摘要：【中文】本发明属于能量收集领域，具体涉及一种能量收集装置、组串、系统，本发明装置包括天线单元、匹配电路、整流电路、储能单元；所述天线单元包括第一导电面、第二导电面、隔离板；所述第二导电面接地；所述第一导电面用于人体站立，以形成单极子天线；所述隔离板设置于所述一导电面和第二导电面之间将两个导电面隔离以形成馈电端口；所述天线单元的馈电端口与所述匹配电路电连接；所述匹配电路与所述整流电路电连接；所述整流电路与所述储能单元电连接。本发明可以有效的在人体站立在隔离板上后，形成单极子天线，进而稳定的收集能量。 【EN】The invention belongs to the field of energy collection, and particularly relates to an energy collection device, a string and a system, wherein the device comprises an antenna unit, a matching circuit, a rectifying circuit and an energy storage unit; the antenna unit comprises a first conductive surface, a second conductive surface and a separation plate; the second conductive surface is grounded; the first conductive surface is used for a human body to stand so as to form a monopole antenna; the isolating plate is arranged between the first conducting surface and the second conducting surface to isolate the two conducting surfaces to form a feed port; the feed port of the antenna unit is electrically connected with the matching circuit; the matching circuit is electrically connected with the rectifying circuit; the rectifying circuit is electrically connected with the energy storage unit. The invention can effectively form the monopole antenna after the human body stands on the isolation plate, thereby stably collecting energy.

摘要：【中文】本发明属于植入式医疗设备领域，具体涉及了一种全植入式脑机接口系统，旨在解决的问题。本发明系统包括：全植入单元，包括电极、前端采集、微处理器、温度传感器、无线通信模块、无线供电模块、MICS无线通信天线、HF无线供电天线，用于采集被测对象脑电信号并发送至外置监测单元以及采集全植入单元工作温度并结合设定温度阈值调整系统工作模式、报警信息，发送系统工作模式、报警信息至外置监测单元；外置监测单元，包括微处理器、无线通信模块、能量辐射模块、电池、MICS无线通信天线、HF无线供电天线，用于接收全植入式单元传递的信息。本发明可检测高时空分辨率的脑电信号，具有尺寸小、负荷低、配置灵活、不易感染的优点。 【EN】The invention belongs to the field of implantable medical equipment, and particularly relates to a fully-implantable brain-computer interface system, aiming at solving the problem. The system of the invention comprises: the fully-implanted unit comprises an electrode, a front-end acquisition unit, a microprocessor, a temperature sensor, a wireless communication module, a wireless power supply module, an MICS wireless communication antenna and an HF wireless power supply antenna, and is used for acquiring electroencephalograms of a measured object, sending the electroencephalograms to the external monitoring unit, acquiring the working temperature of the fully-implanted unit, adjusting the working mode of the system and alarm information by combining a set temperature threshold, and sending the working mode of the system and the alarm information to the external monitoring unit; the external monitoring unit comprises a microprocessor, a wireless communication module, an energy radiation module, a battery, an MICS wireless communication antenna and an HF wireless power supply antenna and is used for receiving information transmitted by the full-implanted unit. The invention can detect the electroencephalogram signals with high space-time resolution and has the advantages of small size, low load, flexible configuration and difficult infection.

摘要：【中文】本发明属于实验器械领域，具体提供一种超薄切片机及其刀台。本发明旨在解决现有的超薄切片机切出的切片行程较短，无法满足科研需求的问题。为此，本发明的刀台包括与超薄切片机的底座固定连接的刀台底座、与所述刀台底座滑动连接的刀台本体、设置在所述刀台本体上的压电陶瓷片和与所述刀台本体相连接的刀架。其中，压电陶瓷片通电时能够产生1mm的形变量，并且压电陶瓷片通电时能够驱动刀架及安装在刀架上的刀具向样品臂移动，并切割样品臂上的样品。本发明具有上述刀台的超薄切片机切割出的切片能够达到1mm，满足科研人员的需求。 【EN】The invention belongs to the field of experimental instruments, and particularly provides an ultrathin slicer and a knife table thereof. The invention aims to solve the problems that the conventional ultrathin slicing machine has short slicing stroke and cannot meet the scientific research requirement. Therefore, the tool post comprises a tool post base fixedly connected with a base of the ultrathin slicer, a tool post body in sliding connection with the tool post base, a piezoelectric ceramic piece arranged on the tool post body and a tool rest connected with the tool post body. The piezoelectric ceramic piece can generate a deformation amount of 1mm when being electrified, and can drive the tool rest and a tool arranged on the tool rest to move towards the sample arm when being electrified, and cut a sample on the sample arm. The slice cut by the ultrathin slicer with the cutter table can reach 1mm, and the requirements of scientific researchers are met.

摘要：【中文】本发明属于智能驾驶技术领域，具体涉及了一种基于加速度前馈的异构车队协同自适应巡航控制方法，旨在解决现有的协同自适应巡航控制方法依赖开发人员的专业经验，设计难度大的问题。本发明方法包括：建立被控车辆纵向动力学模型，定义跟车策略；获取被控车辆与前车的传递函数，并分析车队频域弦稳定条件；建立被控车队的状态空间模型；确定被控车队时域弦稳定条件；基于预设车辆间隔时间参数，获取前馈控制参数和反馈控制参数进行车辆巡航控制。本发明方法方便使用计算工具求解问题的可行解，降低了设计难度；获取最小间隔时间参数，从而获得具有最好跟随性能的协同自适应巡航控制器，提高了整个车队的通行能力。 【EN】The invention belongs to the technical field of intelligent driving, and particularly relates to an acceleration feedforward-based heterogeneous fleet cooperative adaptive cruise control method, aiming at solving the problems that the conventional cooperative adaptive cruise control method depends on the professional experience of developers and is difficult to design. The method comprises the following steps: establishing a longitudinal dynamic model of a controlled vehicle, and defining a following strategy; obtaining a transfer function of a controlled vehicle and a front vehicle, and analyzing a motorcade frequency domain chord stability condition; establishing a state space model of a controlled fleet; determining the time domain string stability condition of a controlled fleet; and acquiring a feedforward control parameter and a feedback control parameter to carry out vehicle cruise control based on a preset vehicle interval time parameter. The method of the invention is convenient to use a calculation tool to solve feasible solutions of problems, and reduces the design difficulty; and acquiring the minimum interval time parameter so as to obtain a cooperative self-adaptive cruise controller with the best following performance and improve the traffic capacity of the whole motorcade.

摘要：【中文】本发明公开一种消防机器人，涉及消防设备的技术领域，其包括移动平台，所述移动平台上搭载有第二摄像装置、第一摄像装置、灭火装置以及本地控制器,第一摄影装置为全景摄影装置，第二摄影装置为单向摄影装置，所述本地控制器内写入有控制系统，控制系统对第二摄像装置和第一摄像装置拍摄的影像进行火源识别与定位，并根据定位后确定的位置信息控制移动平台和灭火装置的移动，还能够控制灭火装置进行灭火作业。本发明其能够全方位的对火源自动识别与定位，并自动控制灭火装置瞄准火源并进行灭火。 【EN】The invention discloses a fire-fighting robot, which relates to the technical field of fire-fighting equipment and comprises a mobile platform, wherein a second camera device, a first camera device, a fire extinguishing device and a local controller are mounted on the mobile platform, the first camera device is a panoramic camera device, the second camera device is a one-way camera device, a control system is written in the local controller, the control system identifies and positions a fire source of an image shot by the second camera device and the first camera device, controls the mobile platform and the fire extinguishing device to move according to position information determined after positioning, and can also control the fire extinguishing device to carry out fire extinguishing operation. The fire extinguishing device can automatically identify and position the fire source in an all-around manner, and automatically control the fire extinguishing device to aim at the fire source and extinguish fire.

摘要：【中文】本发明属于自动驾驶技术领域，具体涉及一种基于云端的辅助驾驶控制系统及方法，旨在解决驾驶员操作驾驶模拟器进行远程辅助驾驶控制无法实现一对多车的接管控制与调度的问题。本系统包括设置于远程服务器的云端、设置于被控车辆的车辆端；车辆端与云端通过无线通讯链路连接；车辆端，配置为基于控制状态切换指令切换至远程控制状态时，获取车辆行驶环境数据、车辆驾驶行为数据并发送至云端，获取云端发送的车辆控制数据，以对车辆进行控制；云端，配置为接收车辆行驶环境数据、车辆驾驶行为数据，基于预设的自动驾驶控制模型，获取对应被控车辆的车辆控制数据，并发送至对应的车辆端。本发明通过云端实现了一对多车的接管控制及调度。 【EN】The invention belongs to the technical field of automatic driving, and particularly relates to a cloud-based auxiliary driving control system and method, aiming at solving the problem that the control and scheduling of the take-over of one-to-many vehicles cannot be realized by a driver operating a driving simulator to carry out remote auxiliary driving control. The system comprises a cloud end arranged on a remote server and a vehicle end arranged on a controlled vehicle; the vehicle end is connected with the cloud end through a wireless communication link; the vehicle end is configured to acquire vehicle running environment data and vehicle driving behavior data and send the vehicle running environment data and the vehicle driving behavior data to the cloud end when switching to a remote control state based on the control state switching instruction, and acquire vehicle control data sent by the cloud end to control the vehicle; and the cloud end is configured to receive vehicle running environment data and vehicle driving behavior data, acquire vehicle control data corresponding to the controlled vehicle based on a preset automatic driving control model, and send the vehicle control data to the corresponding vehicle end. The invention realizes the control and dispatching of the take-over of one-to-many vehicles through the cloud.

摘要：【中文】本发明属于深度学习领域，具体涉及一种基于FPGA的加速卷积计算的系统、卷积神经网络，旨在为了解决解决现有技术中的上述问题。本发明包括：参数量化模块，存储各卷积层的定点化后的权值参数、尺度、偏置；参数加载模块，将定点化后的CNN模型参数文件加载到FPGA中；输入模块，获取输入数据定点化后的低比特数据；卷积计算模块，将输入数据的特征图矩阵拆分为多个小矩阵依次加载到FPGA中，根据卷积核的数量分批进行卷积计算；输出模块，各小矩阵对应的卷积计算结果进行合并作为下一层的输入图像；本发明在硬件FPGA上保证了网络模型精度损失很小的前提下，减少网络模型的存储，实现加速卷积计算。 【EN】The invention belongs to the field of deep learning, and particularly relates to a system for accelerating convolution calculation based on an FPGA (field programmable gate array) and a convolution neural network, aiming at solving the problems in the prior art. The invention comprises the following steps: the parameter quantization module stores the fixed-point weight parameters, the scales and the offsets of the convolution layers; the parameter loading module is used for loading the fixed-point CNN model parameter file into the FPGA; the input module is used for acquiring low-bit data after the input data is fixed in point; the convolution calculation module divides the characteristic diagram matrix of the input data into a plurality of small matrixes which are sequentially loaded into the FPGA, and performs convolution calculation in batches according to the number of convolution kernels; the output module is used for combining convolution calculation results corresponding to the small matrixes to be used as an input image of the next layer; the invention reduces the storage of the network model and realizes the acceleration of the convolution calculation on the premise of ensuring small precision loss of the network model on the hardware FPGA.

摘要：【中文】本发明属于计算机深度强化学习、模式识别领域，具体涉及了一种基于形状注意力机制的三维目测检测方法、系统、装置，旨在解决单阶段检测器精度低于两阶段检测器，而两级检测器耗时大、不适用于实时系统的问题。本发明包括：通过三维网格体素表示点云数据；提取特征并编码空间稀疏特征图；投影到顶视图后提取不同尺度特征；采用反卷积层合并特征；通过注意力权重及卷积编码层提取形状注意力特征图；通过目标分类网络和回归定位网络获取目标类别和目标位置、尺寸、方向。本发明使用基于距离约束的采样策略以及基于形状先验的注意力机制，缓解数据分布不均匀导致的不稳定，改善单阶段检测器缺乏形状先验的问题，精度高、耗时短、实时性强、鲁棒性好。 【EN】The invention belongs to the field of computer deep reinforcement learning and pattern recognition, and particularly relates to a three-dimensional visual inspection detection method, system and device based on a shape attention mechanism, aiming at solving the problems that the precision of a single-stage detector is lower than that of a two-stage detector, and the two-stage detector consumes much time and is not suitable for a real-time system. The invention comprises the following steps: representing the point cloud data by three-dimensional grid voxels; extracting features and coding a space sparse feature map; extracting different scale features after projecting to a top view; adopting a deconvolution layer merging characteristic; extracting a shape attention feature map through attention weight and a convolution coding layer; and acquiring the target category, the target position, the target size and the target direction through a target classification network and a regression positioning network. The invention uses a sampling strategy based on distance constraint and an attention mechanism based on shape prior, relieves the instability caused by uneven data distribution, improves the problem that a single-stage detector lacks shape prior, and has high precision, short time consumption, strong real-time performance and good robustness.

摘要：【中文】本发明属于特种机器人技术领域，旨在实现运载机构在线上行走时的稳定性，本发明提供了一种用于线上行走的运载机构，包括由第一安装板、第二安装板和纵向移动板组成的运载平台、行走装置、夹持装置、驱动装置和用于调整运载机构姿态的自平衡控制装置；纵向移动板可滑动地设置于与第一安装板平行固定连接的第二安装板之上；行走装置为至少两组沿行进方向设置的行走轮；夹持装置可滑动地设置于纵向移动板的下侧；驱动装置包括驱动行走轮滚动的第一驱动装置和驱动纵向移动板移动的第二驱动装置，夹持装置在纵向移动板的带动下夹紧或松开待巡检目标物。本发明的有益效果为：通过行走装置、夹持装置和自平衡控制装置配合使用，有效保证行车稳定性。 【EN】The invention belongs to the technical field of special robots, and aims to realize the stability of a carrying mechanism when the carrying mechanism walks on a line; the longitudinal moving plate is slidably arranged on a second mounting plate which is fixedly connected with the first mounting plate in parallel; the walking device is at least two groups of walking wheels arranged along the traveling direction; the clamping device is slidably arranged on the lower side of the longitudinal moving plate; the driving device comprises a first driving device for driving the traveling wheels to roll and a second driving device for driving the longitudinal moving plate to move, and the clamping device is driven by the longitudinal moving plate to clamp or loosen the object to be inspected. The invention has the beneficial effects that: through running gear, clamping device and self-balancing controlling means cooperation use, effectively guarantee driving stability.

摘要：【中文】本发明涉及一种深度二值神经网络训练方法及系统，所述训练方法包括：初始化浮点型的深度神经网络，得到初始化网络模型；基于交替方向乘子法，根据所述初始化网络模型，采用目标传播算法，得到具有二值激活和浮点型权重的优化深度神经网络；基于交替方向乘子法，根据所述优化深度神经网络，得到深度二值神经网络。本发明深度二值神经网络训练方法通过交替方向乘子法优化框架，将权重和激活分别进行二值化，能够减轻同时二值化带来的耦合效应，提高深度二值神经网络的训练效果；采用目标传播算法优化具有二值激活的深度神经网络，能够减小量化过程不可微导致的深度神经网络量化优化困难问题。 【EN】The invention relates to a deep binary neural network training method and a deep binary neural network training system, wherein the training method comprises the following steps: initializing a floating point type deep neural network to obtain an initialized network model; based on an alternating direction multiplier method, obtaining an optimized deep neural network with binary activation and floating point type weight by adopting a target propagation algorithm according to the initialized network model; and obtaining a deep binary neural network according to the optimized deep neural network based on an alternating direction multiplier method. According to the deep binary neural network training method, the frame is optimized through the alternative direction multiplier method, the weight and the activation are respectively binarized, the coupling effect caused by simultaneous binarization can be reduced, and the training effect of the deep binary neural network is improved; the deep neural network with binary activation is optimized by adopting a target propagation algorithm, so that the problem of difficulty in quantization optimization of the deep neural network caused by irreducible quantization process can be solved.