ID	原文	译文
6674	该算法运用集员方法和共轭梯度,避免重复计算,得到有效的权重向量,保证良好的收敛性能。	Using the set membership method and conjugate gradient, the algorithm to avoid double-counting, get effective weight vector, ensure good convergence performance.
6675	又对算法进行计算复杂度和收敛性能分析。	And the algorithm computational complexity and convergence performance analysis.
6676	仿真结果表明,与其他传统算法相比,该算法在保证良好的收敛性能的同时,大大减少了计算复杂度。	The simulation results show that compared with other traditional algorithm, this algorithm in ensuring good convergent performance at the same time, greatly reduce the computational complexity.
6677	近年来通用航空发展迅速,由此引发的安全问题日益引起重视。	General aviation has developed rapidly in recent years, the resulting security problem is becoming more and more attention.
6678	但是由于通用航空器种类繁多,样本之间的差异性较大,传统的统计分析技术在通用航空风险预测方面就显得无能为力。	But because there are so many types of general aircraft, the differences between samples is larger, the traditional statistical analysis techniques in the aspect of general aviation risk prediction is powerless.
6679	提出一种基于稀疏降噪自编码神经网络的通用航空风险预测方法,稀疏降噪自编码模型(sparse de-noising auto-encoder,SDAE)可以学习相对稀疏简明的数据特征,更好地表达输入数据。	In this paper, a neural network based on sparse noise reduction since the coding of the general aviation risk prediction method, the sparse noise reduction since the coding model (sparse DE - noising auto - encoder, SDAE) can learn relatively sparse data characteristics of concise, better express the input data.
6680	利用收集到从2012年1月至2015年12月共48个月各个不同事件发生原因的其他不安全事件总数量,建立民航事故征候万架次率的神经网络预测模型,将其他不安全事件的发生与事故征候联系起来。	Using collected from January 2012 to December 2015, a total of 48 months various events for the total number of other unsafe events, vehicles set up signs of civil aviation accident rate of neural network prediction model, the other is not linked to the occurrence of security incidents and accidents signs.
6681	通过实例说明,构建的SDAE模型可以根据输入的其他不安全事件的数量对当月的事故征候万架次率做出较为准确的预测。	With practical examples, can build SDAE model according to the number of input other unsafe events for the month of vehicles accident symptom rate to make more accurate predictions.
6682	无人机"精准农业"任务要求无人机能够响应环境信息进行自主决策,从而完成喷绘任务。	Drones "precision agriculture" mission requirements to respond to environmental information decisions independently, to complete the print tasks.
6683	首先,研究了面向作业要求的空间属性设计规则,根据精准作业条件属性因子的值域分布特点及其样本密度,提出了一种基于高斯分布密度的反正切属性精细离散方法。	First, the requirements of spatial attribute oriented design rules, according to the range of values of the precise operation condition attribute factor distribution characteristics and the sample density, this paper proposes a fine the arctangent attributes based on gaussian distribution density of discrete method.