多径环境下多基地超宽带雷达定位与跟踪技术研究
发布时间：2018-10-14 12:21
【摘要】：随着地理位置服务的发展,室内定位技术有着广阔的应用前景,如室内监控、商场导航等。超宽带技术由于系统结构简单、抗多径、穿透性强、测距精度高、共存性好,在复杂多径环境下的室内定位中占有优势,因而成为近十年来的研究热点。但是,该技术在复杂环境下超宽带信号传播特性分析、实用高精度测距算法设计等方面仍面临诸多难题。超宽带定位技术还可应用于战场侦察与监视。相对于单基地高频地面侦察雷达,多基地超宽带地基侦察雷达具有很多优势,如抗摧毁、抗干扰、可探测隐蔽目标等,是战场侦察雷达的重要发展方向。在近程多基地超宽带雷达中,如何在复杂未知环境下实现对运动人员的可靠定位与持续稳健跟踪是其关键技术问题和重要研究内容。本文针对室内和树林等多径环境,结合多项科研项目,对超宽带技术在目标定位与跟踪中的应用问题展开研究。研究内容包括复杂环境下超宽带信号传播特性、实用的超宽带测距算法、多基地系统理论上的定位及跟踪性能、多基地定位中虚假定位点剔除和多目标定位、多径环境下多基地稳健跟踪等。经仿真和实测数据验证的研究成果可以应用到超宽带定位、多基地战场侦察雷达设计、多基地定位和跟踪等方面。1、针对测距和定位应用,分析了IEEE 802.15.4a信道模型在不同环境下的部分统计特性,如首径后第1条多径的位置及幅度分布、首径后2ns内多径的条数及幅度的分布等,所得结论为测距算法设计奠定了基础。提出了两段多径模型以描述脉冲超宽带信号在发射站—非合作目标—接收站间传播时的信道特性。基于实测数据分析了超宽带信号的穿墙传播特性,结果表明多径在超宽带信号穿墙传播中存在。对背景相消后的剩余回波分析证明了复杂环境下被监视目标运动时,存在与目标相关的多径现象。同时,由于环境的改变,特别是相对位置关系的变化,会导致跟随目标的多径发生变化。2、为降低TOA估计性能对门限的严格要求,提出了能量检测接收机下简单实用的脉冲超宽带TOA估计新方法。该方法利用了接收信号统计特性,即不同帧噪声样本超过门限的随机性和首径样本的确定性。在密集多径环境下,提出了一种基于熵的无需门限的首径检测新方法。该方法从另一个角度考虑首径样本检测问题。通过对信道的分析,可知密集多径环境下多径区域具有稳定性,特别是首径附近,新方法通过检测熵序列的突变检测多径区域,进而确定首径样本。为避免使用复杂的chi平方分布,本文提出对ED接收机输出样本序列进行幂变换,然后拟合变换数据序列的分布,在最佳近似分布基础上估计TOA,获得了与基于较复杂的理论上的chi平方分布的TOA估计方法类似的性能。分析了ED接收机部分参数,并提出了基于延时器的具有更高精度的TOA估计方法。3、从理论上系统地分析了多基地雷达系统两个方面的性能:不同状态的可观测性和不同状态估计的CRLB。分别分析了两种情形下的可观测性,多站瞬间位置和速度的可观测性以及不同航迹的跟踪可观测性。分别分析推导了多基地下瞬间目标状态估计的CRLB和跟踪的CRLB。理论分析和仿真结果可以为系统设计提供指导。4、在分析多基地下单接收站距离或多普勒观测误差变化对定位和测速的影响基础上,首先提出并推导了速度一致性准则,即在同一个时间,目标只可能处于一个位置,速度大小和方向应唯一。不同观测组合估计得到的目标速度应是一致的,而虚假观测组合估计的数个速度不一致。据此可判别真假观测组合,然后基于真假组合的判定和速度一致性完成不同站观测的正确配对和多目标定位。仿真结果表明该方法具有很高的虚假组合识别成功率。5、提出了多径环境下的基于轨迹关联性的真假目标轨迹辨识算法,该算法把伴随动目标的多径回波信号看作来自一个虚假目标,根据双站距离和和多普勒观测得到目标的当前航迹,然后求出反射点,根据路径关系识别真实轨迹和虚假轨迹,当目标向基线靠近时识别率较高。在稳健跟踪方面,为应对数据缺失和问题观测,设计了不同观测维度下的稳健跟踪流程,该流程通过自适应改变滤波器结构实现对目标的持续跟踪;为应对干扰等带来的错误观测,提出了基于错误观测组合辨识的复杂环境下稳健跟踪方法,其使用预测与观测最邻近、速度和位置一致性准则来辨识错误组合。针对实际战场上常见的之字形战术机动人员机动跟踪问题,提出了基于分段CV模型的IMM算法。新方法建立了机动人员的运动模型,可更加准确的描述目标真实的运动状态。所提方法中的模型集可以及时适应目标运动方向的变化,在突变发生时可对跟踪模型集的速度方向不断调整直至达到稳态,使状态分量估计更接近目标真实的速度方向。仿真表明所提算法可以获得比传统机动算法更好的跟踪性能。
[Abstract]:With the development of geographical location service, indoor positioning technology has wide application prospect, such as indoor monitoring, market navigation and so on. Ultra-wideband technology has the advantages of simple system structure, multi-diameter, strong penetrability, high ranging precision and good coexistence, and has an advantage in indoor positioning under complex multi-path environment, thus becoming a hot spot in the last decade. However, this technology still faces many difficulties in the analysis of ultra-wide band signal propagation characteristics, practical high precision distance measurement algorithm design and so on in complex environment. Ultra-wideband positioning technology can also be applied to battlefield reconnaissance and surveillance. Compared with single-base high-frequency ground reconnaissance radar, multi-base ultra-wideband ground reconnaissance radar has many advantages, such as anti-destruction, anti-jamming, detection of hidden targets and so on, and is an important development direction of battlefield reconnaissance radar. In short-range multi-base ultra-wideband radar, how to realize reliable positioning and continuous robust tracking of moving personnel in complex unknown environment is its key technology problem and important research content. In view of the multi-path environment such as indoor and wood, combining several scientific research projects, this paper studies the application of ultra-wideband technology in target positioning and tracking. The research content includes the characteristics of ultra-wide band signal propagation in complex environment, practical ultra-wideband distance measuring algorithm, positioning and tracking performance in multi-base system theory, false localization point elimination in multi-base location and multi-target positioning, multi-base robust tracking in multi-path environment, etc. The results of simulation and data verification can be applied to the aspects of ultra-wide band positioning, multi-base battlefield reconnaissance radar design, multi-base location and tracking, etc. 1. For ranging and location applications, some statistical properties of IEEE 802.15. 4a channel model in different environments are analyzed. The results are as follows: the location and amplitude distribution of the first multi-path, the number of multi-paths and the distribution of the amplitude within 2ns after the first diameter, and the conclusion is the foundation for the design of the distance-measuring algorithm. A two-stage multi-path model is proposed to describe the channel characteristics of a pulsed ultra-wideband signal propagating between a transmitting station and a non-cooperative target transceiver station. Based on the measured data, the wall propagation characteristics of ultra-wideband signals are analyzed, and the results show that the multi-path exists in the through-wall propagation of ultra-wideband signals. The residual echo analysis after the background cancellation proves that there is a multi-path phenomenon related to the target when the target motion is monitored in a complex environment. At the same time, due to the change of the environment, especially the relative positional relation, the multi-path change of the following target can be caused, and the new method of the simple and practical pulse ultra-wideband TOA estimation under the energy detection receiver is proposed in order to reduce the requirement of TOA estimation performance to the threshold. The method utilizes the statistical characteristic of the received signal, that is, the randomness of the different frame noise samples exceeds the threshold and the certainty of the first path sample. In the dense multi-path environment, a new method for detecting head diameter without threshold based on entropy is proposed. The method considers the first-diameter sample detection problem from another point of view. Through the analysis of the channel, it can be seen that the multi-path region has stability, especially near the first diameter, in the dense multi-path environment, and the new method detects the multi-path region by detecting the mutation of the entropy sequence, and then determines the first diameter sample. In order to avoid the use of complex chi-square distribution, this paper proposes a power transformation to the output sample sequence of ED receiver, then fits the distribution of transform data sequence, and estimates TOA on the basis of the optimal approximate distribution. A similar performance to TOA estimation methods based on a more complex theoretical chi-square distribution is obtained. In this paper, some parameters of ED receiver are analyzed, and a TOA estimation method based on delayer is proposed. The performance of two aspects of multi-base radar system is systematically analyzed theoretically: the observability of different states and the CRLB of different state estimation. The observability of two cases, the observability of multi-station instantaneous position and velocity, and the tractability of different flight paths are analyzed respectively. CRLB and CRLB for tracking target state estimation under multi-base are analyzed respectively. Theoretical analysis and simulation result can provide guidance for system design. 4. On the basis of analyzing the influence of distance or Doppler observation error of single receiving station on positioning and velocity measurement, this paper first puts forward and deduces the speed consistency criterion, that is, at the same time, The target may only be in one position, and the speed and direction should be unique. The target velocities obtained by different observation combinations are consistent, and several velocities of false observation combinations are inconsistent. and then the correct pairing and multi-target positioning of different station observations are completed based on the judgment and the speed consistency of the true and false combination. The simulation results show that the method has a high success rate of false combination recognition. 5. An algorithm based on track correlation based on track correlation in multi-path environment is proposed, which treats the multi-path echo signal with moving target as a false target. According to the two-station distance and the Doppler observation, the current track of the target is obtained, then the reflection point is obtained, the real track and the false track are identified according to the path relation, and the recognition rate is higher when the target is close to the baseline. In the aspect of robust tracking, in order to cope with the data loss and problem observation, a robust tracking process in different observation dimensions is designed, which realizes the continuous tracking of the target through the adaptive changing filter structure, and the error observation caused by the interference and the like is achieved, A robust tracking method based on error observation combined identification is presented, which uses prediction and observation nearest neighbor, velocity and position consistency criterion to identify the error combination. In this paper, a new IMM algorithm based on segmented CV model is proposed. The new method establishes the motion model of maneuver personnel, and can describe the target's real motion state more accurately. The model set in the proposed method can adapt to the change of the target motion direction in time, and the speed direction of the tracking model set can be continuously adjusted until the steady state is reached when the mutation occurs, so that the state component estimation is closer to the target real speed direction. Simulation results show that the proposed algorithm can achieve better tracking performance than traditional maneuvering algorithm.
【学位授予单位】：国防科学技术大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN957.51
，
本文编号：2270436
[Abstract]:With the development of geographical location service, indoor positioning technology has wide application prospect, such as indoor monitoring, market navigation and so on. Ultra-wideband technology has the advantages of simple system structure, multi-diameter, strong penetrability, high ranging precision and good coexistence, and has an advantage in indoor positioning under complex multi-path environment, thus becoming a hot spot in the last decade. However, this technology still faces many difficulties in the analysis of ultra-wide band signal propagation characteristics, practical high precision distance measurement algorithm design and so on in complex environment. Ultra-wideband positioning technology can also be applied to battlefield reconnaissance and surveillance. Compared with single-base high-frequency ground reconnaissance radar, multi-base ultra-wideband ground reconnaissance radar has many advantages, such as anti-destruction, anti-jamming, detection of hidden targets and so on, and is an important development direction of battlefield reconnaissance radar. In short-range multi-base ultra-wideband radar, how to realize reliable positioning and continuous robust tracking of moving personnel in complex unknown environment is its key technology problem and important research content. In view of the multi-path environment such as indoor and wood, combining several scientific research projects, this paper studies the application of ultra-wideband technology in target positioning and tracking. The research content includes the characteristics of ultra-wide band signal propagation in complex environment, practical ultra-wideband distance measuring algorithm, positioning and tracking performance in multi-base system theory, false localization point elimination in multi-base location and multi-target positioning, multi-base robust tracking in multi-path environment, etc. The results of simulation and data verification can be applied to the aspects of ultra-wide band positioning, multi-base battlefield reconnaissance radar design, multi-base location and tracking, etc. 1. For ranging and location applications, some statistical properties of IEEE 802.15. 4a channel model in different environments are analyzed. The results are as follows: the location and amplitude distribution of the first multi-path, the number of multi-paths and the distribution of the amplitude within 2ns after the first diameter, and the conclusion is the foundation for the design of the distance-measuring algorithm. A two-stage multi-path model is proposed to describe the channel characteristics of a pulsed ultra-wideband signal propagating between a transmitting station and a non-cooperative target transceiver station. Based on the measured data, the wall propagation characteristics of ultra-wideband signals are analyzed, and the results show that the multi-path exists in the through-wall propagation of ultra-wideband signals. The residual echo analysis after the background cancellation proves that there is a multi-path phenomenon related to the target when the target motion is monitored in a complex environment. At the same time, due to the change of the environment, especially the relative positional relation, the multi-path change of the following target can be caused, and the new method of the simple and practical pulse ultra-wideband TOA estimation under the energy detection receiver is proposed in order to reduce the requirement of TOA estimation performance to the threshold. The method utilizes the statistical characteristic of the received signal, that is, the randomness of the different frame noise samples exceeds the threshold and the certainty of the first path sample. In the dense multi-path environment, a new method for detecting head diameter without threshold based on entropy is proposed. The method considers the first-diameter sample detection problem from another point of view. Through the analysis of the channel, it can be seen that the multi-path region has stability, especially near the first diameter, in the dense multi-path environment, and the new method detects the multi-path region by detecting the mutation of the entropy sequence, and then determines the first diameter sample. In order to avoid the use of complex chi-square distribution, this paper proposes a power transformation to the output sample sequence of ED receiver, then fits the distribution of transform data sequence, and estimates TOA on the basis of the optimal approximate distribution. A similar performance to TOA estimation methods based on a more complex theoretical chi-square distribution is obtained. In this paper, some parameters of ED receiver are analyzed, and a TOA estimation method based on delayer is proposed. The performance of two aspects of multi-base radar system is systematically analyzed theoretically: the observability of different states and the CRLB of different state estimation. The observability of two cases, the observability of multi-station instantaneous position and velocity, and the tractability of different flight paths are analyzed respectively. CRLB and CRLB for tracking target state estimation under multi-base are analyzed respectively. Theoretical analysis and simulation result can provide guidance for system design. 4. On the basis of analyzing the influence of distance or Doppler observation error of single receiving station on positioning and velocity measurement, this paper first puts forward and deduces the speed consistency criterion, that is, at the same time, The target may only be in one position, and the speed and direction should be unique. The target velocities obtained by different observation combinations are consistent, and several velocities of false observation combinations are inconsistent. and then the correct pairing and multi-target positioning of different station observations are completed based on the judgment and the speed consistency of the true and false combination. The simulation results show that the method has a high success rate of false combination recognition. 5. An algorithm based on track correlation based on track correlation in multi-path environment is proposed, which treats the multi-path echo signal with moving target as a false target. According to the two-station distance and the Doppler observation, the current track of the target is obtained, then the reflection point is obtained, the real track and the false track are identified according to the path relation, and the recognition rate is higher when the target is close to the baseline. In the aspect of robust tracking, in order to cope with the data loss and problem observation, a robust tracking process in different observation dimensions is designed, which realizes the continuous tracking of the target through the adaptive changing filter structure, and the error observation caused by the interference and the like is achieved, A robust tracking method based on error observation combined identification is presented, which uses prediction and observation nearest neighbor, velocity and position consistency criterion to identify the error combination. In this paper, a new IMM algorithm based on segmented CV model is proposed. The new method establishes the motion model of maneuver personnel, and can describe the target's real motion state more accurately. The model set in the proposed method can adapt to the change of the target motion direction in time, and the speed direction of the tracking model set can be continuously adjusted until the steady state is reached when the mutation occurs, so that the state component estimation is closer to the target real speed direction. Simulation results show that the proposed algorithm can achieve better tracking performance than traditional maneuvering algorithm.
【学位授予单位】：国防科学技术大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN957.51
，
本文编号：2270436
本文链接：https://www.wllwen.com/kejilunwen/wltx/2270436.html
