169results about How to "Maintain accuracy" patented technology

The invention discloses an intelligent visualization and text association method for a multi-modal knowledge graph, and relates to the technical field of information processing. According to the intelligent visualization and text association method for the multi-modal knowledge graph, multi-modal data in an internal database of an enterprise is obtained, and a multi-modal feature vector is generated by using a data preprocessing and feature extraction technology; the feature vectors are mapped to a unified vector space through a cross-modal embedding algorithm, semantic alignment, association and fusion among multi-modal data are realized, and a complete multi-modal knowledge graph is constructed. Through bidirectional association and multi-modal joint query, the information acquisition efficiency is improved, and visual display of the knowledge graph is realized by using the graph neural network. Meanwhile, the knowledge graph is dynamically updated through a graph reasoning algorithm, and the timeliness and accuracy of the knowledge graph are ensured. A more efficient solution is provided for an enterprise in the aspects of management and application of multi-modal data, and the intelligent process of knowledge management is promoted.

The invention discloses a multi-modal information fusion and dynamic updating group behavior analysis method. The method comprises the steps of information integration and dynamic updating, multi-modal information obtaining and information preprocessing, feature extraction, feature representation conversion, alignment and fusion, and multi-modal information real-time processing and dynamic updating through a stream processing framework. A self-learning mechanism and a self-optimization mechanism are introduced; group portrait generation: extracting population statistical characteristics, geographic position distribution, behavior and emotional psychological characteristics, social relationships, social media and economic characteristics from the multi-modal data to generate a group portrait; analyzing group behaviors; and outputting visualization, and outputting group behavior prediction results in a visualized manner, including a behavior trend graph, a group clustering graph, an association rule graph and an abnormal behavior detection graph. According to the method, multi-modal data from various sources and forms are effectively integrated and analyzed, self-learning and self-optimization mechanisms are introduced, and finally group behavior analysis is performed.

The invention relates to the technical field of camera adjustment, in particular to a camera adaptive adjustment method and system and a storage medium. The method comprises the following steps: acquiring a multi-source environment sensing data set and an initial environment image; performing environment feature recognition on the initial environment image according to the multi-source environment sensing data set to obtain a scene environment feature map; performing semantic scene segmentation on the scene environment feature map to obtain a semantic scene macroscopic feature map; performing time sequence attention target tracking on the scene environment feature map to obtain a target vehicle motion feature map; performing feature constraint fusion on the semantic scene macroscopic feature map and the target vehicle motion feature map to obtain a target vehicle-scene motion semantic map; and performing parameter adjustment on the camera according to the target vehicle-scene motion semantic map to obtain a camera parameter adjustment strategy set. According to the invention, the performance and the image quality of the camera in a changeable environment are improved.

The invention discloses a power grid operation and maintenance deployment method and system based on a knowledge graph, and relates to the technical field of power grid operation and maintenance deployment, and the method comprises the steps: collecting key data, and carrying out the preprocessing of the key data; extracting entities from the key data, identifying relationships between the entities, constructing a knowledge graph based on the relationships between the entities, and storing the knowledge graph in a graph database; dynamically identifying newly appearing entities based on the key data, synchronously adjusting the entity relationship, and updating the knowledge graph based on the new entities and the entity relationship; extracting training data from the knowledge graph, and constructing a graph neural network model based on the training data; and predicting potential faults in the power system based on the knowledge graph and the graph neural network model, and generating corresponding decision suggestions. According to the method, the entities are extracted from the key data and the relationship between the entities is identified, so that the structured expression of the complex relationship in the power system is realized, and the multi-source heterogeneous data is converted into operable knowledge graph nodes and edges.

A three-dimensional target detection and tracking method fusing multi-sensor information belongs to the technical field of computer vision, and can make up for the defect of a single mode by collecting camera and radar image information and fusing camera images and laser radar point clouds. Multi-target tracking is carried out by using a multi-data coupling method of 2D + 3D data, a remote object is identified through a camera, when the object enters an induction range of the laser radar, the laser radar can carry out more accurate positioning, and position estimation of the target in a 2D image is corrected by using a more accurate 3D position provided by the laser radar. The method is used for solving the problems of inaccurate target identification, target shielding and the like under complex road conditions; a traditional single-mode three-dimensional target detection algorithm cannot effectively fuse the advantages of multi-sensor data and is difficult to accurately identify objects in a distance or under a complex road condition.

The invention discloses a multi-source information fusion query method based on RAG Fusion, which comprises the following steps: receiving original query data input by a user, and carrying out preprocessing and expansion processing on the original query data to obtain an expanded query set; retrieving each extended query in the extended query set in a preset vector library by using a plurality of retrieval strategies to obtain a document set corresponding to each retrieval strategy, and reordering all documents in the document set to obtain a document list; dynamic context construction is carried out on documents in the document list, and a context document set with weights is generated; performing sentence-level segmentation on each document in the context document set, extracting fragments related to the original query data from segmented sentences by utilizing a pre-trained question and answer model, and performing duplicate removal and combination on the extracted fragments to generate an information unit set; and inputting the original query data, the context document set and the information unit set into a pre-trained large-scale language model to generate a target answer.

The invention relates to an enterprise park meteorological fire risk assessment method based on an adaptive random forest model, and the method comprises the steps: building a risk assessment model and a visual early warning platform, enabling the risk assessment model to collect alarm data, meteorological data, enterprise data and park division data, and forming a risk assessment data set, and integrating internal characteristics and external data of the park into a data set, inputting new data into the self-adaptive random forest model by means of a real-time data flow technology, updating and evaluating a fire risk index value of the park in real time, and optimizing a model decision tree to automatically adjust a risk evaluation model and adapt to changes of environment and operation. Accurate early warning information is provided for a visual early warning platform, the risk level of the park is displayed in real time, and a risk alarm is given out in time.

The invention discloses an intelligent electric vehicle driving state safety estimation method under network attacks, and relates to automobile intelligent safety and automatic driving. Firstly, a three-degree-of-freedom dynamic model of the intelligent electric vehicle is established, the model comprehensively considers transverse, longitudinal and yaw dynamic characteristics of the vehicle, and accurate description is provided for subsequent state estimation. And designing an attack signal and state security estimation method based on an adaptive tolerance extended Kalman filtering algorithm. According to the method, the real-time and accurate estimation of the vehicle state is realized by fusing the information of various sensors such as GPS, LiDAR, IMU and the like, and malicious data introduced due to network attack can be effectively eliminated. Meanwhile, the method also has the capability of tolerating non-ideal network environments such as sensor data packet loss and the like, and the safety and reliability of the self-driving automobile are further enhanced. The intelligent electric vehicle can still maintain accurate driving state estimation when subjected to network attack, and safe driving of the vehicle is ensured.

The invention belongs to the technical field of electronic circuits, and provides a low-dropout linear voltage stabilizing circuit, a low-dropout linear voltage regulator and electronic equipment. The low-dropout linear voltage stabilizing circuit comprises an error amplification module, a buffer module, a power module, a feedback module and a transient enhancement module, the buffer module is connected with the error amplification module, the power module and the transient enhancement module, the power module is connected with the transient enhancement module, the feedback module and a load, and the feedback module is connected with the error amplification module. When the current on the load changes, the transient enhancement module provides the first current for the power module according to the target voltage, so that the power module quickly adjusts the target voltage according to the first current and quickly responds to the change of the load; the feedback module outputs feedback voltage according to the target voltage; the error amplification module outputs an error signal according to the feedback voltage and the reference voltage; the buffer module outputs a driving signal to the power module according to the error signal, so that the power module maintains the target voltage stable according to the driving signal.

The invention discloses a document proofreading processing method based on a large model. The document proofreading processing method comprises the following steps: step 1, preprocessing a document and identifying the document as a text; step 2, segmenting the text into independent text blocks; step 3, performing vectorization processing on the text blocks; step 4, performing retouching on the text block by using the random degree large model to obtain a new text block; step 5, performing vectorization processing on the new text block; 6, comparing the vector similarity of the text block and the new text block, if the similarity is high, entering the next step, and if the similarity is low, repeating the steps 4-6; 7, comparing the text of the text block with the text of the new text block, and marking difference characters in pairs; and 8, proofreading the text blocks and the new text blocks according to the document, and integrating and outputting the text blocks and the new text blocks. According to the method, more accurate and comprehensive document proofreading service can be provided, dynamic change of languages and diversified document requirements can be met, and the specialty and credibility of documents are remarkably improved.

The invention relates to the technical field of data processing, in particular to an intelligent recommendation system fusing big data and AI, and the system comprises a scheme obtaining module which obtains a plurality of target marketing schemes; the scheme analysis module analyzes a plurality of target marketing schemes to determine a plurality of marketing features; the data acquisition module collects a plurality of historical behavior data of a user; the data analysis module analyzes a plurality of historical behavior data to determine user characteristics; the association determination module analyzes association between the user features and the plurality of marketing features based on the initial association AI model, and determines prediction priorities of the plurality of target marketing schemes based on an association analysis result; a recommendation module constructs a user recommendation page based on the predicted priority, and determines target priorities of a plurality of target marketing schemes based on the real-time behavior data of the user; and the feedback module compares a comparison result of the prediction priority and the target priority to adjust the initial association AI model to obtain a target association AI model. According to the invention, the accuracy of personalized marketing scheme recommendation of the user is improved.

The invention belongs to the technical field of natural language processing, and discloses a text data labeling method and system based on weak supervised learning. The method comprises the following steps: performing pseudo label generation on a plurality of historical text data by using a weak supervised learning algorithm to obtain a plurality of source domain data provided with real labels and a plurality of target domain data provided with pseudo labels; constructing a text data labeling model by using a deep learning and adversarial training algorithm, and adjusting the text data labeling model by using a continuous learning algorithm; and labeling the real-time text data by using the adjusted text data labeling model to obtain labeled real-time text data, and updating the adjusted text data labeling model by using a continuous learning algorithm. The problems that in the prior art, manual labeling cost is high, labeling quality is uneven, and dependence on label data is high are solved.

The invention relates to the field of meteorological temperature measurement, in particular to a meteorological temperature measurement method and system based on an optical fiber sensor, and the method comprises the steps: collecting data information in an environment through the optical fiber sensor, carrying out the preprocessing of the obtained different parameter information, and providing a data basis for the meteorological temperature measurement method; establishing a meteorological temperature prediction neural network model according to the preprocessed data information, and predicting the meteorological temperature; the meteorological temperature measurement process is improved according to the requirements of different meteorological environments and the prediction result of the meteorological temperature; and evaluating the improvement result of the specific process of meteorological temperature measurement, establishing a temperature measurement evaluation model, and evaluating the effect of the temperature measurement result in practical application. High-precision temperature measurement is provided, meteorological temperature data can be monitored in real time through manual intervention, the change condition of the meteorological temperature is reflected in time, and the whole measurement process has high reliability and stability.

The invention relates to the field of computer vision and artificial intelligence, in particular to a real-time multi-person motion analysis system based on motion perception and semantic understanding, which comprises a motion perception module, a motion tracking module, a motion semantic understanding module and a motion performance analysis module. The motion area segmentation unit adopts a background modeling or inter-frame difference method to identify a motion area and generate a motion mask; the target detection unit is connected with the central processing unit through a data bus and used for processing video frame data and detecting the position and the bounding box of a moving target. The motion area segmentation unit is connected with the target detection unit and is used for generating a motion mask. By integrating a plurality of modules, the motion data can be comprehensively captured and analyzed, and the accuracy and real-time performance of motion analysis are improved. Through combination of the target detection unit and the motion area segmentation unit, the system can accurately identify a motion target and a motion area, and a solid foundation is laid for subsequent tracking and analysis.

The invention discloses a thermal generator set load prediction system based on machine learning, and the system overcomes the defects of a conventional load prediction method through an end-to-end solution composed of a comprehensive data collection module, a multi-level feature engineering module, a multi-model integration module, a self-adaptive optimization module and the like. The system collects multi-dimensional data from a thermal generator set control system, an environment monitoring system, a fuel detection system and a historical database, wherein the multi-dimensional data comprises load data, environment data, fuel characteristics and equipment operation data. According to the multi-dimensional data integration mode, the model not only depends on single historical load data, but combines multiple influence factors such as environment, fuel and equipment states, the problem that a traditional method depends on a single data source is solved, the load prediction precision is effectively improved, and the load prediction efficiency is improved. And the adaptive capacity of the prediction model to different environment conditions and equipment states is stronger.

The invention relates to the technical field of data processing and prediction, in particular to a large hydrological model construction method integrating multi-source fusion and dynamic optimization. The method comprises the following steps: acquiring hydrological monitoring data; performing data preprocessing on the hydrological monitoring data to obtain standard hydrological monitoring data; performing multi-dimensional feature recognition on the standard hydrological monitoring data to obtain multi-dimensional hydrological feature data; and carrying out runoff feature extraction on the multi-dimensional hydrological feature data to obtain hydrological runoff feature data. According to the invention, through a data processing technology, a mode identification technology, a geographic information system technology and a deep learning technology, seasonal change influence evaluation is carried out on hydrological monitoring data; and the hydrological monitoring model is dynamically optimized according to the hydrological geographic position information and the hydrological change evaluation data, so that the response speed of the hydrological monitoring model is higher, and the precision of a prediction result is higher.

The invention discloses a full-automatic feeding rubbing color fastness device which comprises a rack, a wiping rag feeding mechanism, a wiping rag conveying mechanism and a sample cloth feeding mechanism, the wiping rag feeding mechanism conveys a wiping rag body to a feeding path of the wiping rag conveying mechanism, and the wiping rag conveying mechanism comprises a shifting assembly, a mounting frame, a wiping rag adsorption head and a wiping rag clamping head; the wiping rag mold is provided with a wiping rag containing groove, a groove is formed in the inner wall of the wiping rag containing groove, a clamping ring is movably connected into the groove in a sleeved mode, and the wiping rag clamping head is inserted into the clamping ring so that the wiping rag clamping head can be matched with the clamping ring to clamp the wiping rag body. The wiping rag body is attached to the end face of the wiping rag clamping head so as to conduct a friction test on the sample cloth body on the friction position. According to the full-automatic feeding rubbing color fixing device, the clamping ring can be easily withdrawn from the wiping rag clamping head, so that the wiping rag body can be detached from the clamping head, the wiping rag body can be conveniently replaced, and the test accuracy and the detection efficiency are kept.

The invention relates to the technical field of projectors, in particular to a projector control method and a projector. The method comprises the following steps: performing image projection signal processing on target projection image data to generate initial projection picture data; performing projection point cloud space-time registration on the initial projection picture data by using the dual laser radar assembly to obtain projection scanning point cloud space data; performing shielding area geometric feature analysis on the projection scanning point cloud spatial data, determining an ideal projection area, and generating ideal rectangular projection area data; performing intelligent projection control parameter processing on the ideal rectangular projection area data to obtain intelligent projection control parameters; and performing projection attitude control instruction processing on the intelligent projection control parameters to obtain projection control optimization effect data. According to the invention, the projection control parameters are automatically adjusted through intelligent identification of the projection surface, so that the projector can maintain a stable projection effect in different environments, and the quality of a projection image is remarkably improved.

The invention relates to a fault prediction method and system for an IP network. The method comprises the following steps: collecting network link data; extracting network link characteristics according to the network link data; training a GRU model through the network link features to obtain a prediction model; and inputting real-time network data into the prediction model to obtain a prediction result. According to the invention, the efficiency and accuracy performance of performance monitoring and fault prediction can be improved.

The invention discloses a generalizable deep forgery detection method based on multi-classification guidance, and belongs to the technical field of image processing. According to the technical scheme, data enhancement is carried out on an image, and then the image is input into a binary classification network and a multi-classification network of a double-flow neural network; combining the features of the two branches through a fusion module; the fused features are input into a binary classifier to determine whether the input is authentic or counterfeited. The method has the beneficial effects that the generalizable deep counterfeiting detection method based on multi-classification guidance shows remarkable beneficial effects in the aspects of improving the model generalization ability, enhancing the feature expression ability, optimizing the loss function design, preventing over-fitting, improving the calculation efficiency, enhancing the label consistency and the like; the method is beneficial to suppressing the propagation of false information, maintaining the right to know and security of the public, enhancing the vigilance of people to the authenticity of information, laying a solid foundation for the applications of identity authentication, digital forensics and the like in multiple fields, and making an important contribution to the construction of a safer and more credible information society.

The invention belongs to the technical field of machine vision detection, and discloses a product appearance defect identification method and device, equipment and a medium. The method comprises the following steps: obtaining a to-be-recognized image of a product appearance, and processing the to-be-recognized image into a binary image; carrying out contour extraction in the OpenCV model to obtain a contour boundary of the product defect in the to-be-identified image; using a Douglas-Peucker algorithm to extract contour vertex coordinates from the contour boundary, and calculating the contour area of the product defect based on the contour vertex coordinates; determining a minimum bounding box containing a contour boundary through a contour searching method, extracting vertex coordinates of the bounding box, and calculating the area of the bounding box based on the vertex coordinates of the bounding box; and based on the contour area and the bounding box area, calculating an area error ratio of the contour area to the bounding box area, and based on the area error ratio, determining an identification result including qualified product appearance and unqualified product appearance. The detection efficiency can be effectively improved, and the product omission ratio is reduced.

The invention discloses a fire risk assessment and early warning system, and relates to the technical field of machine learning. The fire risk assessment early warning system comprises the following steps: (1) dividing potential factors of a fire risk into physical factors and environmental factors, and collecting related data; (2) constructing a risk assessment data learning sample; (3) dividing fire risk grades; (4) determining a fire risk calculation and evaluation model; and (5) comparing the risk score obtained by the established model with the divided fire risk level to obtain a building risk level. According to the embodiment of the invention, the method can facilitate the detection and early warning of the fire condition, and improves the safety.

The invention relates to the technical field of network security, and discloses a virus monitoring and early warning method based on network traffic deep analysis, which comprises the following steps: collecting network traffic data in real time through a deep packet inspection technology, extracting traffic features, and constructing a traffic feature sequence based on a time window; establishing an anomaly detection model based on the traffic feature sequence, and identifying an anomaly mode deviating from normal behavior distribution; according to an abnormal behavior detection result, constructing a dynamic model of an equipment infection state, and quantifying an equipment infection risk; and constructing a propagation graph model in combination with an equipment interaction relationship, and simulating a diffusion path and a propagation trend of the virus in the network based on propagation dynamics. According to the method, hidden behaviors of unknown viruses and variant viruses can be accurately monitored, a virus propagation path and an influence range are dynamically predicted, optimal allocation of defense resources is realized in combination with a game theory and linear programming, and the real-time performance of virus monitoring and defense is improved.

The invention discloses a compressor air cylinder base grinding device, and relates to the technical field of air cylinder machining and grinding, the compressor air cylinder base grinding device comprises a machining base and further comprises a clamping device, a machining frame is fixedly installed at the bottom of the machining base, a servo motor is fixedly installed on the front side of the machining frame, and a lead screw is fixedly installed at the output end of the servo motor; a moving block is slidably installed on the inner wall of the machining frame and is in threaded connection with the lead screw, the clamping device comprises an electric push rod, a stabilizing plate, a detection rod, a detection plate, an elastic telescopic block, a shielding plate, a button and a limiting hole, the shielding plate moves to break away from contact with the button and release shielding of the button, the clamping force of the stabilizing plate is controlled through limiting, and the clamping force is adjusted to be larger than the clamping force of the limiting hole. Therefore, clamping pressure is uniform and moderate, deformation of the workpiece can be avoided, button shielding can be relieved only after the workpiece is clamped stably, and accidental starting of grinding equipment caused by manual mistaken touch can be reduced.

The food safety detection equipment disclosed by the invention comprises a sampling mechanism, a multi-station turntable device, a centrifugal device, a conveying mechanism, a detection device, a reagent adding device and an extraction mechanism. The sampling mechanism is provided with a sampling table; the multi-station turntable device is provided with a first containing part; the centrifugal device comprises a rotating frame and a first driving mechanism, and the rotating frame is provided with a second containing part; the first conveying mechanism is used for transferring the sample containers among the sampling mechanism, the multi-station turntable device and the centrifugal device; the detection device is provided with a third containing part, a fourth containing part, a first image acquisition device and a second image acquisition device; the reagent adding device is used for adding reagents to form a first sample and a second sample; the extraction mechanism comprises a manipulator and a pipette and is used for extracting the first sample and the second sample. The device has the advantages of high automation degree, good soaking and purifying effects, high detection precision and integration of multiple detection means.

The invention discloses a construction method based on a typical engineering project investment prediction model of a transformer substation. The construction method comprises the following specific steps: S1, collecting a representative typical power transformation engineering scheme and historical price data thereof; s2, preprocessing the collected data, including noise reduction, normalization processing and invalid data removal; s3, substituting the processed historical price data one by one by adopting a control variable method, performing mathematical modeling by utilizing a linear regression model, and quantifying the influence of each variable on the total cost; s4, monitoring independent variable information flow and dynamically adjusting the weight based on a dynamic asymmetric information regulation and control and self-adaptive offset mechanism; s5, verifying and correcting the linear regression model, and optimizing a prediction formula in combination with regional characteristics; and S6, according to dynamic changes of market and regional economy, adjusting the model in real time, and generating final cost prediction data. The problems that the regional adaptability is poor, and the cost scheme is not updated in time are effectively solved.

The invention discloses an agricultural voice question answering system based on large model reasoning and intelligent auxiliary error correction, which relates to the technical field of agricultural knowledge intelligent question answering and comprises a robot terminal interaction module, a user interaction layer, a question answering processing layer, a performance evaluation optimization module and a data layer. According to the method, the GPT large model based on the Transform architecture is utilized, agricultural field adaptation is carried out, the powerful language understanding and generating capacity of the large model can be fully utilized, agricultural related problems can be better processed after fine tuning, initial answers with certain accuracy are generated, a powerful knowledge source and an inference engine are provided for the system, and the system has good application prospects. According to the method, the initial answer generated by the large model can be effectively evaluated and corrected by learning the characteristic difference of a large number of agricultural question and answer data pairs in cooperation with the established BERT error correction model, so that the accuracy and reliability of system answers are improved, and wrong answers possibly occurring in the large model are reduced.

The invention relates to the technical field of OLED micro-displays, and discloses a temperature sensor calibration and external temperature compensation method for an OLED micro-display, and the method comprises a temperature compensation parameter storage module which is used for storing the temperature calibration data of the OLED micro-display through a one-time programmable memory OTP. According to the temperature sensor calibration and external temperature compensation method for the OLED micro-display, the actual working temperature Tactual is calculated by using the current temperature register value REG1 measured by the temperature sensor and factory calibration data T0 and REG0, the brightness change is compensated by adjusting the common voltage Vcom in real time, and the temperature of the OLED micro-display is compensated by combining a temperature compensation formula. The brightness of the display is accurately controlled by adjusting the electric field intensity, the Vcom is improved in a low-temperature environment to compensate brightness reduction, and the Vcom is reduced in a high-temperature environment to prevent excessive light emission or thermal damage, so that the problem of inconsistent display effects in long-term use is solved, and the user experience is improved.

The invention belongs to the technical field of computers, particularly relates to a heart sound signal denoising method based on an improved wavelet threshold value, and aims to solve the problem that a heart noise signal component and a heart sound signal component cannot be accurately determined by an IMF component obtained after a heart sound signal is decomposed through CEEMDAN. In the prior art, the selection of a threshold value and a threshold value function in a wavelet threshold value denoising process affects the denoising effect of a signal, when the threshold value is selected too large, some effective components in the heart sound signal are lost, and when the threshold value is selected too small, part of noise in the heart sound signal is not completely eliminated. The method comprises the following steps: decomposing heart sound signals by using CEEMDAN, defining noisy heart sound IMF components and heart sound signal IMF components by using de-trending fluctuation analysis, and de-noising the heart noise IMF components defined by a de-trending fluctuation analysis algorithm by using a wavelet threshold de-noising method of an improved wavelet threshold selection mode. And reconstructing to obtain a de-noised heart sound signal.

The invention relates to the field of house building, and discloses a detachable formwork for construction of an elevator shaft and a water-collecting well and an anti-floating reinforcing method.The detachable formwork comprises the following assemblies: a hollow box-shaped formwork composed of a side formwork assembly and a bottom formwork assembly which are connected through concave and convex points; the opposite-pull water stop screw comprises a screw body and a water stop piece, and the screw body penetrates through the side mold assembly and the bottom mold assembly of the hollow box-shaped mold plate and is used for connecting the side mold assembly and the bottom mold assembly in a fastening mode; the adjustable jacking comprises a lead screw and a supporting plate, the lower end of the lead screw is connected with the top of the hollow box-shaped formwork, and the supporting plate is fixed to the lead screw through threads and used for adjusting the supporting height of the top of the formwork. By adopting the light and high-strength box-shaped formwork and a detachable connecting mode, the formwork is more convenient and faster to build and disassemble, the construction efficiency is improved, meanwhile, the stability and accuracy of the formwork can be kept through the fixing effect of the pull water stop screw and the adjustable jacking, and therefore the construction quality is improved.