的水下环境、复杂的水下移动对象运动轨迹和目标分类识别还存在不足。本文主要针对以上不足,对水下移动对象识别过程中的时空数据模型和识别方法进行研究。主要内容如下:1)系统地研究水下移动对象的轨迹特征和属性特征,并详细分析这些特征对水下移动对象识别精度的影响。2)针对水下移动对象轨迹的时空特点,建立水下移动对象模型(Underwater Moving ObjectDatabase,UMOD)。首先,提出了基于球坐标的轨迹分解方法,降低了轨迹的整体复杂程度;其次,UMOD模型采用移动对象时空模型中的动态属性方法,提出了针对水下未知目标的动态阈值位置更新策略,辅助轨迹分解并对误差点进行排除;最后,提出了轨迹拟合策略,对水下移动对象进行全时域轨迹描述,从而系统地得到水下移动对象运动过程中的轨迹识别特征。3)分析和研究水下移动对象的轨迹特征和属性特征,建立水下移动对象特征数据库,同时提出水下移动对象混合特征识别方法(Underwater Moving Objects Mixed Feat-ure Recognition Method,UMOMFRM)。首先,UMOMFRM通过模糊推理的方法排除目标的最小概率分类;其次,采用贝叶斯分类算法计算未知目标对应剩余项目的概率;最后,对这些概率进行概率阈值筛选并进行最大概率比较,若存在最大概率,则其所对应的类别为分类结果,否则该目标所属类别为其他类。4)采用模拟数据集进行对比实验。首先,根据本文提出的UMOD模型对轨迹数据进行处理,验证UMOD模型的准确性。最后,进行UMOMFRM与传统水下目标识别方法的对比试验,比较识别率,以验证UMOMFRM的准确性和有效性。更多还原

【Abstract】 With the rapid development of GIS and Wireless communication, the tracking and monitoring of the underwater moving objects becomes true. In the study of the underwater objects,we have noticed that it is not enough to indentify the kinds of objects and their motion track which are in the harsh underwater environment by traditional methods. In order to solve that, the method of identification of the underwater moving objects has been studied,also its temporal data model is given in this paper. The main contents are as follows,1) The characteristics of properties and tracks of the underwater moving objects have been studied systematically, the influences of them on the accuracy of identification are also analyzed in detail2) For the temporal track characteristics, UMOD, Underwater Moving Object Database,is established. First of all,put forward the trajectory decomposition method based on the spherical coordinates,to reduce the whole complexity of the trajectory;The second,proposed in view of the underwater unknown target of dynamic threshold position update strategy,which introduces the dynamic properties method in the Moving Objects Spatio-Temporal model,exclude the Secondary path decomposition and the error points;At last,the fitting track strategy is proposed to describe the underwater moving objects real-timely, then obtaining the track characteristics of the underwater objects during their moving processes.3) With the study of the characteristics of properties aOnd tracks, the corresponding database is established. At the same time,UMOMFRM, Underwater Moving Objects Mixed Feat -ure Recognition Method, is proposed to indentify the unknown objects.First of all,UMOMFRM excluded the minimum probability of target classification,which uses a fuzzy reasoning method; The second,classify unknown target corresponds to the rest of the calculation of probability by Using bayes classification algorithm;At last,compare the probability of maximum on the basis of the probability threshold value,if there is maximum probability,it is the classification result which category the probability corresponds to,otherwise, the target category for other class.4) Contrast experiment was carried out using simulated data set.First of all, according to the method of UMOD model, to verify the validation of the UMOD model. At last, the UMOMFRM is compared with the existing methods of indentifying the underwater objects through the contrast experiment , to verify the accuracy and validation of UMOMFRM model.更多还原