Towards fast focal mechanism inversion of shallow crustal earthquakes in the Chinese mainland

We have developed an automatic regional focal mechanism inversion system based on the Earthquake Rapid Report(ERR) system and the real-time three-component seismic waveform stream of 1 000 broadband seismic stations provided by the China Earthquake Networks Center(CENC). The system can rapidly provide a double couple solution and centroid depth within 5–15 min after receiving earthquake information from the ERR system.The data processing is triggered by earthquake information obtained from the ERR system. The system is capable of determining the focal mechanism of all shallow-depth earthquakes in the Chinese mainland with a magnitude of 5.5–6.5. It utilizes waveform data recorded by seismic stations located within 500 km from the epicenter,enabling the reporting of a focal mechanism solution within 5–15 min of an earthquake occurrence. Additionally,the system can assign a corresponding grade(A B C) to the focal mechanism solution. We processed a total of 301earthquakes that occurred from 2021 to June 2022, and after the quality control, 166 of them were selected.These selected solutions were manually checked, and 160 of them were compiled in a focal mechanism catalog.This catalog can be conveniently downloaded online via the Internet. The automatic focal mechanism solution of earthquakes in eastern China exhibits a good agreement with that provided by the Global Centroid Moment Tensor(GCMT), when available. The average Kagan angle between this catalog and GCMT is 22°, and the average difference in MWis 0.17. Furthermore, compared with GCMT, the minimum magnitude of our catalog has been reduced from approximately 5.0 to 4.0. The correlation between the centroid depth and crustal thickness in the Chinese mainland confirms the distribution of the centroid depth.

Automatic Keyphrase Extraction from Scientific Chinese Medical Abstracts Based on Character-Level Sequence Labeling

Purpose:Automatic keyphrase extraction(AKE)is an important task for grasping the main points of the text.In this paper,we aim to combine the benefits of sequence labeling formulation and pretrained language model to propose an automatic keyphrase extraction model for Chinese scientific research.Design/methodology/approach:We regard AKE from Chinese text as a character-level sequence labeling task to avoid segmentation errors of Chinese tokenizer and initialize our model with pretrained language model BERT,which was released by Google in 2018.We collect data from Chinese Science Citation Database and construct a large-scale dataset from medical domain,which contains 100,000 abstracts as training set,6,000 abstracts as development set and 3,094 abstracts as test set.We use unsupervised keyphrase extraction methods including term frequency(TF),TF-IDF,TextRank and supervised machine learning methods including Conditional Random Field(CRF),Bidirectional Long Short Term Memory Network(BiLSTM),and BiLSTM-CRF as baselines.Experiments are designed to compare word-level and character-level sequence labeling approaches on supervised machine learning models and BERT-based models.Findings:Compared with character-level BiLSTM-CRF,the best baseline model with F1 score of 50.16%,our character-level sequence labeling model based on BERT obtains F1 score of 59.80%,getting 9.64%absolute improvement.Research limitations:We just consider automatic keyphrase extraction task rather than keyphrase generation task,so only keyphrases that are occurred in the given text can be extracted.In addition,our proposed dataset is not suitable for dealing with nested keyphrases.Practical implications:We make our character-level IOB format dataset of Chinese Automatic Keyphrase Extraction from scientific Chinese medical abstracts(CAKE)publicly available for the benefits of research community,which is available at:https://github.com/possible1402/Dataset-For-Chinese-Medical-Keyphrase-Extraction.Originality/value:By designing comparative experiments,our study demonstrates that character-level formulation is more suitable for Chinese automatic keyphrase extraction task under the general trend of pretrained language models.And our proposed dataset provides a unified method for model evaluation and can promote the development of Chinese automatic keyphrase extraction to some extent.

Development of Automatic Programming System for NC Laser Cutting of Chinese Characters

This Paper presents an automatic programming system developed for NC laser onning of Chinese charaCterswhich combines AutoCAD with NC Under Windows environment, and generates numerical controlled Programs by extracting outline (stroke) figures of Chinese Characters from outline character base, optimizing cutting routes, transferring them to AutoCAD for editing operations, such as rotation transfer, enlarging an mirroring, the system can transferto a laser cutting machine througy RS-232 series interface to accomplish laser cutting of large Chinese characters.

Automatic Cooking Robot for Chinese Dishes

Until now, most of the kitchen works are done manually, which often make people bored and suffer from cooking oil smoke pollution. With the development of the robotic technology, it becomes more and more urgent for the appearance of the automatic cooking machines that can substitute man for most of those works. With this aim in mind, in this paper, a kind of automatic cooking robot is presented, which mainly consists of five parts ： the wok mechanism, the stirring-fry and dispersing mechanism, the feeding mechanism, fire control system and the assistant ingredients processing mechanism. Experiment results have proved that the robot has achieved the goal with respect to appearance, smell, and taste of the dishes cooked by the robot, according to the master cook＇s view.

Effective Analysis of Chinese Word-Segmentation Accuracy

Automatic word-segmentation is widely used in the ambiguity cancellation when processing large-scale real text,but during the process of unknown word detection in Chinese word segmentation,many detected word candidates are invalid.These false unknown word candidates deteriorate the overall segmentation accuracy,as it will affect the segmentation accuracy of known words.In this paper,we propose several methods for reducing the difficulties and improving the accuracy of the word-segmentation of written Chinese,such as full segmentation of a sentence,processing the duplicative word,idioms and statistical identification for unknown words.A simulation shows the feasibility of our proposed methods in improving the accuracy of word-segmentation of Chinese.

A method for improving the accuracy of automatic indexing of Chinese-English mixed documents

Purpose: The thrust of this paper is to present a method for improving the accuracy of automatic indexing of Chinese-English mixed documents.Design/methodology/approach: Based on the inherent characteristics of Chinese-English mixed texts and the cybernetics theory,we proposed an integrated control method for indexing documents. It consists of 'feed-forward control','in-progress control' and 'feed-back control',aiming at improving the accuracy of automatic indexing of Chinese-English mixed documents. An experiment was conducted to investigate the effect of our proposed method.Findings: This method distinguishes Chinese and English documents in grammatical structures and word formation rules. Through the implementation of this method in the three phases of automatic indexing for the Chinese-English mixed documents,the results were encouraging. The precision increased from 88.54% to 97.10% and recall improved from97.37% to 99.47%.Research limitations: The indexing method is relatively complicated and the whole indexing process requires substantial human intervention. Due to pattern matching based on a bruteforce(BF) approach,the indexing efficiency has been reduced to some extent.Practical implications: The research is of both theoretical significance and practical value in improving the accuracy of automatic indexing of multilingual documents(not confined to Chinese-English mixed documents). The proposed method will benefit not only the indexing of life science documents but also the indexing of documents in other subject areas.Originality/value: So far,few studies have been published about the method for increasing the accuracy of multilingual automatic indexing. This study will provide insights into the automatic indexing of multilingual documents,especially Chinese-English mixed documents.

Adjusting precipitation measurements from the TRwS204 automatic weighing gauge in the Qilian Mountains, China

With the popularity of the automatic precipitation gauges in national weather stations,testing their performance and adjusting their measurements are top priorities. Additionally,because different climatic conditions may have different effects on the performance of the precipitation gauges, it is also necessary to test the gauges in different areas. This study mainly analyzed precipitation measurements from the single-Altershielded TRwS204 automatic weighing gauge(TRwS_(SA)) relative to the adjusted manual measurements(reference precipitation) from the Chinese standard precipitation gauge in a doublefence wind shield(CSPG_(DF)) in the Hulu watershed in the Qilian Mountains, China. The measurements were compared over the period from August 2014 to July2017, and the transfer function derived from the work by Kochendorfer et al.(2017 a) for correcting windinduced losses was applied to the TRwS_(SA) measurements. The results show that the average loss of TRwS_(SA) measurements relative to the reference precipitation decreased from 0.55 mm(10.7%) to 0.51 mm(9.9%) for rainfall events, from 0.35 mm(8.5%)to 0.22 mm(5.3%) for sleet events, and from 0.49 mm(18.9%) to 0.33 mm(12.7%) for snowfall events after adjustment. The uncorrected large biases of TRwS_(SA) measurements are considered to be mainly caused by specific errors of TRwS_(SA), different gauge orifice area and random errors. These types of errors must be considered when comparing precipitation measurements for different gauge types, especially in the mountains.

JSON-ASR:A lightweight data storage and exchange format for automatic systematic reviews of TCM

Objectives:The aim of this study was to investigate and develop a data storage and exchange format for the process of automatic systematic reviews(ASR)of traditional Chinese medicine(TCM).Methods:A lightweight and commonly used data format,namely,JavaScript Object Notation(JSON),was introduced in this study.We designed a fully described data structure to collect TCM clinical trial information based on the JSON syntax.Results:A smart and powerful data format,JSON-ASR,was developed.JSON-ASR uses a plain-text data format in the form of key/value pairs and consists of six sections and more than 80 preset pairs.JSON-ASR adopts extensible structured arrays to support the situations of multi-groups and multi-outcomes.Conclusion:JSON-ASR has the characteristics of light weight,flexibility,and good scalability,which is suitable for the complex data of clinical evidence.

日光温室薄膜全自动清洗机研制与试验

针对日光温室薄膜人工清洗劳动强度大、自动化清洗设备缺乏的现状,该研究设计了一种日光温室薄膜全自动清洗机。为减小整机质量并降低成本,采用双蜗轮蜗杆电机分别驱动清洗主机的毛刷与爬升轴;设计地面移位换行装置,用以承载清洗主机沿温室长度方向移动;为实现自动换行清洗,设计棚顶移位换行装置,调整棚顶吊绳安装高度,避免吊绳在换行作业时与棚面产生干涉而影响换行质量或损伤薄膜。基于多传感器融合与数据校验技术,保证清洗主机、棚顶与地面移位装置间的协作实时性、一致性及可靠性。为验证设计方案的合理性与可行性,以参照日光温室跨度、脊高、肩高参数10:1制作模型温室,按外形尺寸5:1加工清洗样机,并进行对齐、倾斜偏移及清洗效果验证试验。结果表明,清洗主机升降时的水平偏移量在±3°以内,左右偏移量在±7 mm以内(换行操作时各偏移量无累积);地面与棚顶移位换行装置的单次换行误差与多次换行累计误差均在1 mm左右;毛刷材料选型与关键参数设计合理,可对薄膜表面灰尘进行有效清洗,洗净率为95.4%,清洗效果明显。该清洗机在丰富国内日光温室清洗装备类型的同时,有效解决了团队前期研发清洗机质量大、易伤膜、自动化水平低等问题,为温室大棚薄膜相关清洗设备的设计研发提供了参考。

子午工程二期自动磁通门经纬仪

地磁台站观测有相对记录和绝对观测2种,相对记录数据需经基线值改正后方可得到地磁场要素的实际值.磁通门经纬仪是获取基线值的基本设备之一,在地磁观测中发挥着重要作用.目前台站采用的磁通门经纬仪需人工操作,人为因素会对观测数据质量产生影响,因在偏远地区和环境恶劣地区无法实现地磁绝对观测,造成地磁绝对观测的空间覆盖空白.子午工程二期将在我国内地地区首次开展自动化磁偏角和磁倾角地磁绝对观测,必将推动地磁绝对观测技术创新与发展.自动磁通门经纬仪是我国自主研发的磁偏角和磁倾角自动化观测设备,采用无磁材料和压电电机,通过优化设计与改进加工工艺,以及引入多参量误差补偿算法,有效克服和消除了系统误差,提高了测量精度.该设备在河北涉县台、吉林合隆台、陕西乾陵台和北京白家疃台等台站开展了长时间的实际观测,并与台站的人工磁通门经纬仪观测结果进行了对比.结果表明,自动磁通门经纬仪的主要性能指标达到了人工磁通门经纬仪水平.该设备也通过了中国地震局前兆设备入网测试,功能和性能指标符合地磁台站地磁绝对观测要求.

基于特征融合的汉语被动句自动识别研究

汉语中的被动句根据有无被动标记词可分为有标记被动句和无标记被动句。由于其形态构成复杂多样,给自然语言理解带来很大困难,因此实现汉语被动句的自动识别对自然语言处理下游任务具有重要意义。该文构建了一个被动句语料库,提出了一个融合词性和动词论元框架信息的PC-BERT-CNN模型,对汉语被动句进行自动识别。实验结果表明,该文提出的模型能够准确地识别汉语被动句,其中有标记被动句识别F1值达到98.77%,无标记被动句识别F1值达到96.72%。

古汉语通假字资源库的构建及应用研究

古籍文本中的文字通假现象较为常见,这不仅为人理解文意造成了困难,也是古汉语信息处理面临的一项重要挑战。为了服务于通假字的人工判别和机器处理,该文构建并开源了一个多维度的通假字资源库,包括语料库、知识库和评测数据集三个子库。其中,语料库收录11000余条包含通假现象详细标注的语料;知识库以汉字为节点,通假和形声关系为边,从字音、字形、字义多个角度对通假字与正字的属性进行加工,共包含4185个字节点和8350对关联信息;评测数据集面向古汉语信息处理需求,支持通假字检测和正字识别两个子任务的评测,收录评测数据19678条。在此基础上,该文搭建了通假字自动识别的系列基线模型,并结合实验结果分析了影响通假字自动识别的因素与改进方法。进一步地,该文探讨了该资源库在古籍整理、人文研究和文言文教学中的应用。

基于二阶段对比学习的中文自动文本摘要方法研究

在中文自动文本摘要中,暴露偏差是一个常见的现象。由于中文文本自动摘要在序列到序列模型训练时解码器每一个词输入都来自真实样本,但是在测试时当前输入用的却是上一个词的输出,导致预测词在训练和测试时是从不同的分布中推断出来的,而这种不一致将导致训练模型和测试模型直接的差异。本文提出了一个两阶段对比学习框架以实现面向中文文本的生成式摘要训练,同时从摘要模型的训练以及摘要评价的建模进行对比学习。在大规模中文短文本摘要数据集(LCSTS)以及自然语言处理与中文计算会议的文本数据集(NLPCC)上的实验结果表明,相比于基线模型,本文方法可以获得更高的面向召回率的摘要评价方法(ROUGE)指标,并能更好地解决暴露偏差问题。

中药自动抓配系统的设计

随着中医药的发展,中药文化在传统医学体系中扮演着重要的角色,为了解决医药医生抓配中药困难的问题,设计了一款基于STM32的中药抓配系统。设备上位机接收医生开出的药单,通过串口通信下发指令给抓配设备的主控,抓药设备根据接收的指令,控制药柜前的横向与纵向步进电机、称重传感器等设备,实现对中药材的自动抓配。医生通过上位机对中药抓配过程进行远程监控。该系统通过数字化、智能化的手段,提高了中药抓配的精准性和效率,也促进了中医药的传统文化与现代科技的融合。

基于随机森林的中医电子病历数据分类系统设计

应用当前系统进行电子病历数据分类时,分类精度低,效率也低。为此设计基于随机森林的中医电子病历数据分类系统。以S3C2440A和PGL22G分别为核心设计运行控制和通信控制模块实现硬件部分对软件功能的支撑。结合中医电子病历数据特点,选择数据分类特征。由CART算法生成随机森林基础分类器,利用训练数据集对其训练,确定分类参数完成分类系统设计。通过实验测试,该系统的分类精度高于92%,且分类效果优于两对比系统,运行性能优越。

医院中药配方颗粒自动化调配药房调剂操作规程的应用

中药配方颗粒是根据中医药理论和临床应用需求,对中药饮片采用特殊的加工手法制成的,它保持了中药汤剂的特点,服用过程简单,易于携带、运输,逐步为广大患者认识和接受。自动化中药配方颗粒药房的建立和运行,应按照标准化调配和操作原则规范进行,调剂设备应符合中医临床用药习惯,调剂操作规程要防止调配差错、药品污染,使用的调剂软件对调剂过程实现可追溯。文章介绍了医院中药配方颗粒自动化调剂及其操作规程,对其中处方审核、调配、药品养护、贮存环境、人员等环节以及操作规程作了较为详细的说明,为推动自动化中药配方颗粒药房建设和管理提供参考。

用于中药中农药残留检测的全自动快速样品处理仪验证与评价

采用中药农药残留的样品处理方法作为验证方式,建立了全自动快速样品处理(QuEChERS)设备的验证与评价方法。验证的设备按照标准方法的步骤执行样品处理过程,对影响最终实验结果的基本性能和应用效果进行验证。验证的农药化合物在0.005~0.25μg/mL质量浓度范围内与其相应的色谱峰面积具有较好的线性关系,相关系数均大于0.99,检出限均为0.002 mg/kg。中药样品在不同浓度的加标水平下的加标平均回收率为71.6%~128.8%,相对标准偏差为0.1%~8.1%。全自动快速样品处理设备实验操作的一致性程度较高,设备在分液、移液和加盐等步骤的操作可以满足实验的要求。采用自动化样品处理方式应用于中药农药残留检测,能有效减少人为误差,提高检测的准确性、重现性和溯源性,为分析仪器的自动化和智能化技术验证提供了技术支持。

Automatic Extraction of Tongue Coatings from Digital Images:A Traditional Chinese Medicine Diagnostic Tool

In traditional Chinese medicine, the coating on the tongue is considered to be a reflection of various pathologic factors. However, the conventional method to examine the tongue lacks an acceptable standard and does not provide the means for sharing information. This paper describes a segmentation method to extract tongue coatings. First, the tongue body was extracted from the original image using the watershed transform. Then, a threshold method was applied to the image to eliminate the light from the camera flash. Finally, a threshold method using the Otsu model in combination with a splitting-merging method was used in the red, green, and blue （RGB） space to extract the thin coating. The combination of the above two methods is applied in the hue, saturation, and value （HSV） space to extract the thick coating. The feasibility of this method is tested by experiments, and the accuracy of segmentation is 95.9%.

基于移动云计算模式的中医典籍英译自动校准系统

现有的几种翻译校准系统难以有效计算中医典籍中英汉语义的相似性,为提高自动校准正确率,基于移动云计算模式设计中医典籍英译自动校准系统。确定收发器芯片组成结构。计算语义相似特征,以此建立中医典籍英译相似度预测模型,设计汉译英自适应校准算法,实现汉译英自动校准。根据实验结果可知,系统准确度约为87%,高出现有的三种自动校准系统,因此可知该系统提高了中医典籍英译自动校准的正确率。

医学汉语水平考试对话流完形填空智能出题系统研究

针对医学汉语水平考试(Medical Chinese Test,简称MCT)题库建设成本高昂及资料匮乏等问题,本文提出MCT对话流完形填空智能出题系统研发思路,旨在通过智能化方法实现对话流的高效生成和完形填空题的挖空出题。该系统主要包括对话流智能生成和完形填空智能出题两部分。在对话流智能生成部分,该系统先依据电子病历构建知识图谱,再采用图神经网络实现基于知识图谱的问题与问题链的生成,从而获得对话流数据;在完形填空智能出题部分,该系统首先基于多维复杂医学约束知识对对话流文本进行篇章语义解析,筛选出符合要求的对话流作为出题语料,然后进行对话流文本的知识标注,最后完成挖空并生成干扰项。结果表明,该系统能够有效地生成大量MCT对话流完形填空题目,经人工检测与评估,所生成的题目质量较高。