摘要：【中文】本发明公开了一种基于板壳理论的深海管道屈曲传播压力的计算方法，具体步骤为：测量深海管道的半径R，和壁厚t；测量深海管道材料的拉伸应力‑应变关系，并用线性硬化弹塑性本构关系拟合得到屈服强度σ_s和线性硬化模量E₁；计算屈服影响系数m_y、应变硬化系数m_h和薄膜拉伸因子n；通过屈曲传播压力计算公式计算深海管道屈曲传播压力。本发明基于板壳理论，考虑壳体结构的薄膜拉伸效应在屈曲传播过程中的影响，整体地分析深海管道屈曲传播现象，提出更精确、更合理的屈曲传播压力计算方法。 【EN】The invention discloses a method for calculating deep sea pipeline buckling propagation pressure based on a plate shell theory, which comprises the following specific steps of: measuring the radius R and the wall thickness t of the deep sea pipeline; measuring the tensile stress-strain relation of deep sea pipeline material, and fitting by using linear hardening elastoplasticity constitutive relation to obtain yield strength sigma_sAnd linear hardening modulus E₁(ii) a Calculating yield influence coefficient m_yStrain hardening coefficient m_hAnd a film stretch factor n; and calculating the buckling propagation pressure of the deep sea pipeline through a buckling propagation pressure calculation formula. The invention is based on the plate-shell theory, considers the influence of the film stretching effect of the shell structure in the buckling propagation process, integrally analyzes the buckling propagation phenomenon of the deep sea pipeline, and provides a more accurate and more reasonable buckling propagation pressure calculation method.

摘要：【中文】本发明公开一种电池多极耳固定装置，包括通过铰接部进行相互铰接的第一固定件及第二固定件；第一固定件包括第一卡合部与第一把手部，第二固定件包括第二卡合部与第二把手部；第一卡合部与第二卡合部的相互靠近的两侧均设有可导电的压合板，其中一个压合板连接有一个外接极耳；第一把手部与第二把手部之间设有一个收紧机构；收紧机构同时收紧第一把手部与第二把手部将第一卡合部与第二卡合部进行压合。本发明通过卡扣连接代替传统的焊接，避免了焊接中存在的缺点，工艺简单，安装效率更高，提升了电池的可维修性与可替换性。 【EN】The invention discloses a multi-tab fixing device for a battery, which comprises a first fixing piece and a second fixing piece which are mutually hinged through a hinge part; the first fixing piece comprises a first clamping part and a first handle part, and the second fixing piece comprises a second clamping part and a second handle part; conductive laminated plates are arranged on two sides, close to each other, of the first clamping part and the second clamping part, and one laminated plate is connected with an external tab; a tightening mechanism is arranged between the first handle part and the second handle part; the tightening mechanism simultaneously tightens the first handle part and the second handle part to press the first clamping part and the second clamping part. The invention replaces the traditional welding by the buckle connection, avoids the defects in the welding, has simple process and higher installation efficiency, and improves the maintainability and the replaceability of the battery.

摘要：【中文】本发明公开了一种含脱层复合材料层合结构剩余承载强度的计算方法，步骤如下：确定复合材料层合结构的铺层方式以及单层板的材料参数；确定贯穿矩形脱层的脱层发生的位置、大小以及层数；根据脱层X方向的发生位置将层合板划分为若干个子板，根据脱层Z方向的发生位置将子板划分为若干个子区域，计算每个子板中每个子区域的弯曲刚度、拉伸刚度及拉弯耦合刚度；将子板中每个子区域相加等效为一个不含脱层的完整子板；建立用挠度表示的平衡微分方程；将满足平衡微分方程及定解条件的最小解作为该脱层层合板的临界屈曲荷载。该计算方法通过刚度等效原则避免考虑脱层界面的复杂接触性能，解决含多个脱层的复合材料层合结构的剩余承载能力的计算问题。 【EN】The invention discloses a method for calculating the residual bearing strength of a laminated structure of a composite material containing delamination, which comprises the following steps: determining a layering mode of a composite material laminating structure and material parameters of a single-layer plate; determining the location, size, and number of layers of delamination that occurred throughout the rectangular delamination; dividing the laminated board into a plurality of sub-boards according to the occurrence position in the delamination X direction, dividing the sub-boards into a plurality of sub-areas according to the occurrence position in the delamination Z direction, and calculating the bending rigidity, the tensile rigidity and the stretch-bending coupling rigidity of each sub-area in each sub-board; adding each sub-area in the daughter board is equivalent to a complete daughter board without delamination; establishing a balance differential equation expressed by deflection; and taking the minimum solution which meets the equilibrium differential equation and the definite solution condition as the critical buckling load of the delamination laminated plate. The calculation method avoids considering the complex contact performance of a delaminating interface through a rigidity equivalence principle, and solves the calculation problem of the residual bearing capacity of a composite material laminating structure containing a plurality of delayers.

摘要：【中文】本发明公开一种电池浆料制备方法及电池浆料，包括：S1：将磷酸铁锂粉末和粘接剂粉体进行混合，获得混合物料A；S2：向混合物料A中加入溶剂、悬浮剂和冷却水混合，获得混合物料B；S3：向混合物料B中加入沉淀剂混合并过滤，获得混合物料C；S4：将混合物料C刮料，获得混合物料D；S5：向混合物料D加入导电剂混合并进行分散、挤压，获得混合物料E；S6：对混合物料E进行真空脱泡、真空干燥，得到电池浆料。本发明通过向制备磷酸铁锂浆料的搅拌罐中加入悬浮剂，并在搅拌之后再加入沉淀剂搅拌并过滤，使得制备出来的磷酸铁锂浆料分散效果好，混合均匀，精简了配料工序，节约了人工成本和能耗。 【EN】The invention discloses a battery slurry preparation method and a battery slurry, which comprise the following steps: s1: mixing lithium iron phosphate powder and adhesive powder to obtain a mixed material A; s2: adding a solvent, a suspending agent and cooling water into the mixed material A, and mixing to obtain a mixed material B; s3: adding a precipitator into the mixed material B, mixing and filtering to obtain a mixed material C; s4: scraping the mixed material C to obtain a mixed material D; s5: adding a conductive agent into the mixed material D, mixing, dispersing and extruding to obtain a mixed material E; s6: and (5) carrying out vacuum defoaming and vacuum drying on the mixed material E to obtain the battery slurry. According to the invention, the suspending agent is added into the stirring tank for preparing the lithium iron phosphate slurry, and the precipitator is added after stirring and filtering are carried out, so that the prepared lithium iron phosphate slurry has a good dispersing effect and is uniformly mixed, the batching process is simplified, and the labor cost and the energy consumption are saved.

摘要：【中文】本发明公开一种人群密度估计方法，包括如下步骤：获取人群的视频图像；对所述人群的视频图像进行目标检测，一个行人利用一个Bbox进行标记，所有Bbox构成与人群对应的Bbox集合；根据所述Bbox集合，通过计算平均占地面积获得人群的密度。本发明的有益效果是：通过自动采集监控视频，并进行目标检测，根据检测结果通过求取平均占地面积的方法进行人群密度进行计算，实现了对人群密度进行自动估计，减少了人工工作量，提高了效率和估计精度，将人数与占地面积相结合，提高了估计精度，增强了密度估计的可比性，便于比较、分级。本发明还公开一种可读存储介质。 【EN】The invention discloses a crowd density estimation method, which comprises the following steps: acquiring a video image of a crowd; carrying out target detection on the video image of the crowd, marking a pedestrian by using a Bbox, and forming a Bbox set corresponding to the crowd by all the Bboxes; and according to the Bbox set, obtaining the density of the crowd by calculating the average floor area. The invention has the beneficial effects that: by automatically acquiring the monitoring video, detecting the target and calculating the crowd density by a method of calculating the average occupied area according to the detection result, the crowd density is automatically estimated, the manual workload is reduced, the efficiency and the estimation precision are improved, the number of people and the occupied area are combined, the estimation precision is improved, the comparability of density estimation is enhanced, and comparison and classification are facilitated. The invention also discloses a readable storage medium.

摘要：【中文】本发明提供的基于Mask R‑CNN和Soft‑NMS融合的群养粘连猪实例分割方法，包括建立群养粘连猪视频和图像数据集；构建基于Mask R‑CNN的分割网络，包括基础骨干网络、RPN网络和三分支网络；通过基础骨干网络对图像数据集进行多尺度特征的提取，得到特征映射；基于特征映射，RPN生成感兴趣区域RoIs，并对RoIs进行目标分类以及回归，利用Soft‑NMS去除冗余的检测框，获取最好的目标RoIs；对每个目标RoIs进行双线性插值操作，将每个RoIs处理为相同大小；利用三分支网络实现对相同大小的RoIs进行检测、分类和分割操作。本发明提供的群养粘连猪实例分割方法，针对猪只重叠、粘连及其他复杂环境导致目标猪只漏检问题，在基本的Mask R‑CNN框架上采用Soft‑NMS算法代替传统的NMS算法，能有效降低粘连猪的漏检率。 【EN】The invention provides a mass-cultured adherent pig example segmentation method based on Mask R-CNN and Soft-NMS fusion, which comprises the steps of establishing a mass-cultured adherent pig video and image data set; constructing a Mask R-CNN-based segmented network, which comprises a basic backbone network, an RPN network and a three-branch network; extracting multi-scale features of the image data set through a basic backbone network to obtain feature mapping; generating region-of-interest (RoIs) by the RPN based on feature mapping, performing target classification and regression on the RoIs, and removing redundant detection frames by using Soft-NMS to obtain the best target RoIs; carrying out bilinear interpolation operation on each target Rois, and processing each Rois into the same size; and the three-branch network is utilized to realize the detection, classification and segmentation operations of the RoIs with the same size. According to the method for segmenting the group-fed adherent pig example, provided by the invention, aiming at the problem that the target pig is missed to be detected due to overlapping, adhesion and other complex environments of the pig, the Soft-NMS algorithm is adopted on the basic Mask R-CNN framework to replace the traditional NMS algorithm, so that the missed detection rate of the adherent pig can be effectively reduced.