摘要：【中文】本发明涉及一种移动机器人SLAM问题中的实时回环检测方法，该方法包括以下步骤：1)获取移动机器人存储的子图；2)获取移动机器人的当前扫描帧，基于移动机器人存储的子图对当前扫描帧进行初步回环检测，获取处于同一环境地图位置的扫描帧‑子图对；3)对处于同一地图位置的扫描帧‑子图对进行相关性度量和复杂性度量计算；4)基于相关性度量和复杂性度量的计算结果，获得最终回环检测结果。与现有技术相比，本发明采用相关性度量和复杂性度量对初步回环检测进行了验证，既保证了精度又保证了实时性，采用分枝定界法和预计算栅格获取相对位姿，有效减小了计算量，提高了回环检测效率。 【EN】The invention relates to a real-time loop detection method in the SLAM problem of a mobile robot, which comprises the following steps: 1) acquiring a subgraph stored by the mobile robot; 2) acquiring a current scanning frame of the mobile robot, performing primary loop detection on the current scanning frame based on sub-images stored by the mobile robot, and acquiring scanning frame-sub-image pairs at the same environment map position; 3) carrying out correlation measurement and complexity measurement calculation on scanning frame-sub-image pairs at the same map position; 4) and obtaining a final loop detection result based on the calculation results of the correlation metric and the complexity metric. Compared with the prior art, the method adopts the relevance measurement and the complexity measurement to verify the preliminary loop detection, not only ensures the precision, but also ensures the real-time performance, adopts the branch-and-bound method and the pre-calculation grid to obtain the relative pose, effectively reduces the calculated amount and improves the loop detection efficiency.

摘要：【中文】本发明公开了一种用于全自动泊车的实时高精度定位方法及装置，方法包括：设置停车位原点，将原点作为坐标系原点构建坐标系；获取传感器数据，对传感器数据进行数据融合处理，得到定位可利用的数据；基于定位可利用的数据中的车辆里程、速度、方向盘转角、档位、陀螺仪数据和加速度，推导出当前的车辆位置；基于定位可利用的数据中的摄像头识别的车位信息和超声波雷达识别的障碍物信息在坐标系下的位置，对推导出的当前的车辆位置进行补正，输出当前的车辆精确位置。本发明能够通过图像识别以及超声波识别技术，识别出车位信息以及周边障碍物等信息，在定位时再和其他传感器数据进行融合，实现实时高精度定位。 【EN】The invention discloses a real-time high-precision positioning method and a device for full-automatic parking, wherein the method comprises the following steps: setting an origin of a parking space, and constructing a coordinate system by taking the origin as the origin of the coordinate system; acquiring sensor data, and performing data fusion processing on the sensor data to obtain positioning available data; deriving a current vehicle position based on vehicle mileage, speed, steering wheel angle, gear, gyroscope data, and acceleration in the location-available data; and correcting the deduced current vehicle position based on the positions of the parking space information identified by the camera and the obstacle information identified by the ultrasonic radar in the available positioning data in the coordinate system, and outputting the current accurate position of the vehicle. The invention can identify the parking space information, the surrounding obstacles and other information through image identification and ultrasonic identification technology, and then fuse the information with other sensor data during positioning, thereby realizing real-time high-precision positioning.

摘要：【中文】本发明公开了一种抗噪声数据的人脸识别模型训练方法，包括如下步骤：S1、构建人脸识别数据集，按照个体ID进行分类；S2、对步骤S1所获得的数据集进行预处理；S3、根据步骤S2的预处理结果，构建20层的ResNet网络结构提取特征，使用ArcFace损失函数训练人脸识别模型；S4、统计最近一定数量Batch中cosθ值的分布情况，以cosθ值为横坐标，同一个cosθ的数量为纵坐标，绘制曲线图P；S5、重复步骤S4，直到cosθ分布呈现两个高斯分布，分别计算两个高斯分布顶点横坐标α、β和交点横坐标λ；S6、根据α、β值的大小，判断模型拟合情况，动态调整损失函数。本发明能够自动过滤噪声数据扰，提高模型训练效果。 【EN】The invention discloses a training method of a face recognition model of anti-noise data, which comprises the following steps of S1, constructing a face recognition data set, classifying according to individual IDs, S2, preprocessing the data set obtained in the step S1, S3, constructing a 20-layer ResNet network structure according to a preprocessing result of the step S2, extracting features, training the face recognition model by using an ArcFace loss function, S4, counting the distribution situation of cos theta values in a certain number of recent Batch, drawing a curve graph P by using the cos theta values as horizontal coordinates and the same cos theta values as vertical coordinates, S5, repeating the step S4 until the cos theta values show two Gaussian distributions, respectively calculating the horizontal coordinates α, β of the two Gaussian distribution vertexes and the horizontal coordinate lambda of an intersection point, S6, dynamically adjusting the loss function according to the sizes of the α and β values, and automatically filtering the loss data disturbance and improving the training effect of the model.

摘要：【中文】本发明涉及一种模块化高效热交换结构，包括若干散热器模块，散热器模块包括由若干三维变形管并列排布而成的管束，管束的外侧覆盖有散热器壳体，散热器壳体的一端设置有进口水管，散热器壳体的另一端设置有出口水管，相邻散热器模块的进口水管与出口水管之间可拆卸连接有连通弯头，进口水管与出口水管分别与管束的两端相连通，散热器壳体远离进口水管一端的上端面开设有进风口，散热器壳体靠近进口水管的一端开设有出风口，进风口安装有风口竖直向下设置的风机。通过采用上述技术方案，减少相对应的经济损失，以及具有适应不同功率需要下的热交换的优点。有效的增大了传热温差，从而增大了该设备的换热功率；可以节约设备制作材料。 【EN】The invention relates to a modular efficient heat exchange structure which comprises a plurality of radiator modules, wherein each radiator module comprises a pipe bundle formed by arranging a plurality of three-dimensional deformation pipes in parallel, a radiator shell covers the outer side of each pipe bundle, an inlet water pipe is arranged at one end of each radiator shell, an outlet water pipe is arranged at the other end of each radiator shell, a communication elbow is detachably connected between the inlet water pipe and the outlet water pipe of each adjacent radiator module, the inlet water pipe and the outlet water pipe are respectively communicated with two ends of each pipe bundle, an air inlet is formed in the upper end face of one end, far away from the inlet water pipe, of each radiator shell, an air outlet is formed in one end, close to. By adopting the technical scheme, the corresponding economic loss is reduced, and the heat exchange device has the advantages of adapting to heat exchange under different power requirements. The heat transfer temperature difference is effectively increased, so that the heat exchange power of the equipment is increased; the equipment manufacturing material can be saved.

摘要：【中文】本发明提供一种图像识别方法、装置及系统，该方法，包括：获取当前的高精度地图；获取高精度地图中标注的地图要素；根据标注的地图要素逆向生成目标样本数据；将目标样本数据输入到目标深度学习模型的样本库中，通过目标深度学习模型识别出图像获取设备所拍摄的图像中的地图要素。本发明可以从高精度地图中获取标注的地图要素，然后基于该标注的地图要素逆向生成目标样本数据，利用该目标样本数据对深度学习模型进行训练，从而可以使得训练出的目标深度学习模型具有更高的识别精度。 【EN】The invention provides an image identification method, device and system, wherein the method comprises the following steps: acquiring a current high-precision map; obtaining map elements marked in a high-precision map; reversely generating target sample data according to the marked map elements; target sample data is input into a sample library of a target deep learning model, and map elements in an image shot by the image acquisition equipment are identified through the target deep learning model. According to the method, the marked map elements can be obtained from the high-precision map, target sample data are generated reversely based on the marked map elements, and the deep learning model is trained by using the target sample data, so that the trained deep learning model has higher identification precision.

摘要：【中文】本发明提供了一种适用于涂胶和打螺丝的装置，包括机柜、打螺丝装置、涂胶装置、夹具以及控制装置；所述打螺丝装置、涂胶装置、夹具都安装在所述机柜的内部；所述打螺丝装置、涂胶装置分别与控制装置电连接，其中机柜中设置有两台螺丝枪和一台涂胶机，能够同时进行产品涂胶和打螺丝的作业，提供了作业效率，同时本发明将整套装置固定在一个机柜内，并通过垂直交叉的多滑轨结构设置充分利用了设备的内部空间，且能够更直观且有效地确保各个机构的垂直度符合工艺要求，使整个装置的体积变小，提高了设备空间利用率，实用性强。 【EN】The invention provides a device suitable for gluing and screwing, which comprises a cabinet, a screwing device, a gluing device, a clamp and a control device, wherein the gluing device is arranged on the cabinet; the screw driving device, the gluing device and the clamp are all arranged inside the cabinet; the screw driving device and the gluing device are respectively and electrically connected with the control device, two screw guns and a gluing machine are arranged in the cabinet, so that the gluing and screw driving operations of products can be simultaneously carried out, the operation efficiency is improved, meanwhile, the whole device is fixed in the cabinet, the internal space of equipment is fully utilized through the arrangement of the vertically crossed multi-slide rail structure, the verticality of each mechanism can be more intuitively and effectively ensured to meet the technological requirements, the size of the whole device is reduced, the space utilization rate of the equipment is improved, and the practicability is high.

摘要：【中文】本发明涉及电池管理技术领域，其公开了一种基于自循环校正的锂电池SOH估算系统及方法，通过对SOH估算值的自循环校正，在确保预测结果收敛的前提下，提高锂电池SOH估算的精确度，提升电池管理系统的可靠性。该系统包括：数据采集模块，用于采集锂电池系统的运行数据；SOH估算模块，用于利用锂电池系统的运行数据进行SOH估算；自循环校正模块，用于通过自循环充放电数据分析，对比SOH估算值与预先获取的同批次电池循环充放电实验数据之间的误差，对电池SOH进行校正。 【EN】The invention relates to the technical field of battery management, and discloses a lithium battery SOH estimation system and method based on self-circulation correction. The system comprises: the data acquisition module is used for acquiring the operation data of the lithium battery system; the SOH estimation module is used for estimating the SOH by utilizing the operation data of the lithium battery system; and the self-circulation correction module is used for comparing errors between the SOH estimated value and the pre-acquired battery circulation charge-discharge experimental data of the same batch through self-circulation charge-discharge data analysis and correcting the SOH of the battery.

摘要：【中文】本发明涉及电池管理技术领域，其公开了一种磷酸铁锂电池SOC估算系统，实现对磷酸铁锂电池的荷电状态的实时、可靠估算。该系统包括：数据采集模块，用于采集磷酸铁锂电池的运行状态参数，包括电压、电流和运行时间值；SOC计算模块，用于根据所述运行状态参数获取电池当前SOC值：当判定当前电量位于第一阈值与第二阈值的区间范围内时，通过安时积分法计算当前SOC值，否则，通过开路电压查询当前SOC值；温度补偿模块，用于采集电池运行实时温度，根据实时温度为SOC计算模块提供相应的温度补偿信号，辅助SOC计算模块获取电池当前SOC值；SOC校正模块，用于根据电池电压变化与容量变化的比值与SOC值呈现的关系，对SOC计算模块获取的SOC值进行校正。 【EN】The invention relates to the technical field of battery management, and discloses a lithium iron phosphate battery SOC estimation system which can realize real-time and reliable estimation of the state of charge of a lithium iron phosphate battery. The system comprises: the data acquisition module is used for acquiring the running state parameters of the lithium iron phosphate battery, including voltage, current and running time values; the SOC calculation module is used for acquiring the current SOC value of the battery according to the running state parameters: when the current electric quantity is judged to be within the interval range of the first threshold value and the second threshold value, calculating the current SOC value through an ampere-hour integration method, and otherwise, inquiring the current SOC value through open-circuit voltage; the temperature compensation module is used for acquiring the running real-time temperature of the battery, providing a corresponding temperature compensation signal for the SOC calculation module according to the real-time temperature and assisting the SOC calculation module to acquire the current SOC value of the battery; and the SOC correction module is used for correcting the SOC value acquired by the SOC calculation module according to the relation between the ratio of the voltage change of the battery to the capacity change and the SOC value.

摘要：【中文】本发明公开的一种多目标测控通信天线阵列光纤闭环校准方法，旨在解决天线误差校准测试参数多、天线单元互耦在内的各接收通道的不一致性，测试量大的问题。本发明通过下述技术方案予以实现：标校信号产生单元基于起始指令产生接收和发射校准基带信号，接收校准基带信号通过标校变频单元将接收校准信号上变频至射频并传输至光传输设备，通过光纤传输至天线子阵的光收/发模块，并分别耦合至各接收通道，处理后通过光纤传输至幅相一致性测量单元，实现接收通道的闭环校准；发射校准基带信号经开关网络与耦合器输出至收/发校准开光网络，传输至后端光传输设备，通过下变频得到发射校准基带信号并传输至幅相一致性测量单元，实现发射通道的闭环校准。 【EN】The invention discloses a multi-target measurement and control communication antenna array optical fiber closed-loop calibration method, which aims to solve the problems of more antenna error calibration test parameters, inconsistency of receiving channels with antenna units coupled with each other and large test quantity. The invention is realized by the following technical scheme: the calibration signal generating unit generates a receiving calibration baseband signal and a transmitting calibration baseband signal based on the initial instruction, the receiving calibration baseband signal up-converts the receiving calibration signal to radio frequency through the calibration frequency conversion unit and transmits the radio frequency to the optical transmission equipment, the radio frequency is transmitted to the optical receiving/transmitting module of the antenna subarray through the optical fiber and is respectively coupled to each receiving channel, and the received calibration signal is transmitted to the amplitude-phase consistency measuring unit through the optical fiber after being processed, so that the closed-loop calibration of the receiving channels is realized; the emission calibration baseband signal is output to a receiving/emission calibration open optical network through a switch network and a coupler, transmitted to rear-end optical transmission equipment, subjected to down-conversion to obtain an emission calibration baseband signal, and transmitted to an amplitude-phase consistency measuring unit, so that closed-loop calibration of an emission channel is realized.

摘要：【中文】本发明公开了一种新型耕种施肥一体机，涉及农业机械技术领域，该新型耕种施肥一体机，包括机架，所述机架的外部从右至左依次安装有地轮、施肥器、旋耕机、开槽块、播种机构、覆土板、镇土机构和开沟机构，通过设置的播种机构，使用时，通过三个不同位置的离合器，可使三个不同位置轴杆获取相应的动力，进而带动不同的导轮转动，同时在通孔的作用下，可使装置能够将不同品种的种子送入至下料管完成播种步骤，从而使装置能能够实现一机多用，大大降低了农作物播种的成本，通过设置的镇土机构，使用时，可转动调节杆使限位管沿着固定管进行移动，进而可改变镇压轮与地面之间的距离，以便使镇压轮能够对不同厚度的松土进行镇压。 【EN】The invention discloses a novel integrated machine for tilling and fertilizing, which relates to the technical field of agricultural machinery, and comprises a frame, wherein a land wheel, a fertilizer applicator, a rotary cultivator, a grooving block, a seeding mechanism, an earth covering plate, an earth pressing mechanism and a ditching mechanism are sequentially arranged outside the frame from right to left, and when the novel integrated machine for tilling and fertilizing is used, through the arranged seeding mechanism, shaft levers at three different positions can obtain corresponding power through clutches at three different positions so as to drive different guide wheels to rotate, and simultaneously, under the action of a through hole, the device can feed different types of seeds into a discharging pipe to complete a seeding step, so that the device can realize multiple purposes, the cost of crop seeding is greatly reduced, through the arranged earth pressing mechanism, when the novel integrated machine for tilling and fertilizing can rotate an adjusting rod so as to enable a limiting pipe to move along a fixed pipe so as to change the distance between the earth pressing wheel and the ground, so that the press wheel can press the loose soil with different thicknesses.