摘要：【中文】本发明涉及一种基于二维异质薄膜的谐振频率自适应调控系统，其包括二维异质薄膜、支撑衬底、压电陶瓷驱动器、外部可调光源、振动探测器以及信号调理电路，异质薄膜上的纳米金属颗粒能够产生等离共振效应，二维异质薄膜置于支撑衬底之上，支撑衬底与压电陶瓷驱动器实现固定连接，外部可调光源对二维异质薄膜进行照射，压电陶瓷驱动器驱动二维异质薄膜振动，振动探测器感知二维异质薄膜的振动并产生相应电信号，信号调理电路将振动探测器生成的电信号进行处理并转换成驱动控制信号，从而实现对压电陶瓷驱动器的振动控制。本发明具有非接触、大范围和自适应特点，同时能实现灵活调控和高精度探测。 【EN】The invention relates to a resonant frequency self-adaptive control system based on a two-dimensional heterogeneous film, which comprises the two-dimensional heterogeneous film, a supporting substrate, a piezoelectric ceramic driver, an external adjustable light source, a vibration detector and a signal conditioning circuit, wherein nano metal particles on the heterogeneous film can generate a plasma resonance effect, the two-dimensional heterogeneous film is arranged on the supporting substrate, the supporting substrate is fixedly connected with the piezoelectric ceramic driver, the external adjustable light source irradiates the two-dimensional heterogeneous film, the piezoelectric ceramic driver drives the two-dimensional heterogeneous film to vibrate, the vibration detector senses the vibration of the two-dimensional heterogeneous film and generates corresponding electric signals, and the signal conditioning circuit processes the electric signals generated by the vibration detector and converts the electric signals into driving control signals, so that the vibration control of the piezoelectric ceramic driver is realized. The invention has the characteristics of non-contact, large range and self-adaption, and can realize flexible regulation and control and high-precision detection.

摘要：【中文】本发明公开了一种测量可逆式模型机组功率损失的方法及其使用的装置，包括连接可逆式机组的测功力臂、负荷传感器、标定装置以及底座，负荷传感器上设穿过负荷传感器受力孔的顶针杆，并连接底座；标定装置为两组，包括连接杆和连接螺杆，连接杆通过活动件可转动地设于底座上，连接杆的两端分别设砝码盘和连接螺杆，连接螺杆分别位于测功力臂的两个不同力矩方向上，并通过测功力臂、顶针杆的端部作用于负荷传感器上，适用于顺时针或逆时针运行可逆式机组时的功率损失的测量，很好地适应了可逆式机组输出功率损失的特点及位置要求，具有双向原位标定传感器和测功的功能，相比单向测功方式可靠性更高，结构简单，安装、操作、维护更为方便。 【EN】The invention discloses a method for measuring power loss of a reversible model unit and a device used by the method, and the device comprises a power measuring arm, a load sensor, a calibration device and a base which are connected with the reversible model unit, wherein the load sensor is provided with a thimble rod which penetrates through a stress hole of the load sensor and is connected with the base; the calibration devices are two groups and comprise connecting rods and connecting screw rods, the connecting rods are rotatably arranged on the base through moving parts, the two ends of the connecting rods are respectively provided with a scale pan and a connecting screw rod, the connecting screw rods are respectively positioned on two different moment directions of the power measuring force arm and act on the load sensor through the power measuring force arm and the end part of the ejector rod, the calibration device is suitable for measuring power loss when the reversible unit runs clockwise or anticlockwise, the calibration device well adapts to the characteristics and the position requirements of the output power loss of the reversible unit, has the functions of bidirectional in-situ calibration of the sensor and power measurement, and has higher reliability compared with a unidirectional power measuring mode, simple structure, installation, operation and maintenance are more convenient.

摘要：【中文】本发明公开了一种基于分词权重判断文章所属分类的方法、装置和存储介质，属于自然语言处理技术领域。所述方法包括：对文本语料进行分词和去停用词处理，得到训练集语料库；将训练集中的特征项转变成词频矩阵，统计每个特征项的TF-IDF值，作为特征权重；将训练集的权重矩阵和初始标签传给分类器，训练得到分类模型；获取待分类文章的权重矩阵，使用训练好的分类模型对文章进行分类。本发明通过分词，构建基于权重的词向量空间，再使用贝叶斯分类器可以对文章类型直接进行判断，可以在短时间内判断出大量的文本类型，且具有准确、稳定的优点。 【EN】The invention discloses a method, a device and a storage medium for judging article classification based on word segmentation weight, and belongs to the technical field of natural language processing. The method comprises the following steps: performing word segmentation and word stop removal processing on the text corpus to obtain a training set corpus; converting the characteristic items in the training set into a word frequency matrix, and counting the TF-IDF value of each characteristic item as a characteristic weight; transmitting the weight matrix and the initial label of the training set to a classifier, and training to obtain a classification model; and acquiring a weight matrix of the article to be classified, and classifying the article by using the trained classification model. The invention constructs a weight-based word vector space by word segmentation, can directly judge the article types by using the Bayesian classifier, can judge a large number of text types in a short time, and has the advantages of accuracy and stability.

摘要：本发明公开了一种基于分词权重判断文章所属分类的方法、装置和存储介质，属于自然语言处理技术领域。所述方法包括：对文本语料进行分词和去停用词处理，得到训练集语料库；将训练集中的特征项转变成词频矩阵，统计每个特征项的TF‑IDF值，作为特征权重；将训练集的权重矩阵和初始标签传给分类器，训练得到分类模型；获取待分类文章的权重矩阵，使用训练好的分类模型对文章进行分类。本发明通过分词，构建基于权重的词向量空间，再使用贝叶斯分类器可以对文章类型直接进行判断，可以在短时间内判断出大量的文本类型，且具有准确、稳定的优点。