摘要：【中文】本发明提供了一种采用注塑方式制作带外部散热片的功率集成电路部件的方法，把自带散热片的功率集成电路封装体与导电片形成的整体注塑镶件、外部散热片等放入到壳体注塑模具中，向壳体注塑模具中注塑高温塑料形成壳体，脱模后即完成功率集成电路部件的制作。该方法减少了粘接外部散热片、灌封密封胶、粘接保护盖板等生产环节，具有降低生产成本，高效生产，制作出的产品一致性好并可适用于多种应用方式且能在各种环境温度尤其是高温下满功率工作。 【EN】The invention provides a method for manufacturing a power integrated circuit part with an external radiating fin by adopting an injection molding mode, wherein an integral injection molding insert formed by a power integrated circuit packaging body with the radiating fin and a conducting strip, the external radiating fin and the like are placed in a shell injection mold, high-temperature plastic is injected into the shell injection mold to form a shell, and the manufacture of the power integrated circuit part is finished after demolding. The method reduces the production links of adhering external radiating fins, encapsulating sealant, adhering protective cover plates and the like, has the advantages of reducing production cost, realizing efficient production, and manufacturing products with good consistency, can be suitable for various application modes and can work at full power at various environmental temperatures, particularly high temperature.

摘要：【中文】本申请提供一种模型训练的方法和装置，所述方法包括：监控模型的训练进度和训练进度对应的模型的损失值，在所述模型的训练进度到达预设的检查点的情况下，保存所述模型在检查点对应的模型参数，根据所述模型的损失值，在确定所述模型训练异常的情况下，直接设定所述模型的模型参数为所述模型在检查点对应的模型参数。直接调整所述模型的训练参数，使得模型的训练过程具有更大的随机性，从而继续进行模型的训练，确保模型的训练过程不停顿，能够加快模型的训练速度。 【EN】The application provides a method and a device for model training, wherein the method comprises the following steps: monitoring the training progress of the model and the loss value of the model corresponding to the training progress, saving the model parameters of the model corresponding to the check point when the training progress of the model reaches a preset check point, and directly setting the model parameters of the model as the model parameters of the model corresponding to the check point when the abnormal training of the model is determined according to the loss value of the model. The training parameters of the model are directly adjusted, so that the training process of the model has higher randomness, the training of the model is continued, the training process of the model is ensured to be not stopped, and the training speed of the model can be accelerated.

摘要：【中文】本申请提供一种手动取词的方法及装置，用于取词模型，所述方法包括：获取目标图像，并对所述目标图像进行手部图像检测；在检测到所述目标图像包含手部图像的情况下，对所述手部图像进行指尖识别，确定指尖位置；确定所述指尖位置对应的候选区域，对所述指尖位置对应的候选区域进行处理，以生成目标词单元，从而实现根据手部的指尖精确定位至对应的词单元。 【EN】The application provides a method and a device for manually fetching words, which are used for fetching word models, wherein the method comprises the following steps: acquiring a target image, and carrying out hand image detection on the target image; under the condition that the target image is detected to contain a hand image, carrying out fingertip identification on the hand image, and determining a fingertip position; and determining a candidate region corresponding to the fingertip position, and processing the candidate region corresponding to the fingertip position to generate a target word unit, so that the purpose of accurately positioning to the corresponding word unit according to the fingertip of the hand is realized.

摘要：【中文】本申请提供一种二维码空间位置确认方法及装置，其中所述二维码空间位置确认方法包括：利用增强现实工具扫描二维码，并构建对应的三维空间；获取所述二维码在显示屏幕中至少三个不共线的识别点和所述二维码在所述三维空间中的三维特征点集合；在所述三维特征点集合中选取与所述识别点对应的目标三维特征点；根据所述目标三维特征点确定所述二维码在所述三维空间中的位置。通过本申请提供的方法，在实现识别二维码的基础上，还可以快速准确地确定二维码在三维空间中的位置和朝向，为后续开发提供了方便，提高了工作效率。 【EN】The application provides a method and a device for confirming a spatial position of a two-dimensional code, wherein the method for confirming the spatial position of the two-dimensional code comprises the following steps: scanning the two-dimensional code by using an augmented reality tool, and constructing a corresponding three-dimensional space; acquiring at least three non-collinear identification points of the two-dimensional code in a display screen and a three-dimensional feature point set of the two-dimensional code in the three-dimensional space; selecting a target three-dimensional feature point corresponding to the identification point from the three-dimensional feature point set; and determining the position of the two-dimensional code in the three-dimensional space according to the target three-dimensional feature point. By the method, on the basis of realizing the identification of the two-dimensional code, the position and the orientation of the two-dimensional code in the three-dimensional space can be quickly and accurately determined, convenience is provided for subsequent development, and the working efficiency is improved.

摘要：【中文】一种化机浆高浓滤液处理方法，包括如下步骤：步骤一：将化机浆生产线产生的高浓滤液送入化学浆热水槽中，加入洗涤用水混合；步骤二：将化学浆热水槽中的高浓滤液与洗涤用水的混合液抽出，送入鼓式真空洗浆机进行浆层过滤；步骤三：将经过鼓式真空洗浆机内浆层过滤的高浓滤液与洗涤用水的混合液送入黑液槽暂时储存；步骤四：将黑液槽中的液体抽出，经过机械过滤后送入碱回收车间蒸发工段，经过多效蒸发器进行蒸发浓缩；步骤五：经过蒸发浓缩到工艺要求后送入碱回收锅炉进行燃烧处理，融物进入绿液槽，经过苛化处理回收碱液。本发明能除去大部分的细小纤维，有效的减轻碱回收蒸发器结垢现象，提高蒸发效率，节约能源消耗。 【EN】A method for treating high-concentration filtrate of chemi-mechanical pulp comprises the following steps: the method comprises the following steps: sending high-concentration filtrate generated by a chemical-mechanical pulp production line into a chemical pulp hot water tank, and adding washing water for mixing; step two: pumping out a mixed solution of high-concentration filtrate and washing water in a chemical pulp hot water tank, and feeding the mixed solution into a drum type vacuum pulp washer for pulp layer filtration; step three: sending a mixed solution of high-concentration filtrate filtered by a pulp layer in a drum type vacuum pulp washer and washing water into a black liquid tank for temporary storage; step four: pumping out the liquid in the black liquid tank, mechanically filtering, sending the liquid to an evaporation section of an alkali recovery workshop, and carrying out evaporation concentration by a multi-effect evaporator; step five: after the process is carried out by evaporation and concentration, the mixture is sent into an alkali recovery boiler for combustion treatment, the molten substance enters a green liquor tank, and alkali liquor is recovered by causticization treatment. The invention can remove most of fine fibers, effectively reduce the scaling phenomenon of the alkali recovery evaporator, improve the evaporation efficiency and save the energy consumption.

摘要：【中文】本申请提供一种多文本分析方法及装置，其中所述多文本分析方法包括：将至少两个待分析文本和待回答问题输入预先训练好的文本分析模型中进行处理，生成所述待回答问题的至少两个候选答案和每个候选答案对应的候选答案向量；提取每个所述候选答案的特征向量；根据每个所述候选答案的所述特征向量和所述候选答案向量生成对应的拼接向量；根据所述拼接向量在所述候选答案中选取所述待回答问题的最佳答案。结合多种类型的特征之后选出最佳答案，减小了答案误判的可能性，提高了选择答案的准确率。 【EN】The application provides a multi-text analysis method and a multi-text analysis device, wherein the multi-text analysis method comprises the following steps: inputting at least two texts to be analyzed and questions to be answered into a text analysis model trained in advance for processing, and generating at least two candidate answers of the questions to be answered and a candidate answer vector corresponding to each candidate answer; extracting a feature vector of each candidate answer; generating a corresponding splicing vector according to the feature vector of each candidate answer and the candidate answer vector; and selecting the best answer of the question to be answered from the candidate answers according to the splicing vector. The best answer is selected after combining various types of features, the possibility of misjudgment of the answer is reduced, and the accuracy of answer selection is improved.

摘要：【中文】本发明公开了一种信号接收方法及系统，涉及通信技术领域。信号接收方法包括：接收携带业务数据的光信号，并将光信号转换成电压信号，其中，业务数据的传输速率处于限定范围内；对电压信号进行模数转换，得到采样信号，其中，采样速率大于限定范围的上限；根据业务数据的传输速率，对采样信号进行均衡滤波处理；对均衡滤波处理的输出数据进行数模转换，得到恢复数据并输出到相应的物理媒介适配接口。本发明能够接收多种传输速率的业务数据，并提高链路性能。 【EN】The invention discloses a signal receiving method and a signal receiving system, and relates to the technical field of communication. The signal receiving method comprises the following steps: receiving an optical signal carrying service data, and converting the optical signal into a voltage signal, wherein the transmission rate of the service data is within a limited range; performing analog-to-digital conversion on the voltage signal to obtain a sampling signal, wherein the sampling rate is greater than the upper limit of a limited range; according to the transmission rate of the service data, carrying out balanced filtering processing on the sampling signal; and D/A conversion is carried out on the output data subjected to the equalization filtering processing to obtain recovery data and the recovery data is output to a corresponding physical medium adaptive interface. The invention can receive the service data with various transmission rates and improve the link performance.

摘要：【中文】本申请提供一种语句选词模型的训练方法及装置，其中所述语句选词模型的训练方法包括：获取预训练的语句选词模型和样本数据，其中，所述样本数据包括待选词语句和至少两个候选词，所述至少两个候选词中包括一个目标词；将所述待选词语句与每个候选词分别拼接，获得拼接语句，并根据每个候选词的字数在所述拼接语句中生成对应数量的掩膜标签；将每个包括所述候选词和所述掩膜标签的拼接语句输入所述语句选词模型，生成与每个所述拼接语句对应的候选词向量和掩膜向量；根据所述候选词向量和所述掩膜向量生成每个所述拼接语句对应的分值；根据每个拼接语句的分值调整所述语句选词模型的参数，直至所述目标词对应的拼接语句的分值最高。 【EN】The application provides a training method and a device of a sentence word selection model, wherein the training method of the sentence word selection model comprises the following steps: obtaining a pre-trained sentence word selection model and sample data, wherein the sample data comprises a sentence to be word selected and at least two candidate words, and the at least two candidate words comprise a target word; splicing the sentence of the words to be selected with each candidate word respectively to obtain a spliced sentence, and generating a corresponding number of mask labels in the spliced sentence according to the word number of each candidate word; inputting each spliced statement comprising the candidate words and the mask label into the statement word selection model, and generating candidate word vectors and mask vectors corresponding to each spliced statement; generating a score corresponding to each spliced statement according to the candidate word vector and the mask vector; and adjusting parameters of the sentence word selection model according to the score of each spliced sentence until the score of the spliced sentence corresponding to the target word is the highest.

摘要：【中文】本申请提供一种成语推荐模型的训练方法及装置，其中所述方法包括：获取训练样本和对应的训练标签，其中，所述训练样本包括含有缺失空白的原始语料以及多个常见成语，所述训练标签包括每个常见成语在每个缺失空白中对应的推荐分数；通过所述训练样本和对应的训练标签对成语推荐模型进行训练，得到所述成语推荐模型，所述成语推荐模型使得所述训练样本与所述训练标签相关联。本申请通过大规模的语料训练，实现了根据上下文语义信息对常见成语进行打分的智能模型，通过该模型能够帮助用户在写作时进行常见成语的推荐，使得用户能够又快又好的完成高质量文笔的文章，提升了用户的体验。 【EN】The application provides a method and a device for training a idiom recommendation model, wherein the method comprises the following steps: acquiring a training sample and a corresponding training label, wherein the training sample comprises an original corpus containing missing blanks and a plurality of common idioms, and the training label comprises a recommendation score corresponding to each common idiom in each missing blank; and training a idiom recommendation model through the training samples and the corresponding training labels to obtain the idiom recommendation model, wherein the idiom recommendation model enables the training samples to be associated with the training labels. According to the method and the device, the intelligent model for scoring the common idioms according to the context semantic information is realized through large-scale corpus training, the user can be helped to recommend the common idioms when writing through the model, the user can quickly and well complete the high-quality text pen article, and the user experience is improved.

摘要：【中文】本发明公开了一种开口销拆卸器，包括夹钳和拔销机构，夹钳用于夹持销轴，拔销机构包括连接架、手柄和牵引部，连接架的底部设置于夹钳上，连接架的顶部与手柄的中部转动连接，手柄的一端连接有牵引部，牵引部用于连接销轴上的开口销。本发明通过牵引部连接开口销端部的圆孔，扳动夹钳夹住销轴，防止拆卸开口销过程圆轴转动，扳动拔销机构的手柄，拉出开口销，有效解决了使用撬棍和手锤敲打开口销的操作难度，减低了劳动强度，且适用于多种规格开口销，操作方便，提高了拆卸效率。 【EN】The invention discloses a cotter detacher which comprises a clamp and a cotter pulling mechanism, wherein the clamp is used for clamping a pin shaft, the cotter pulling mechanism comprises a connecting frame, a handle and a traction part, the bottom of the connecting frame is arranged on the clamp, the top of the connecting frame is rotatably connected with the middle of the handle, one end of the handle is connected with the traction part, and the traction part is used for connecting the cotter on the pin shaft. According to the invention, the traction part is connected with the round hole at the end part of the cotter pin, the clamp is pulled to clamp the pin shaft, the round shaft is prevented from rotating in the cotter pin dismounting process, the handle of the cotter pin pulling mechanism is pulled to pull out the cotter pin, the operation difficulty of knocking the cotter pin by using a crowbar and a hand hammer is effectively solved, the labor intensity is reduced, the cotter pin dismounting device is suitable for cotter pins of various specifications, the operation is convenient, and the dismounting efficiency is improved.