ID	原文	译文
6384	干扰检测和参数估计性能受干扰强度、干扰带宽以及压缩率变化的影响。	Interference detection and parameter estimation performance disturbance intensity, the interference of bandwidth and changes in the compression ratio.
6385	干扰强度越强、干扰带宽越小、压缩率越大,干扰检测和参数估计效果越好。	The stronger the intensity of interference, interference bandwidth is smaller, the compression ratio, the greater the effect of interference detection and parameter estimation, the better.
6386	针对传统效能评估中指标权重相对固定的不足,基于变权理论和灰靶理论,提出了一种动态评估方法,解决了指标评价值为精确实数、区间数和三角模糊数的混合型多属性指挥控制系统效能评估问题。	Aimed at the shortage of the traditional effectiveness evaluation index weight is relatively fixed, based on the variable weight theory and grey target theory, this paper proposes a dynamic evaluation method, solves the index value of accurate real number, interval number and triangular fuzzy number of hybrid multi-attribute command and control system effectiveness evaluation problem.
6387	首先,引入指标作用度的概念,解决混合型指标评价值相对重要度度量问题,采用改进序关系法求解指标初始常权;	First of all, introduce the concept of index effect degree, solve the problem of relative important degree measure the hybrid index evaluation, with the improved sequence relations act to solve the initial weights indicators;
6388	其次,以指标评价值与正、负靶心距离的相对值作为变量,定义了一种新的优势度函数用于刻画指标的优劣性,并基于变权理论构造局部状态变权函数对初始常权进行动态修正;	Secondly, in order to index value and relative value of positive and negative the bull 's-eye distance as the variable, defines a new kind of advantage degree function is used to depict index and inferiority, and based on the variable weight theory construct the local state variable weight function of initial weights dynamic correction;
6389	最后,以评估对象变权投影靶心距作为度量标准,采用投影灰靶方法确定了评估对象综合效能及排序。	Finally, to evaluate the object variable weight as metrics, projection distance to target by projection gray target method to determine the comprehensive efficiency evaluation object and sorting.
6390	实例分析验证了所提方法的可行性。	The example analysis verify the feasibility of the proposed method.
6391	针对超短波通信中特定信号的识别问题,提出一种将时频谱图和卷积神经网络相结合的超短波特定信号识别方法。	For a particular signal identification in ultrashort wave communication problems, a spectrum and convolution is put forward combined ultrashort wave specific signal recognition method of neural network.
6392	该方法首先对特定信号进行短时傅里叶变换得到时频谱图,然后使用时频谱图对改进的卷积神经网络模型进行训练,最后测试网络模型,实现超短波特定信号识别。	This method firstly the specific signal short-time Fourier transform to get the spectrum diagram, then use the spectrum diagram of improved convolution neural network model for training, the final test network model, realize the ultrashort wave specific signal recognition.
6393	实验结果表明,该方法对特定信号的识别率能达到98%,在信噪比为0dB时仍能达到97%的识别率,并且在混叠50%时识别率达到了90%。	Experimental results show that this method is of certain signal recognition rate can reach 98%, when the signal-to-noise ratio of 0 db still can achieve the recognition rate of 97%, and 50% in the aliasing recognition rate reached 90%.