摘要：【中文】本发明公开了一种三坐标设备位置校准方法，在非参数回归的基础上，采用改进遗传算法优化网络参数，以误差样本数据为后验条件，执行非参数估计，依据最大概率原则计算对应输入位置条件下的网络的输出，即对应输入条件下的位置误差，通过遗传算法更新，找到最优网络结构，实现误差的精确预测与修正。该方法以径向基网络为基础，采用改进遗传算法进行网络参数优化，具有良好的非线性逼近性能，训练更为方便，计算精度较高，计算速度较快，满足误差修正高精度和实时性要求。 【EN】The invention discloses a position calibration method for three-coordinate equipment, which is characterized in that on the basis of nonparametric regression, an improved genetic algorithm is adopted to optimize network parameters, error sample data is taken as a posterior condition, nonparametric estimation is executed, the output of a network under the condition of corresponding input position, namely the position error under the condition of corresponding input, is calculated according to the maximum probability principle, and the optimal network structure is found through updating of the genetic algorithm, so that the accurate prediction and correction of the error are realized. The method is based on the radial basis network, adopts the improved genetic algorithm to optimize the network parameters, has good nonlinear approximation performance, is more convenient to train, has higher calculation precision and higher calculation speed, and meets the requirements of high precision and real-time performance of error correction.

摘要：【中文】本发明公开了一种多主机多深度摄像头的深度图像拼接融合方法及系统，投影主机连接多个节点主机，每个节点连接多个深度摄像头；将被采集识别画面的区域划分为多个摄像区域，每个摄像区域对应一个深度摄像头；所有深度摄像头同时采集对应摄像区域的深度数据；每个节点主机接收其对应的多个深度摄像头的深度数据，并基于每个深度摄像头的深度环境背景数据和三维空间识别范围参数，分别计算并线性变换深度数据，得到可处理和显示的第一深度图像；每个节点主机拼接融合多个深度摄像头的第一深度图像，得到第二深度图像；投影主机接收多个节点主机的第二深度图像，按指定顺序拼接第二深度图像，识别互动者的位置及显示拼接后的第三深度图像。 【EN】The invention discloses a depth image splicing and fusing method and system of a multi-host multi-depth camera.A projection host is connected with a plurality of node hosts, and each node is connected with a plurality of depth cameras; dividing the area of the collected identification picture into a plurality of camera shooting areas, wherein each camera shooting area corresponds to a depth camera; all the depth cameras simultaneously acquire depth data of corresponding camera shooting areas; each node host receives depth data of a plurality of depth cameras corresponding to each node host, and respectively calculates and linearly converts the depth data based on the depth environment background data and the three-dimensional space identification range parameters of each depth camera to obtain a first depth image which can be processed and displayed; each node host splices and fuses the first depth images of the multiple depth cameras to obtain a second depth image; and the projection host receives the second depth images of the plurality of node hosts, splices the second depth images according to an appointed sequence, identifies the position of an interdynamic person and displays the spliced third depth images.

摘要：【中文】本发明公开了一种基于光学感应识别的互动装置，包括：不见光发射端，用于发射不见光；不见光接收端，不见光接收端相对应不见光发射端设置，用于接收不见光发射端发射的不见光；MCU、灯光设备和声音播放设备，MCU的输入端与不见光接收端相连，MCU的输出端与灯光设备和声音播放设备相连；当MCU检测到不见光发射端与不见光接收端之间有物体经过时，向灯光设备和声音播放设备发出控制指令，改变灯光设备和声音播放设备的状态。本发明可实现对运动物体感应的准确识别，在识别后进行发出不同音效和灯光作为互动，为人们的体验增加了互动乐趣。 【EN】The invention discloses an interaction device based on optical induction recognition, which comprises: a invisible light emitting end for emitting invisible light; the invisible light receiving end is arranged corresponding to the invisible light emitting end and is used for receiving invisible light emitted by the invisible light emitting end; the system comprises an MCU, a lighting device and a sound playing device, wherein the input end of the MCU is connected with a non-visible light receiving end, and the output end of the MCU is connected with the lighting device and the sound playing device; when the MCU detects that an object passes through between the invisible light emitting end and the invisible light receiving end, the MCU sends a control instruction to the lighting equipment and the sound playing equipment to change the states of the lighting equipment and the sound playing equipment. The invention can realize accurate identification of the moving object induction, and sends out different sound effects and lamplight after identification to be used as interaction, thereby increasing interactive pleasure for the experience of people.

摘要：【中文】本发明公开了一种船舶围护系统安装平台的脚手架拆除移动平台，所述移动平台包括：底座框架，所述底座框架的底部固定连接有多个万向轮；两排相对设置的立管，所述立管垂直连接在所述底座框架的左右两侧上，所述立管在不同高度上设置有多组盘扣；连接横梁，所述连接横梁架设在所述立管中间，所述连接横梁的两端分别通过所述盘扣固定；连接梁，所述连接梁固定连接在所述连接横梁上，所述连接梁上设置有起重滑轮；以及绞车。本发明适用于安装平台的特殊位置的脚手架拆除，移动平台通过有一定高度的可移动框架，将立柱迁移到平面上，避免人工搬运破坏LNG船液货舱围护系统的波纹板。 【EN】The invention discloses a scaffold dismantling mobile platform of a ship containment system installation platform, which comprises: the bottom of the base frame is fixedly connected with a plurality of universal wheels; two rows of vertical pipes which are arranged oppositely are vertically connected to the left side and the right side of the base frame, and a plurality of groups of disc buckles are arranged on the vertical pipes at different heights; the connecting cross beam is erected in the middle of the vertical pipe, and two ends of the connecting cross beam are fixed through the disc buckles respectively; the connecting beam is fixedly connected to the connecting cross beam, and a hoisting pulley is arranged on the connecting beam; and a winch. The invention is suitable for dismantling scaffolds at special positions of an installation platform, and the movable platform transfers the upright post to a plane through a movable frame with a certain height, so that the corrugated plate of the LNG ship cargo tank containment system is prevented from being damaged by manual carrying.

摘要：【中文】本发明公开了一种图形图像识别方法及识别系统，包括：计算平台的均值背景深度图；采集平台上覆盖有红外反光材料的被测道具的红外图和实时深度图，预处理得到红外二值图像；将投影图像和红外二值图像自动校点对齐；识别红外二值图像的坐标和长宽值；将均值背景深度图与实时深度图进行差值运算，得到差值深度图；通过红外二值图像识别的坐标值，确定被测道具在差值深度图内的高度值；通过外部设置每种被测道具的几何参数，对被测道具进行分类；将被测道具的识别点坐标映射到投影图像的显示坐标处，将被测道具类别和变换后的坐标发送到互动投影效果软件。本发明可简单快速的实现图形图像的识别，识别过程为实时处理，当图形改变会立即识别出图形。 【EN】The invention discloses a graphic image recognition method and a recognition system, comprising the following steps: calculating a mean background depth map of the platform; acquiring an infrared image and a real-time depth image of a detected prop covered with an infrared reflective material on a platform, and preprocessing to obtain an infrared binary image; automatically correcting and aligning the projected image and the infrared binary image; identifying coordinates and length and width values of the infrared binary image; performing difference operation on the mean background depth map and the real-time depth map to obtain a difference depth map; determining the height value of the detected prop in the difference depth map through the coordinate value identified by the infrared binary image; classifying the tested props by externally setting the geometric parameters of each tested prop; and mapping the coordinates of the identification points of the detected props to the display coordinates of the projected images, and sending the categories and the transformed coordinates of the detected props to the interactive projection effect software. The invention can simply and quickly realize the recognition of the graphic image, the recognition process is real-time processing, and the graphic can be recognized immediately when the graphic changes.

摘要：【中文】本公开提供了一种隔墙连接结构，包括：至少两个隔墙模块，相邻的两个隔墙模块形成任意形结构，每个隔墙模块包括龙骨和装饰板，其中龙骨依次首尾相连形成龙骨模块，装饰板设置在龙骨模块每个开口的侧面；连接模块，连接模块包括连接龙骨和配套装饰板，连接模块用于连接至少两个隔墙模块。本公开还提供了一种隔墙组合方式。本公开的隔墙系统，住户只需拆装局部隔墙模块并重新组合，就能实现不同户型的转换。 【EN】The present disclosure provides a partition wall connection structure, include: the partition wall module comprises at least two partition wall modules, wherein every two adjacent partition wall modules form an arbitrary structure, each partition wall module comprises a keel and a decorative plate, the keels are sequentially connected end to form a keel module, and the decorative plate is arranged on the side surface of each opening of the keel module; the connecting module comprises a connecting keel and a matched decorative plate, and the connecting module is used for connecting at least two partition wall modules. The present disclosure also provides a partition wall combination mode. According to the partition wall system, a resident can realize conversion of different house types only by disassembling and assembling the local partition wall modules and recombining the local partition wall modules.

摘要：【中文】本发明公开了一种多计算机视频和软件同步运行控制系统和方法，其中，控制系统包括：集中控制主机、交换机和节点主机；集中控制主机与交换机通信连接，多台节点主机分别与交换机相连接，节点主机与播放设备相连接；集中控制主机通过交换机同步向所有节点主机发送控制数据，节点主机接收控制数据，并通过播放设备同步播放视频或运行软件。通过本发明的技术方案，能够保证视频播放全程都有良好的同步性，以及实现节点主机同步启动软件并运行的功能。 【EN】The invention discloses a multi-computer video and software synchronous operation control system and a method thereof, wherein the control system comprises: the system comprises a centralized control host, a switch and a node host; the centralized control host is in communication connection with the switch, the plurality of node hosts are respectively connected with the switch, and the node hosts are connected with the playing equipment; the centralized control host synchronously sends control data to all the node hosts through the switch, and the node hosts receive the control data and synchronously play videos or run software through the playing equipment. By the technical scheme, the video playing system and the video playing method can ensure that the video playing is well synchronized in the whole process, and realize the function of synchronously starting software and running by the node host.

摘要：【中文】本发明公开了一种船舶围护系统安装平台的支撑腿调高机构，所述调高机构包括：基座，所述基座的底部与箱体固定连接，所述基座上部设置有凹入的球面结构；以及螺纹轴，所述螺纹轴的底端设置有与所述球面结构相配合的球帽，所述螺纹轴的顶端螺纹连接有调节螺母，所述调节螺母与所述支撑腿固定连接，所述螺纹轴上开设有穿孔，所述穿孔中设置有扳手。本发明的调高机构安装在支撑腿和底部支撑之间，起到调节安装平台支撑腿高度的作用，使得整个安装平台的处于一个平面上，同时用于支撑腿侧力，保证安装平台工作平面和绝缘箱受力不超过设计载荷。 【EN】The invention discloses a supporting leg height adjusting mechanism of a ship containment system mounting platform, which comprises: the bottom of the base is fixedly connected with the box body, and the upper part of the base is provided with a concave spherical structure; the bottom end of the threaded shaft is provided with a ball cap matched with the spherical structure, the top end of the threaded shaft is in threaded connection with an adjusting nut, the adjusting nut is fixedly connected with the supporting leg, a through hole is formed in the threaded shaft, and a wrench is arranged in the through hole. The height adjusting mechanism is arranged between the supporting leg and the bottom support, plays a role in adjusting the height of the supporting leg of the mounting platform, enables the whole mounting platform to be positioned on the same plane, and is used for supporting the lateral force of the supporting leg, so that the stress of the working plane of the mounting platform and the stress of the insulating box are not more than the design load.

摘要：【中文】本发明公开了一种弧形墙面互动识别系统和识别方法，其中识别系统包括：上位机数据处理软件、激光雷达和网络通信装置；至少两个激光雷达分别以雷达识别面平行于弧形墙面正切面的方向设置于弧形墙面上方预设高度处；该至少两个激光雷达通过网络通信装置将检测数据均发送至上位机数据处理软件；上位机数据处理软件将激光雷达检测数据进行处理，合并成为弧形墙面的图像。通过本发明的技术方案，既能得到精准的整个弧形墙面的图像，且激光雷达设置方式与安装方式简单便捷。 【EN】The invention discloses an interactive recognition system and a recognition method for an arc-shaped wall surface, wherein the recognition system comprises: the system comprises upper computer data processing software, a laser radar and a network communication device; at least two laser radars are respectively arranged at the preset height above the arc-shaped wall surface in the direction that the radar recognition surface is parallel to the normal tangent plane of the arc-shaped wall surface; the at least two laser radars send detection data to the upper computer data processing software through the network communication device; and the upper computer data processing software processes the laser radar detection data and combines the laser radar detection data into an image of the arc-shaped wall surface. By the technical scheme, accurate images of the whole arc-shaped wall surface can be obtained, and the laser radar setting mode and the laser radar installation mode are simple and convenient.

摘要：【中文】本申请提供一种信息系统的维护方法、装置和计算机存储介质，获取特征向量序列，特征向量序列包括按时间先后顺序依次获取的待预测日期之前N个日期的特征向量，每个日期的特征向量，均包括当天的时间特征和当天的业务特征，然后将特征向量序列输入预先训练的深度学习预测模型，得到待预测日期对应的系统性能指标预测值；其中，深度学习预测模型利用历史数据作为训练集训练得到，训练集包含过去M个日期的特征向量及这M个日期对应的系统性能指标，系统性能指标预测值作为维护信息系统的依据。本方案利用过去若干天的特征向量预测将来的系统性能指标，从而及时地发现将来可能出现的系统异常，起到发现系统隐患并提前预警或扩容的效果。 【EN】The application provides a method and a device for maintaining an information system and a computer storage medium, a characteristic vector sequence is obtained, the characteristic vector sequence comprises characteristic vectors which are sequentially obtained in time sequence and are N dates before a date to be predicted, the characteristic vector of each date comprises a time characteristic of the current day and a service characteristic of the current day, and then the characteristic vector sequence is input into a pre-trained deep learning prediction model to obtain a system performance index prediction value corresponding to the date to be predicted; the deep learning prediction model is obtained by training by using historical data as a training set, the training set comprises feature vectors of past M dates and system performance indexes corresponding to the M dates, and a system performance index prediction value is used as a basis for maintaining an information system. According to the scheme, the characteristic vectors of the past days are used for predicting the performance indexes of the system in the future, so that the possible system abnormity in the future can be found in time, and the effects of finding the hidden danger of the system and early warning or capacity expansion in advance are achieved.