ID 原文 译文
26465 针对应用于飞行器姿态确定中的乘性扩展卡尔曼滤波(multiplicative extended Kalman filter，MEKF)存在精度低、鲁棒性差的缺点，提出了一种基于多重次渐消因子的强跟踪无迹卡尔曼滤波(multiple fading factors strong tracking unscented Kalman filter，MSTUKF)算法。 For application in spacecraft attitude determine the multiplicative extended Kalman filtering (multiplicative extended Kalman filter, MEKF) exist the shortcomings of low precision, poor robustness, this paper proposes a fading factor based on the multiple times of strong tracking Kalman filtering is no trace (multiple fading factors strong tracking unscented Kalman filter, MSTUKF) algorithm.
26466 该滤波算法克服了单渐消因子对多变量跟踪能力差的局限性，通过引入两个多重次渐消因子对预测误差协方差阵进行调整，使得不同的滤波通道具有不同的调节能力，保证预测误差协方差阵的对称性，从而实现滤波算法强跟踪性。 The filtering algorithm to overcome the single fading factor for multivariate tracking capacity limitations, by introducing two multiple times fading factor to adjust prediction error covariance matrix, makes different filtering channels have different adjustment ability, ensure the prediction error covariance matrix of symmetry, so as to realize strong tracking filtering algorithm.
26467 仿真结果表明，MSTUKF的滤波精度和鲁棒性均明显优于MEKF，能够更好地满足工程应用对精度和鲁棒性的要求。 The simulation results show that MSTUKF filtering accuracy and robustness were significantly better than the MEKF, better able to satisfy the engineering application requirements for precision and robustness.
26468 高频地波雷达在一个相参积累时间内通常只能得到频域一次快拍，利用其直接进行波达方向估计性能较差。 High frequency ground wave radar in a coherent accumulation time usually can only get a fast frequency domain, the use of the direct for doa estimation performance is poorer.
26469 针对这一问题，在分析时、频域协方差矩阵特征分解后差异的基础上，在频域采用降维方法估计协方差矩阵。 In order to solve this problem, the analysis of characteristics of time and frequency domain covariance matrix decomposition after differences, on the basis of the frequency dimension reduction method is used to estimate the covariance matrix.
26470 根据频域目标信噪比相对较大的特点，利用最大特征值对应的信号特征向量构造原始的数据矩阵，并对该矩阵进行奇异值分解得到新的噪声子空间，进而构造出新的噪声特征向量，最后利用该噪声特征向量进行方位角估计。 According to the characteristics of frequency domain target SNR is opposite bigger, using the maximum characteristic value of the corresponding signal feature vector to construct the original data matrix, and the matrix singular value decomposition to get the new noise subspace, and then constructs the new noise eigenvectors, finally using the noise characteristic vector is used to estimate azimuth.
26471 仿真和实测数据分析验证了算法的有效性，相比降维Toeplitz算法和前后向空间平滑算法有着更高的分辨力和估计精度。 The effectiveness of the algorithm is demonstrated by the simulation and experimental data analysis, compared before and after the reduced-order Toeplitz algorithm and the spatial smoothing algorithm has higher resolution and estimation precision.
26472 提出了一种基于博弈论和效用理论的认知无线电功率控制算法。 Put forward a kind of based on game theory and utility theory of cognitive radio power control algorithm.
26473 算法将认知用户发射功率净效用函数表示为效用部分和代价部分之差。 Algorithm will cognitive user transmission power net utility function is expressed as the difference between the utility and cost part.
26474 对效用部分，用Sigmoid型效用函数替代David提出的幂指效用函数，并对替代误差进行补偿；对代价部分，提出了参考历史发射功率经验的代价函数，使得代价的变化更加平稳。 Part of utility, use the Sigmoid type put forward by the utility function to replace David exponential utility function, and compensate for alternative error;Transmission power for price part, put forward the reference history experience of cost function, make price changes more smoothly.