机载雷达降维空时自适应处理方法及杂波预滤波技术研究

发布时间：2018-04-14 15:37

本文选题：空时自适应处理 + 降维　；参考：《西安电子科技大学》2015年博士论文

【摘要】：具有灵活机动、覆盖范围大等优点的机载预警雷达在现代化的战争中发挥着重要作用。但是,由于载机快速移动,机载雷达面临着比地基雷达更加复杂的杂波环境。在空域和时域联合抑制杂波的空时自适应处理方法(Space-time adaptive processing,STAP)能有效抑制机载雷达杂波并检测到目标。但是,在实际应用中,全维STAP方法会产生计算量大和训练样本需求量高的问题。而这些问题进一步促进了一些次优的降维或者降秩STAP方法的发展。由于实际环境中缺少独立同分布的训练样本,因此到目前为止,还没有哪一种降维或者降秩STAP方法能够有效的应用到实际的机载雷达中。因此,考虑到实际环境中有限的训练样本数量以及雷达系统实时处理的需求,本文研究了具有计算量小、训练样本数需求量少等优点的降维STAP方法。在空时自适应处理之前对杂波进行预滤波处理可以有效的减少杂波自由度(Degrees of freedom,Do Fs),这样的预处理能提高后续STAP方法的性能,因为自适应处理器有了更多的自由度来检测目标。因此,我们根据杂波模型,研究了计算复杂度低、不受训练样本影响的机载雷达非自适应杂波预滤波方法。我们的工作主要包括以下几个方面:1.为了提高后多普勒自适应处理方法在大阵列条件下的杂波抑制和动目标检测能力,我们提出了一种基于空域分解的后多普勒自适应处理方法。首先将接收到的杂波和目标数据经过多普勒滤波,将滤波后的空域数据分解,然后将FA或者EFA的权系数表示成分离的形式,从而得到一双二次代价函数,然后利用循环迭代的思想求解权系数。实验表明该方法具有快速收敛性,在小样本、大阵列条件下该方法明显优于因子法和扩展因子法。2.由于机载MIMO雷达采用较小的天线规模即可形成很大的虚拟阵列孔径,并且具有高角度分辨率和强杂波抑制能力,近些年来受到了广大研究者和工程人员的极大关注。但是由于机载MIMO雷达系统自由度过高,相对于传统机载相控阵雷达STAP方法,MIMO-STAP将会需要更多的训练样本数和计算量。因此,我们研究了一种能大幅降低MIMO-STAP所需训练样本数和计算量的两级空域分解方法。该方法首先将接收到的杂波和目标数据经过多普勒滤波,将自适应权系数进行分解,使其变为几个向量的Kronecker乘积,然后利用循环迭代的思想求解自适应权。实验表明该方法具有快速收敛性,在小样本大阵列条件下该方法明显优于传统的后多普勒处理方法。3.机载雷达两维两脉冲对消器(Two-dimension pulse-to-pulse canceller,TDPC)能有效的沿着杂波迹抑制杂波而对目标信号没有影响,并且后续级联STAP方法,能有效提高STAP方法的动目标检测性能,不仅适用于正侧视机载雷达,也适用于非正侧视机载雷达。而且该方法仅仅利用雷达工作参数和载机速度等先验知识,滤波器系数可以提前计算好,具有计算量小不受训练样本影响的优点。但是实际中,受到各种因素的影响,估计的雷达参数有可能会和实际存在较大的误差。因此,考虑到参数误差的影响,我们研究了一种稳健的TDPC(Robust TDPC,RTDPC)方法。该方法考虑了实际中参数估计的误差,利用雷达参数、载机平台速度等信息设计杂波预滤波器以抑制大部分杂波,剩余的少量杂波可由发展成熟的降维空时自适应处理方法进行抑制。该方法增强了原TDPC方法在参数存在误差情况下的稳健性,进一步提高了TDPC方法的实用性。4.由于接收载机平台和发射载机平台之间复杂的几何配置,机载双基地雷达杂波呈现强烈的距离依赖性,空时自适应处理(STAP)所需要的杂波协方差矩阵的估计误差变大,这直接会导致STAP算法检测性能的恶化。针对这一问题,提出了一种机载双基地雷达杂波预滤波方法。该方法考虑了实际中载机速度的估计与真实速度的误差,利用雷达系统参数、载机平台速度等信息设计杂波预滤波器抑制大部分杂波,剩余的少量杂波可由发展成熟的空时自适应处理算法进行抑制,进而提高STAP算法的动目标检测能力。计算机仿真实验表明,该方法能有效的对几种典型几何配置下的机载双基地雷达杂波进行抑制,降低杂波自由度,后续级联能进一步改善STAP算法的动目标检测性能。5.在空时自适应处理中,相对于处理器自由度,杂波协方差矩阵是低秩的。根据这一原理,提出了一种利用低秩杂波子空间的杂波对消器(LRCC)以抑制地面强杂波,该方法利用相对较少的线性无关空时导向矢量构造出原杂波,然后再对消相邻脉冲间的杂波回波。该杂波对消器可以作为预滤波器和传统空时匹配或者空时自适应算法级联以增强后续动目标检测算法的性能。仿真结果表明,该方法在正侧视和非正侧视机载雷达中能有效抑制地杂波而对动目标信号没有影响。该杂波对消器作为预滤波器,可以提高后续空时匹配或空时自适应处理算法的动目标检测性能。
[Abstract]:With flexible, airborne early warning radar coverage range and other advantages play an important role in the modern war. However, because the plane moving fast, facing the airborne radar clutter environment is more complex than the ground-based radar. Adaptive processing method combined clutter suppression of air in space and time when (Space-time adaptive processing. STAP) can effectively inhibit the airborne radar clutter and detect targets. However, in practical application, full dimensional STAP method will have problems of large amount of calculation and training demand. And these problems to further promote the development of some sub optimal dimensionality reduction or reduced rank STAP method. Due to the lack of independent and identical distribution the actual environment of the training samples, so far, there is not any reduction or reduced rank STAP method can be effectively applied to airborne radar in practical. Therefore, considering the actual ring Exit in the limited number of training samples and the real-time processing of radar system, we have investigated the small amount of calculation and dimension reduction STAP method has the advantages of less demand. The number of training samples before the space-time adaptive processing for clutter filter can effectively reduce the clutter degree of freedom (Degrees of, freedom, Do Fs), this pretreatment could improve the performance of the subsequent STAP method, because the adaptive processor has more freedom to detect targets. Therefore, we according to the clutter model of airborne radar with low computational complexity, without training effect of non adaptive clutter filtering method. Our main work includes the following several aspects: 1. in order to improve after Doppler adaptive processing method in array under the condition of clutter suppression and moving target detection, we propose a domain decomposition based on Doppler Adaptive processing method. Firstly, the received clutter and target data through the Doppler filter, spatial data after filtering decomposition, and then the right coefficient of FA or EFA into the form of separation, so as to get a pair of two times the cost function, and then solving the weight coefficient by cyclic iterative method. Experimental results show that the method with fast convergence, in the small sample, a large array under the condition of the method is better than factor method and expansion factor method.2. the antenna airborne MIMO radar using small scale to form a virtual array aperture greatly, and has high angular resolution and strong ability to suppress clutter, in recent years has attracted great attention of many researchers and engineering the airborne MIMO radar system. But because of the degree of freedom is too high, compared with the traditional STAP method for airborne phased array radar, MIMO-STAP will need more training samples and calculation Volume. Therefore, we study a kind of MIMO-STAP can significantly reduce the required number of training samples and the amount of calculation of two spatial decomposition method. Firstly, the received clutter and target data through the Doppler filter, the adaptive weighting coefficients of decomposition, so that it becomes a product of several Kronecker vector, and then use the thought to solve the adaptive weight iteration. The experiments show that this method has fast convergence, in the method of Doppler treatment is obviously superior to the traditional method of small sample of the large array of.3. under the conditions of the airborne radar two dimensional two pulse canceller (Two-dimension pulse-to-pulse, canceller, TDPC) can effectively suppress the clutter clutter along the track and have no effect on the target signal, and the subsequent cascade STAP method, STAP method can effectively improve the target detection performance, not only for the Yu Zheng side looking airborne radar, is also suitable for the non side looking airborne radar Da. And this method only uses radar parameters and aircraft speed prior knowledge, the filter coefficients can be calculated in advance, has the advantages of small calculation amount is not affected by the influence of the training samples. But in practice, the influence of various factors, the radar parameter estimation may and the actual errors. Therefore, considering the influence of parameter errors, we study a robust TDPC (Robust TDPC RTDPC) method. This method considers the error estimation parameters, using radar parameters, aircraft speed and other information platform design clutter pre filter to suppress most clutter, a small amount of residual clutter may be reduced by the development of mature STAP method for suppression. The method enhances the original TDPC method under the condition of error robustness in parameter, further improve the practicability of.4. TDPC method for receiving aircraft platform And the launch vehicle complex geometric configuration platform between airborne bistatic radar clutter has a strong dependence on distance, space-time adaptive processing (STAP) error estimation of the clutter covariance matrix to change, this will directly lead to deterioration of the detection performance of the STAP algorithm. This paper proposes a airborne bistatic radar clutter filtering method. This method considers the error estimation of actual load machine speed and actual speed, the use of radar system parameters, aircraft speed and other information platform design clutter pre filter to suppress most clutter, a small amount of residual clutter may be developed by space-time adaptive processing algorithm inhibition, and improve the target detection ability of STAP algorithm. Simulation results show that this method can be effective for some typical configurations of the airborne bistatic radar clutter suppression, reducing clutter The wave of freedom, will further improve the subsequent level STAP algorithm.5. moving target detection performance in space-time adaptive processing, relative to processor freedom, the clutter covariance matrix is low rank. According to this principle, proposes the use of a low rank clutter subspace clutter canceller (LRCC) to suppression of the strong ground clutter, the method using linear relatively independent space-time steering vector to construct the original clutter cancellation, then adjacent pulses between clutter. The clutter canceller can be used as a pre filter and the traditional space-time matching or space-time adaptive algorithm to enhance the performance of target follow-up cascade detection algorithm. The simulation results show that this method is in the side and non side looking airborne radar can effectively suppress the ground clutter and moving target signal has no effect. The clutter canceller as a pre filter, can increase the space time, or empty The dynamic target detection performance of time adaptive processing algorithm.

【学位授予单位】：西安电子科技大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：TN957.51;TN713

【相似文献】