ID 原文 译文
25195 其次，在引入漂移系数更新机制的基础上建立设备退化状态方程，并采用 EKF 算法同步更新设备退化状态与漂移系数; Next, the degradation state equation of the equipment is established based on the drift coefficient update mechanism, and the EKF algorithm is used to update the degradation status and drift coefficient．
25196 然后，采用 EM-EKF 算法实现对退化模型参数的自适应估计; And then, the EM-EKF algorithm is used to adaptively estimate the parameters of the degradation model．
25197 最后，基于全概率公式，推导出设备剩余寿命的概率密度函数。 Finally, based on the full probability formula, the probability density function (PDF) of ＲUL is derived．
25198 通过对单台微机械陀螺仪实测数据进行分析，验证了本文所提方法具有更好的模型拟合性与预测准确性。 By analyzing the measured data of a single micromechanical gyroscope, it is verified that the proposed method has better model fitting and prediction accuracy．
25199 以 x86 + GPU 为代表的当前主流 AI 计算平台，受限于功耗、体积、带宽、环境适应性等因素，无法适用于物端及边缘智能计算场景。 The existing artificial intelligent (AI) computing platform represented by x86 + GPU，limited by power consumption, dimension, bandwidth, environmental adaptability, and other factors, cannot be well adapted to the things and edge intelligent computing scenarios．
25200 提出并研究了一种基于 AＲM + DLP + SＲIO 的嵌入式智能计算系统，从 AI 算力、能效比、IO 带宽三个方面分析了所提嵌入式智能计算系统的设计思路和技术优势，并实验验证了该系统的功能及性能指标。 We proposed an embedded AI computing system based on AＲM (Advanced ＲISC Machine)+ DLP (Deep Learning Processor)+ SＲIO (Serial ＲapidIO), and elaborated the design methods and technical advantages． In study, three aspects of the system were dissertated: AI computing performance, power efficiency, and IO bandwidth, and the function and performance of the system were verified by experiments．
25201 实验结果表明：基 于 AＲM + DLP + SＲIO 的 嵌 入 式 智 能 计 算 系 统 AI 峰 值 算 力 达 到 114. 9TOPS，能 效 比 达 到1. 03TFLOPS/W，IO 带宽达到 20Gbps。 The results show that the peak performance of the embedded AI computing system based on AＲM + DLP + SＲIO is up to 114. 9TOPS, the energy efficiency is up to1. 03TFLOPS/W, and the IO bandwidth is up to 20Gbps．
25202 在智能计算系统领域，其能效比优于国内其它已知同类板卡或系统，嵌入式环境适应能力优于传统台式机和服务器，可作为物端及边缘环境下 AI 计算任务的通用硬件加速平台。 In the field of AI computing systems, its energy efficiency is better than other similar boards or systems in China, and its embedded environmental adaptability is better than that of traditional desktops and servers, so it can provide a general hardware acceleration platform for AI computing tasks in things and edge computing scenarios．
25203 WiNoC(Wireless Network-on-Chip)中的无线路由器面临着比传统片上路由器更加严峻的拥塞问题，平衡有线/无线链路负载是当前无线片上网络的研究热点之一。 Wireless routers in wireless network on chip confronts more severe congestion problem，so balancing wired/wireless loading has become research focus in WiNoC recently．
25204 为此本文提出并设计了一种基于优先级的交叉开关仲裁方案 PbSA(Priority based Switch allocator)，其将优先级更高的无线数据包优先路由至无线路由器;结合 PbSA 提出了拥塞感知的路由算法 CAＲA(Congestion-Aware Ｒouting Algorithm)，该算法有效平衡有线 /无线链路负载且避免死锁，提高了数据包在网络中的路由效率。 We propose a priority-based switch arbitration scheme, in which data packets more suitable for transmission through wireless channels are routed to wireless routers; We propose a CAＲA (Con-gestion-Aware Ｒouting Algorithm) combined with PbSA, which efficiently balance wired/wireless loading and simultaneously avoid deadlock, improving data routing efficiency.