摘要：【中文】本发明提供了一种图像分类方法，包括：标记训练图像的预设关键点的位置及可见性；根据预设关键点的位置确定关键点位置预测算法；根据关键点的可见性确定可见性预测算法；获取待分类图像的全局特征；根据关键点位置预测算法预测待分类图像的预设关键点的位置；根据可见性预测算法预测待分类图像的预设关键点的可见性；根据预设关键点的位置及可见性提取局部特征；根据全局特征与局部特征拼接后的结果预测对应的分类结果。本发明的图像分类方法能够按照图片中人物的整体印象对图片分类。 【EN】The invention provides an image classification method, which comprises the following steps: marking the position and visibility of a preset key point of a training image; determining a key point position prediction algorithm according to the position of a preset key point; determining a visibility prediction algorithm according to the visibility of the key points; acquiring global features of an image to be classified; predicting the position of a preset key point of the image to be classified according to a key point position prediction algorithm; predicting the visibility of preset key points of the image to be classified according to a visibility prediction algorithm; extracting local features according to the position and visibility of a preset key point; and predicting a corresponding classification result according to the result of splicing the global features and the local features. The image classification method can classify the pictures according to the overall impression of the people in the pictures.

摘要：【中文】本发明公开了一种金属应变速率敏感系数图的建立方法，该方法包括以下步骤：一、将金属在动态载荷条件下进行压缩，得到真应力σ和真应变ε，绘制得到σ‑ε曲线；二、将σ‑ε曲线进行数据处理，得到lnσ和数据；三、将lnσ和数据绘制散点图，然后进行非线性拟合，得到样条曲线；四、将样条曲线求偏导，得到m值；步骤五、将m值进行数据扩展，然后绘制得到金属应变速率敏感系数图。本发明基于真应力σ和真应变ε数据，进行数学处理，得到动态载荷条件下的m值，然后进行数据扩展，建立了全参数金属应变速率敏感系数图，操作简单，无需繁复的公式迭代，适用于大多数金属的应变速率敏感系数的计算，从而进行较优成形工艺参数的选择。 【EN】The invention discloses a method for establishing a metal strain rate sensitive coefficient graph, which comprises the following steps: firstly, compressing metal under a dynamic load condition to obtain a true stress sigma and a true strain epsilon, and drawing to obtain a sigma-epsilon curve; secondly, carrying out data processing on the sigma-epsilon curve to obtain ln sigma sumData; thirdly, sum of ln sigmaDrawing a scatter diagram by using the data, and then carrying out nonlinear fitting to obtain a spline curve; fourthly, solving the partial derivatives of the spline curves to obtain m values; and fifthly, performing data expansion on the m value, and then drawing to obtain a metal strain rate sensitivity coefficient graph. The method is based on true stress sigma and true strain epsilon data, mathematical processing is carried out to obtain an m value under a dynamic load condition, then data expansion is carried out to establish a full-parameter metal strain rate sensitive coefficient diagram, the operation is simple, complex formula iteration is not needed, and the method is suitable for calculating the strain rate sensitive coefficients of most metals, so that the selection of the optimal forming process parameters is carried out.

摘要：【中文】本发明公开了一种金属低速变形再结晶临界应力的确定方法，该方法包括以下步骤：一、将金属在动态载荷条件下进行压缩，得到σ和ε数据；二、将σ和ε数据绘制曲线，得到σ‑ε曲线；三、将σ‑ε曲线求导，得到θ‑σ曲线；四、将θ‑σ曲线对σ求偏导，得到曲线，曲线纵坐标的最小值对应的σ_c为低速变形再结晶临界应力。本发明基于σ和ε数据进行模拟计算，得到金属低速变形再结晶临界应力，得到了动态载荷条件下的金属低速变形再结晶的临界条件，避免了使用经验公式进行估算导致的结果不准确的问题，适用于大多数金属的低速变形再结晶临界应力的确定，实用性强，易于推广。 【EN】The invention discloses a method for determining the critical stress of low-speed deformation recrystallization of metal, which comprises the following steps: firstly, compressing metal under a dynamic load condition to obtain sigma and epsilon data; secondly, drawing a curve by the sigma and epsilon data to obtain a sigma-epsilon curve; thirdly, obtaining a theta-sigma curve by derivation of the sigma-epsilon curve; fourthly, solving the deviation of the theta-sigma curve to sigma to obtainThe curves are shown in the figure, and,σ corresponding to the minimum value of the ordinate of the curve_cIs the critical stress for low-speed deformation recrystallization. The invention is based on sigmaAnd the epsilon data are subjected to simulation calculation to obtain the critical stress of the metal low-speed deformation recrystallization, the critical condition of the metal low-speed deformation recrystallization under the dynamic load condition is obtained, the problem of inaccurate result caused by estimation by using an empirical formula is avoided, and the method is suitable for determining the critical stress of the low-speed deformation recrystallization of most metals, strong in practicability and easy to popularize.

摘要：【中文】本专利公开了一种特高压输电线路巡视检修机器人及作业方法，第一分机(100)和第二分机(200)通过连接臂(10)连接，在连接臂(10)上设置有垂直旋转机构(9)；第一分机(100)和第二分机(200)包括轨道夹持机构、行走机构、控制器(17)、电源(18)、水平旋转机构、无线通讯模块(28)和视频摄相机(26)，分别通过轨道夹持轮(13)的轮缘夹持在防坠落轨道(1)的翼板的两侧；本专利的优点是：工作人员不用登塔，在地面通过人机对话即可完成检修工作；提高了检测效率和供电可靠性，提高了工作人员的安全性，大大减轻了一线人员的劳动强度；对带电线路和停电线路都可以正常完成检测工作。 【EN】The patent discloses an inspection and overhaul robot for an extra-high voltage transmission line and an operation method, wherein a first branch machine (100) is connected with a second branch machine (200) through a connecting arm (10), and a vertical rotating mechanism (9) is arranged on the connecting arm (10); the first branch machine (100) and the second branch machine (200) comprise a track clamping mechanism, a traveling mechanism, a controller (17), a power supply (18), a horizontal rotating mechanism, a wireless communication module (28) and a video camera (26), and the track clamping mechanism and the horizontal rotating mechanism are respectively clamped on two sides of a wing plate of the anti-falling track (1) through the wheel rim of a track clamping wheel (13); the advantage of this patent is: the maintenance work can be completed on the ground through man-machine conversation without climbing a tower by workers; the detection efficiency and the power supply reliability are improved, the safety of workers is improved, and the labor intensity of front-line workers is greatly reduced; the detection work can be normally finished on live lines and power failure lines.