摘要：【中文】本发明公开了一种QAM相干光通信系统中基于矩的精度增强的OSNR监测方法。本发明QAM相干光通信系统中基于矩的精度增强的OSNR监测方法，其特征在于，包括：使用时钟恢复的信号作为输入；将发送信号写为x_n＝R_nе^jθn，其中R_n和θ_n分别是QAM信号的幅度和相位，恢复信号表示为：其中，s是信号功率因数，n是噪声功率因数，z_k是均值为零的高斯白噪声；对时钟恢复后的信号进行平均能量归一化。与传统的基于矩的方法相比，所提出的精度增强的基于矩的方法可以将估计误差降低到0.4dB以下。 【EN】The invention discloses an OSNR monitoring method based on moment precision enhancement in a QAM coherent optical communication system. The invention discloses an OSNR monitoring method based on moment precision enhancement in a QAM coherent optical communication system, which is characterized by comprising the following steps: using the clock recovered signal as an input; writing a Transmission Signal as x_n＝R_nе^jθnWherein R is_nAnd theta_nRespectively, the amplitude and phase of the QAM signal, the recovered signal is represented as:where s is the signal power factor, n is the noise power factor, z_kIs white gaussian noise with a mean value of zero; and carrying out average energy normalization on the signals after clock recovery. The proposed enhanced-precision moment-based method can reduce the estimation error to below 0.4dB compared to the conventional moment-based method.

摘要：【中文】本发明提供一种机器人三维运动空间动作模仿学习方法和系统，涉及人工智能和机器人技术领域。本发明设计了一种基于串并联叠层BP神经网络的机器人三维空间动作模仿学习方法，将模仿学习机制应用于机器人学习系统，在模仿学习机制的流程框架下，通过将机械臂产生的示教信息传递至表示运动策略的串并联叠层BP神经网络模型来进行训练学习。通过学习运动的状态特征矩阵集合和运动的动作特征矩阵集合之间的对应关系，不仅可以再现示教动作，还能够对动作行为进行泛化，在面对不同的任务时，不需要分别进行动作规划，具有较高的智能性。 【EN】The invention provides a method and a system for simulating and learning three-dimensional motion space actions of a robot, and relates to the technical field of artificial intelligence and robots. The invention designs a robot three-dimensional space action simulation learning method based on a series-parallel laminated BP neural network, wherein a simulation learning mechanism is applied to a robot learning system, and teaching information generated by a mechanical arm is transmitted to a series-parallel laminated BP neural network model representing a motion strategy to train and learn under a flow frame of the simulation learning mechanism. By learning the corresponding relation between the state characteristic matrix set of the movement and the action characteristic matrix set of the movement, the teaching action can be reproduced, action behaviors can be generalized, action planning is not needed to be performed when different tasks are faced, and the intelligence is high.

摘要：【中文】本发明公开了一种基于多视角对称非负矩阵分解的跨模态信息检索方法，包括多视角对称非负矩阵分解模型、求解多视角对称非负矩阵分解模型和获得新文本或图像在子空间中的表示三个步骤，通过利用Multi‑view SNMF目标函数O_bj、低维子空间表示H^*和聚类指示矩阵H⁽ⁱ⁾，针对多个模态特征信息的联合建模，通过将不同模态的数据映射到相关子空间，进而建立跨越资源形态差异的联系，以实现信息在不同模态空间中的转换和存取；相较于CCA、PLSR等多元统计方法和以SM为代表的语义匹配方法，本算法在多个指标上具有显著的有益效果，并且本方法具有更好的灵活性和普适性，易拓展到具有非线性关系的复杂应用和任务中。 【EN】The invention discloses a cross-modal information retrieval method based on Multi-view symmetric nonnegative matrix factorization, which comprises three steps of a Multi-view symmetric nonnegative matrix factorization model, solving the Multi-view symmetric nonnegative matrix factorization model and obtaining the representation of a new text or an image in a subspace, and a Multi-view SNMF target function O is utilized_bjLow dimensional subspace representation H^*And a clustering indicator matrix H⁽ⁱ⁾Aiming at the joint modeling of the characteristic information of a plurality of modes, the data of different modes are mapped to related subspaces, and then the connection spanning the resource form difference is established, so that the conversion and the access of the information in different mode spaces are realized; compared with a multivariate statistical method such as CCA (common clear channel), PLSR (partial least squares) and the like and a semantic matching method represented by SM (short message rule), the algorithm has remarkable beneficial effects on multiple indexesAnd the method has better flexibility and universality and is easy to expand into complex application and tasks with nonlinear relation.

摘要：【中文】本发明公开了一种扫描仪生产用贴标装置，其不仅可以实现自动的对标签进行供应，而且本发明通过设置视觉检测与定位机构和角度调节机构，视觉检测与定位机构能够对待贴标的扫描仪的位置进行检测，以便利用所述角度调节机构自动调节好贴标位置，实现对标签的精准位置调节，提高调节的准确度，本发明可以有效的实现自动化贴标，并能够实现与扫描仪输送机构对接，实现流水线作业，便于后续的生产与检测的自动化，本发明的标签控制机构还可以实现标签的自动控制，实现标签进给与锁紧的自动控制，提高控制效率。 【EN】The invention discloses a labeling device for scanner production, which can automatically supply labels, and is provided with a visual detection and positioning mechanism and an angle adjusting mechanism, wherein the visual detection and positioning mechanism can detect the position of a scanner to be labeled, so that the angle adjusting mechanism can be used for automatically adjusting the labeling position, the accurate position adjustment of the labels is realized, and the adjustment accuracy is improved.

摘要：【中文】本发明提出基于平面多机驱动倍频同步振动粉磨装置及参数确定方法，装置包括机座、弹簧系统、偏心激振系统、圆柱形研磨室、进料口、出料口、磨介、物料。通过强制同步或振动自同步或控制同步使两激振器的偏心块实现转速倍频关系，对于超共振状态和亚共振状态，公共周期内考虑倍频比的相位差值分别稳定在‑90°～90°和90°～270°，此时使研磨室的磨介和物料出现非常见的单个椭圆运动轨迹，可实现双椭圆、多椭圆、混沌运动等不同的运动轨迹，这样能够使介质避免单纯的整体运动，使介质之间发生更多的相对摩擦，避免出现较大的低能区，达到更优良的粉磨作用。此外，由于这类磨机采用了振动同步原理，磨介和物料受到的力可远大于一个重力加速度，这使粉磨效率明显提高。 【EN】The invention provides a plane-based multi-machine-driven frequency-doubling synchronous vibration grinding device and a parameter determination method. The eccentric blocks of the two vibration exciters realize the relationship of rotating speed and frequency multiplication by forced synchronization or vibration self-synchronization or control synchronization, phase difference values of frequency multiplication ratio are considered to be stabilized at-90 degrees and 90-270 degrees in a common period for a super-resonance state and a sub-resonance state respectively, the grinding medium and the material of the grinding chamber have unusual single elliptic motion tracks at the moment, and different motion tracks of double ellipses, multiple ellipses, chaotic motion and the like can be realized, so that the medium can avoid simple overall motion, more relative friction is generated between the medium, a larger low-energy area is avoided, and a better grinding effect is achieved. In addition, because the mill adopts the vibration synchronization principle, the force applied to the grinding medium and the material can be far greater than a gravity acceleration, so that the grinding efficiency is obviously improved.

摘要：【中文】本发明提供了一种车牌信息的匹配方法及系统、存储介质、电子装置，其中，所述方法包括：在通过车辆检测器检测到车辆进入到目标区域中的情况下，将触发信号分别发送至图像抓拍装置和RSU，其中，所述触发信号中包括：所述车辆的坐标信息；响应所述触发信号，指示所述图像抓拍装置对所述坐标信息的位置进行抓拍，并将抓拍得到的抓拍信息发送至RSU，其中，所述抓拍信息包括：所述车辆的车牌信息；响应所述触发信号，将所述抓拍信息中的车牌信息和所述车辆的OBU信息中的车牌信息进行匹配，采用上述技术方案，以至少解决相关技术中，由于车辆检测器、相机、RSU之间比较独立，协作能力不强而导致的自由流收费系统中稽查系统捕获、识别准确率低等问题。 【EN】The invention provides a license plate information matching method and system, a storage medium and an electronic device, wherein the method comprises the following steps: under the condition that a vehicle detector detects that the vehicle enters a target area, trigger signals are respectively sent to an image capturing device and an RSU, wherein the trigger signals comprise: coordinate information of the vehicle; responding to the trigger signal, indicating the image capturing device to capture the position of the coordinate information, and sending capturing information obtained by capturing to the RSU, wherein the capturing information comprises: license plate information of the vehicle; and responding to the trigger signal, and matching the license plate information in the snapshot information with the license plate information in the OBU information of the vehicle.

摘要：【中文】本发明提供了一种车辆超限检测系统，包括：车辆检测装置、ETC门架天线、车载标签、后台服务器，其中，所述车辆检测装置，检测目标车辆得到车辆检测信息，将所述车辆检测信息发送给车载标签；所述车载标签，在确定所述目标车辆超限时生成超限检测信息，并将所述超限检测信息通过所述ETC门架天线发送给所述后台服务器；所述后台服务器，用于根据所述超限检测信息对所述目标车辆进行处理，可以解决相关技术中需要视频采集设备提供辅助信息形成完整的证据链才能检测出高速上车辆超载的问题，无需视频采集设备辅助，结合高速动态称重仪器，便可实现高速场景下超限超载车辆的检测。 【EN】The invention provides a vehicle overrun detection system, comprising: the system comprises a vehicle detection device, an ETC portal antenna, a vehicle-mounted tag and a background server, wherein the vehicle detection device detects a target vehicle to obtain vehicle detection information and sends the vehicle detection information to the vehicle-mounted tag; the vehicle-mounted tag generates overrun detection information when the target vehicle is determined to be overrun, and sends the overrun detection information to the background server through the ETC portal antenna; the background server is used for processing the target vehicle according to the overrun detection information, can solve the problem that overload of the vehicle on the high-speed vehicle can be detected only by forming a complete evidence chain by the aid of auxiliary information provided by video acquisition equipment in the related technology, does not need the assistance of the video acquisition equipment, and can realize detection of the overrun overload vehicle in a high-speed scene by combining with a high-speed dynamic weighing instrument.

摘要：【中文】本发明涉及车辆耐久性试验技术领域，具体涉及用户与试验场关联时用户道路最少采集里程计算方法，获取用户道路耐久性载荷谱采集；确定分析对象，获取伪损伤密度；利用非参数检验方法对分析对象所受到的载荷谱进行正态性检验；基于概率统计的理论，从总体服从正态分布的目标里程下伪损伤密度中进行抽样，获取含有均值和方差的抽样分布；定义抽样样本的均值与总体均值的相对偏差；指定采集样本的置信度；获取最小样本量计算公式；最终确定最小采集里程。本发明避免采集数据量过大，同时，提高耐久性试验获得车辆损伤信息的准确性和代表性。 【EN】The invention relates to the technical field of vehicle durability tests, in particular to a method for calculating the minimum collected mileage of a user road when a user is associated with a test field, and load spectrum collection of the durability of the user road is obtained; determining an analysis object, and acquiring a pseudo damage density; carrying out normality test on a load spectrum subjected to an analysis object by using a nonparametric test method; based on the theory of probability statistics, sampling is carried out from the pseudo-damage density under the target mileage generally obeying normal distribution, and sampling distribution containing a mean value and a variance is obtained; defining the relative deviation of the mean value of the sampled samples from the overall mean value; specifying a confidence level of the collected sample; obtaining a minimum sample amount calculation formula; and finally determining the minimum collected mileage. The method avoids overlarge data collection amount, and improves the accuracy and the representativeness of the vehicle damage information obtained by the durability test.

摘要：【中文】本发明提供了一种信息的获取方法和装置、自由流收费系统及存储介质，其中，该方法包括：接收车辆检测子系统发送的第一触发信号，其中，第一触发信号为车辆检测子系统在检测到有车辆通过扫描断面之后发送的触发信号；响应第一触发信号，对与扫描断面对应的目标区域进行扫描；在从目标区域中扫描到OBU的情况下，通过路侧单元RSU与OBU进行跟踪交易，得到交易信息。通过本发明，解决了相关技术的自由流收费系统存在由于各个子系统间的相互协作能力不强，导致交易成功率低的问题。 【EN】The invention provides an information acquisition method and device, a free flow charging system and a storage medium, wherein the method comprises the following steps: receiving a first trigger signal sent by a vehicle detection subsystem, wherein the first trigger signal is the trigger signal sent by the vehicle detection subsystem after a vehicle is detected to pass through a scanning section; responding to the first trigger signal, and scanning a target area corresponding to the scanning section; and under the condition that the OBU is scanned from the target area, tracking transaction is carried out between the RSU and the OBU, and transaction information is obtained. The invention solves the problem that the free flow charging system in the related technology has low transaction success rate due to weak mutual cooperation capability among subsystems.

摘要：【中文】本发明提供了一种治疗痛风的药酒，具有降尿酸、抗炎、消肿止痛等功效。一种治疗痛风的药酒，由以下重量配比的原料制成：大果飞蛾藤15‑20份、麻黄1‑2份、桂枝1‑2份和当归1‑2份。该药酒的制备方法为：将上述原料药洗净、粉碎，置瓷坛或玻璃坛内，加入40‑70度的白酒90‑130份，密封，34‑50天后即得。 【EN】The invention provides medicinal liquor for treating gout, which has the effects of reducing uric acid, resisting inflammation, relieving swelling and pain and the like. A medicinal liquor for treating gout is prepared from the following raw materials in parts by weight: 15-20 parts of plutella xylostella, 1-2 parts of ephedra, 1-2 parts of cassia twig and 1-2 parts of angelica. The preparation method of the medicinal liquor comprises the following steps: cleaning the above materials, pulverizing, placing in a porcelain jar or glass jar, adding 90-130 parts of 40-70% (v/v) Chinese liquor, sealing, and standing for 34-50 days.