ID 原文 译文
11794 针对现有多个弱小目标检测前跟踪(track-before-detect,TBD)算法存在的跟踪精度低,算法复杂度高等问题,提出一种新的基于概率假设密度(probability hypothesis density,PHD)的TBD算法。 In view of the existing multiple weak target detection before tracking (track - before - detect, TBD) algorithm is low tracking precision, the algorithm complexity is high, put forward a new hypothesis is based on probability density (aim-listed probability content, density, PHD) TBD algorithm.
11795 所提算法通过高斯粒子滤波对PHD中的各高斯项进行递归运算、进行多帧能量累积,并提取高斯项的均值为目标的状态,达到检测与跟踪多个弱小目标的目的。 ‭Proposed algorithm by gaussian particle filter is applied to the each gaussian PHD to recursive computation, to frame more energy accumulation, and extract the gaussian item as the target state, to achieve the purpose of detecting and tracking multiple weak small targets.
11796 算法在随机集滤波框架下完成未知数目的多个弱小目标跟踪。 ‭Algorithm in random set filter framework to complete multiple weak target tracking unknown purpose.
11797 不仅充分利用粒子滤波的非线性估计能力,同时避免了传统算法利用模糊聚类进行目标状态提取所带来的跟踪精度低等问题。 Not only make full use of the nonlinear estimation of particle filter ability, at the same time to avoid the traditional algorithm using fuzzy clustering to extract target state caused by the problem of low accuracy of tracking.
11798 仿真结果表明,所提算法与传统方法相比,在降低算法复杂度的同时,对多个红外弱小目标具有更加良好的实时检测和跟踪性能。 ‭The simulation results show that the proposed algorithm is compared with the traditional method, at the same time, reduce the complexity of the algorithm for multiple infrared weak small targets with more good real-time detection and tracking performance. ‭
11799 针对复杂环境下自主水下航行器(autonomous underwater vehicle,AUV)组合导航系统中存在噪声不确定或者易发生变化的情况,提出一种贝叶斯网络增强型交互式多模型(interactive multiple model filter based on Bayesian network,BN-IMM)滤波算法。 In view of the complex environment of autonomous underwater vehicle (autonomous underwater vehicle, the AUV) integrated navigation system exist in the noise uncertainty or easy to change, in this paper, a Bayesian network enhanced interactive multiple model (interactive multiple model filter -based on Bayesian network, BN - IMM) filtering algorithm.
11800 该算法在多模型估计基础上,引入特征变量,并根据变量与系统模型之间存在的因果关系建立贝叶斯网络; ‭ The algorithm based on multiple model estimation, the introduction of characteristic variables, and according to the causal relationship between variables and system model are established the bayesian network;
11801 利用贝叶斯网络参数修正多模型估计中的模型切换概率,能够降低多模型算法中真实模式识别对先验知识的依赖性。 ‭Using bayesian network parameter correction model to estimate the model switching probability, can reduce the real pattern recognition in the multiple model algorithm for the dependence of the prior knowledge.
11802 该算法能够解决交互式多模型(interactive multiple model,IMM)算法中模型转换存在滞后、模型概率易发生跳变等问题,增强多模型算法的自适应能力。 This algorithm can solve the interactive multiple model (interactive multiple model, the IMM) algorithm of model transformation is lagged, prone to jump model probability problems, enhance the multiple model adaptive ability of the algorithm.
11803 以陀螺和加速度计的输出作为特征变量建立贝叶斯网络,对AUV组合导航系统进行仿真。 Gyroscope and accelerometer output as characteristic variables to establish the bayesian network, the AUV integrated navigation system.