ID	原文	译文
57608	为了检测输气管道阀门泄漏，对改进 AlexNet 网络结构进行了研究，提出了基于改进卷积神经网络( CNN) 的阀门泄漏超声信号识别方法.	In order to detect valve leakage in gas pipelines，an improved AlexNet network architecture is studied，an ultrasonic signal recognition method for valve leakage based on an improved convolutional neural network ( CNN) is proposed.
57609	针对泄漏信号短时稳定的窄带线谱特征，从图像邻域信息密度角度出发，将卷积核形状由图像识别领域通常使用的“正方形”改进为“扁横状”.	Due to short-term and narrow-band line spectrum features of the leakage signals，the“square”convolution kernel，commonly used in image recognition，is changed to “flat”based on the perspective of image neighborhood information density.
57610	同时，对 AlexNet 层数进行优化，重新确定卷积核和全连接层神经元数目，并选择小尺寸卷积核，在减少参数量的同时增加网络容量和模型复杂度，防止模型出现过拟合.	At the same time，the Alex- Net layers are optimized，the number of convolution kernel and neurons in the fully connected layers are re-determined，and the small-scale convolution kernel is selected to increase the network capacity and model complexity while reducing the number of parameters to prevent model overfitting.
57611	分别建立二分类和不同泄漏量下的多分类模型，通过输气管道实验平台采集阀门泄漏数据集，生成对应时频图样本，包括不同阀门开度、不同管道压力下的泄漏及背景声信号.	The two-class and multi-class models with different leakages are established respectively，and the data set is collected through experiments to generate corresponding time-frequency diagram samples as well，including leakage signals at different valve openings and pipeline pressures and background acoustic signals.
57612	结果表明，对比传统的 CNN 分类模型，改进 CNN 分类模型在测试集上取得了更高的识别性能.	It is shown that the improved CNN classifier achieves higher recognition performance on the test set than the traditional CNN classifier.
57613	针对无线传感器网络中能耗不均衡问题，提出了一种基于改进萤火虫算法优化反向传播神经网络的非均匀分簇路由协议.	Aiming at solving the problem of uneven energy consumption in wireless sensor networks ( WSNs) ，an uneven clustering routing protocol based on the improved firefly algorithm optimized back propagation( BP) neural network ( IFABPUC) is proposed.
57614	通过在萤火虫算法中引进权重因子并增加 4 个评价指标，来平衡簇内负载和减少簇间的通信距离.	To balance the intra-cluster load and reduce the inter-cluster communication distances，a weighting factor which takes into account four more evaluation indexes than the conventional firefly algorithm is embedded in the improved firefly algorithm.
57615	结合 BP 神经网络，优化路径选择和簇首选举方式，达到最佳成簇效果.	To achieve the best clustering，BP neural network is combined to optimize the way to path selection and cluster head election.
57616	仿真结果表明，改进萤火虫算法优化 BP 神经网络的非均匀分簇路由协议能有效延长网络生命周期，节省能量，并均衡能耗.	Simulations show that IFABPUC can effectively extend the lifecycle of networks，save energy and balance energy consumption.
57617	针对毫微微基站( FBS) 在不同时间段用户数量的差异，研究了最大化下行总信息量的功率分配问题.不同时间段包含忙时和闲时 2 个阶段，忙时用户数量较多，闲时用户数量较少.通过部署一个无人机携带的微微基站( PBS-UAV) 为多个 FBSs 闲时的用户提供服务.FBS 和 PBS-UAV 都具有能量收集功能.在 FBS 忙时，FBS 和 PBS- UAV 同时从宏基站收集能量，并且向用户发送数据.FBS 闲时，由 PBS-UAV 接替多个 FBSs，与用户进行下行通信.将功率分配问题建模为最优化问题，以最大化 FBS 和 PBS-UAV 的下行信息量为目标，同时满足 FBS 和 PBS-UAV能量消耗及发射功率的约束条件.由于建立的最优化问题是凸优化问题，可通过引入增广拉格朗日乘子法获得最优解.仿真结果表明，与 PBS-UAV 参与的等功率及部分功率固定的方法相比，所提出的方法在总信息量方面有不同程度的增加.	Aiming at the difference of users in different time periods for the femto base station ( FBS) ，the power allocation problem of maximizing the total downlink information is investigated. The different time periods include busy time and spare time. There are more users in the busy time and less users in the spare time. By deploying a pico base station carried by unmanned aerial vehicle ( PBS-UAV) ，it provides services for users of multiple FBSs in spare time. Both the FBS and PBS-UAV have energy harvesting function. During the busy time，the FBS and PBS-UAV simultaneously harvest energy from the macro base station，and FBSs transmit data to users. During the spare time，multiple FBSs are replaced by the PBS- UAV to communicate with users in downlink. The power allocation problem is modeled as an optimization problem. The objective is to maximize the amount of downlink information of FBSs and PBS-UAV while satisfying the constraints of FBS and PBS-UAV energy consumption and transmission power. Because the formulated optimization problem is a convex optimization problem，the optimal solution is obtained by u- sing an augmented Lagrange multiplier method. Simulations show that compared with the equal power method and partial fixed power method with PBS-UAV，the proposed method has an increase in terms of total information to different degrees.