27results about How to "Improve mastery" patented technology

The invention discloses a knowledge tracking cold start optimization method and system based on a large language model, and belongs to the field of knowledge tracking. The method comprises the following steps: firstly, acquiring student answering records of an online learning platform, constructing a prompt template to enable a large language model to output question solving steps, labeling knowledge concepts, and generating a binary pairing relationship between the question solving steps and the knowledge concepts; secondly, taking the problem and the knowledge concept as heterogeneous graph nodes, establishing edge connection and generating a semantic embedding vector; then, a heterogeneous graph encoder is used in combination with a graph attention network and a jump knowledge mechanism, and final node embedding of knowledge concepts and problems is output. And finally, predicting the answer of the student to the target question by comparing, learning and training the answer prediction model. According to the method, the performance and accuracy of the knowledge tracking cold start stage on the online learning platform are effectively improved.

The invention discloses a fine-grained knowledge point mastering analysis method based on local graph convolution, and the method comprises the steps: S1, collecting the learning behavior data of students, and constructing an initial knowledge graph based on the learning behavior data; s2, establishing a local graph convolution model based on the initial knowledge graph; s3, updating and generating a knowledge graph with personalized features; s4, based on the personalized knowledge graph, generating a knowledge point mastering condition table by calculating a mastering degree index of each knowledge point; s5, performing clustering analysis on the knowledge points of which the mastering degree indexes are lower than a preset threshold value in the knowledge point mastering condition table; s6, generating knowledge blind spot group feature vectors; and S7, generating a learning strategy for each knowledge blind spot group according to the feature vectors of the knowledge blind spot groups. According to the invention, refined evaluation and real-time dynamic tracking of knowledge point mastering are realized.

The invention discloses a medical science popularization question and answer model fine tuning method, device, equipment, medium and product, and relates to the field of large language model fine tuning, the method comprises the following steps: cleaning data in a medical text question and answer data set, determining a cleaned instruction-input-output normal form data set, and using the data set as a fine tuning training data set; in a pre-trained large language model, positioning a QKV projection matrix in a multi-head self-attention mechanism and a weight matrix in a feedforward neural network; two low-rank matrixes are introduced into each matrix, the pre-trained large language model is finely tuned, and the finely tuned large language model is determined; the fine-tuned large language model is used for generating an answer according to question data in the process of using interaction between the user and the fine-tuned large language model, the large language model is finely tuned, the performance of processing a specific task is improved, the calculation cost is reduced, the grasp ability of the model for specific details is enhanced, and the user experience is improved. And the requirements of specific user groups or application scenes are met.

The invention discloses a big dipper satellite-based dam high-precision deformation monitoring system and method, and belongs to the field of dam deformation monitoring. Based on a big dipper satellite-based high-precision positioning technology, deformation abnormity and deformation uniformity abnormity analysis is carried out on a dam in a water flow impact process, and high-precision accurate monitoring is carried out on dam deformation; according to the method, the dam is monitored, the risk of the dam is comprehensively evaluated, the mastering ability of managers on the dam is improved, the safety of the dam is further improved, and the problem that in the prior art, when dam deformation monitoring is carried out, deformation abnormity and deformation uniformity abnormity analysis cannot be carried out on the dam in the water flow impact process based on the high-precision positioning technology of Beidou satellites is solved. And high-precision accurate monitoring cannot be carried out on dam deformation.

The invention relates to the technical field of artificial intelligence, and discloses a blood glucose management system based on artificial intelligence. According to the system, blood glucose monitoring is reminded through the blood glucose monitoring terminal, a blood glucose curve is drawn every week, abnormal blood glucose conditions are judged for reminding and alarming, the diet terminal is used for guiding low-sugar diet recommendation, high-sugar food intake is avoided, the mastering degree of a patient and family members for food classification is improved, and the exercise terminal determines the exercise time and exercise amount of the patient according to the blood glucose value. Parameters are input into an intelligent system, the system converts and selects a motion mode, time and intensity, a drug treatment terminal calculates the insulin injection amount of a patient every day and every meal according to the blood glucose level and food intake, the insulin dosage is adjusted according to the calculated numerical values, blood glucose fluctuation is reduced, and the treatment effect is improved. The education and self-management terminal guides a patient to improve the cognitive degree of diabetes and provides personalized propaganda and education guidance for the patient, and the complication reminding terminal controls the blood sugar of the patient within a normal range, guides a good lifestyle of the patient and guides behavior specifications.

The present invention relates to the field of artificial intelligence technology, and discloses a self-review knowledge distillation method, system and terminal based on multi-level attention. The method uses a student model and a converged teacher model to perform prediction processing on sample images in a sample set respectively; obtains feature maps output by the teacher model and the student model at each processing stage, and calculates the attention map of the teacher model and the student model at each processing stage. The attention map of the student model at each processing stage is feature-fused with the attention map of the previous processing stage, and the multi-level attention distillation loss is calculated. The standard cross entropy loss and the basic distillation loss are calculated; the three losses are weighted to obtain the total training loss, and gradient descent and back propagation are performed on the basis of the total training loss, so as to optimize the student model parameters, and then the learning model is continuously optimized using the sample images of the next batch of the sample set until the student model converges. The present invention can effectively improve the accuracy of the student model.

The invention discloses a GIS equipment electromechanical state and typical defect simulation test platform and a GIS equipment electromechanical state and typical defect simulation test method, relates to the field of electrical equipment electromechanical state and typical defect simulation tests, and solves the problem of simulation and research of electromechanical states and defect characteristics of gas insulated switchgear (GIS) under different operation conditions. Wherein the test platform comprises an AC voltage source, a GIS scaling model, a vibration table and an excitation power amplifier, defects are preset in the GIS scaling model, the GIS scaling model is provided with a transparent observation window used for observing the internal condition of the model, and the GIS scaling model is fixedly installed on the vibration table; an alternating-current voltage source is loaded on a GIS scaling model bus to simulate distribution of an electric field in a GIS, and vibration is applied to the GIS scaling model through a vibration table and an excitation power amplifier to simulate GIS vibration caused by current.

A transparent visual power grid operation and maintenance system comprises a data acquisition and preprocessing module which is composed of a sensor network and SCADA system docking unit and a data preprocessing unit; the data processing and analysis module is composed of a fault prediction and diagnosis unit, an operation and maintenance strategy optimization unit and a three-dimensional modeling and dynamic simulation unit; the visual display and user interaction module is composed of an AR / VR interface development unit and a user-defined interface and report generation unit; and the system integration and communication module is composed of a standardized API interface and cloud platform unit and a remote access and cooperation unit, and the scheme of the invention solves the problem that the complexity of power grid operation and maintenance system data causes that detection information is difficult to provide visually, thereby affecting the operation and maintenance work quality and efficiency.

The invention provides an efficient underground accurate delineation method for a steep dip thin and fine quartz vein type ore body, and belongs to the technical field of mine underground prospecting, and the scheme comprises the following steps: firstly, utilizing capital construction engineering and development engineering, preliminarily exposing the ore body, constructing a database in three-dimensional mining software, delineating the ore body at a time, and establishing an ore body three-dimensional model; different laser reflection spectral characteristics of targets with different colors are utilized, three-dimensional laser is used for scanning an ore body form and a roadway form, and an ore body model is updated by combining digital geological recording, mechanical sampling and secondary ore body delineating. Three-dimensional laser scanning, digital geological logging and mechanical sampling are fused, on the premise that expenditure and time are limited, the safety of underground operation and digitization of prospecting engineering achievements are improved, the ore body form and the spatial position are mastered more accurately, mining design is guided and optimized, precise implementation of mining is guaranteed, the mining engineering amount is reduced, digital energy saving is achieved, and the working efficiency is improved. The method belongs to a key technical new method for green low-carbon transformation industry development.

The present invention discloses a road condition recognition and reminder system based on an intelligent automobile wire-controlled steering system, comprising a road condition information acquisition module, a road condition recognition module, a driver reminder control module, and a driver reminder execution module. The system recognizes road conditions respectively under different light intensities and working conditions through a combination of a binocular camera, a laser radar, and a series of vehicle sensors, and fuses the recognition results to obtain the road condition information ahead of the vehicle; and provides the driver with the road condition information ahead in advance through multiple methods such as animation display, steering wheel vibration, voice reminder, and signal light flashing, so that the driver can make correct driving decisions, avoid traffic accidents, and protect the driver's life and property.

The invention discloses simulation equipment for feeder automation training and a method thereof, the simulation equipment comprises a training server, a switch, a serial port server, a training client and an operation table, the training server is used for generating a state change strategy in a fault processing process according to a feeder network topology and a feeder automation mode; the switch is used for connecting the training server, the training client and the serial port server; the serial port server is used for converting an RS485 protocol and an Ethernet protocol; the training client is used for accessing the web service of the training server through a browser; the operation table is provided with an indicating lamp and an operation button.

The present invention relates to the field of equipment monitoring, and discloses an equipment monitoring method, equipment and storage medium based on vector coordinates. The method comprises: obtaining N coordinate data of the monitored equipment, wherein N is a positive integer greater than 2; based on the N coordinate data, generating N‑1 vector coordinates and the motion speed corresponding to the N‑1 vector coordinates; judging whether the N‑1 vector coordinates deviate from the preset route range; when they do not deviate from the preset route range, judging whether the motion speed corresponding to the N‑1 vector coordinates has data exceeding the preset speed threshold; when it does not exist, generating visualization data, and sending the visualization data to the preset management port; when it exists, sending speed warning information to the preset management port. In the embodiment of the present invention, the monitoring accuracy of the motion state of the equipment is improved, and the user's grasp of the motion state of the equipment is enhanced by using the visualization data of the vector motion.

The invention discloses a simulation analysis method and system for a coated sand heating and curing process, which belongs to the field of electronic data processing. The invention constructs an image analysis model, imports image data in the coated sand curing process into the image analysis model to perform image anomaly analysis, constructs an environmental analysis model, imports environmental data in the coated sand curing process into the environmental analysis model to perform environmental anomaly analysis, imports acquired image anomaly analysis results and environmental anomaly analysis results into a curing process anomaly analysis strategy to perform curing process anomaly analysis, performs curing process anomaly warning according to the curing process anomaly analysis results, and accurately estimates the curing results of the coated sand in real time by performing real-time comprehensive analysis of the environment and images in the coated sand curing process, so as to provide real-time and accurate warnings of abnormal situations in the coated sand curing process, thereby improving the ability to grasp the coated sand curing process and the pass rate of the coated sand curing process.

The invention provides a local auditing protection method and system for Azure cloud hybrid deployment. The method comprises a hybrid deployment state exploration process, a key data auditing process and a threat protection process. After entering a mixed deployment state exploration process, entering a key data auditing process; after entering a key data auditing process, entering a threat protection process, judging whether rule verification is passed or not, and if a result is yes, ending; and if not, executing the security policy, and ending. According to the invention, through the preposed hybrid deployment state probing subsystem, integration and analysis of a local region hybrid deployment environment are realized, the mastering condition of a user on current domain hybrid deployment configuration is improved, and information support is provided for subsequent auditing and protection.

The present invention provides a counterfactual explanation generation method and system based on a multi-objective genetic algorithm, which obtains the original instance of the learner from the knowledge tracking data set; obtains the number and length of the questions to be added and randomly generates multiple length codes; obtains the questions to be added from the knowledge tracking data set and obtains the corresponding answer values ​​through the deep knowledge tracking model to be explained, forming the interaction pairs of the questions to be added; obtains the counterfactual instances based on the length codes, the interaction pairs of the questions to be added and the original instances; repeatedly obtains different interaction pairs of the questions to be added to obtain different counterfactual instances, generates a population of counterfactual instances and screens them to obtain the optimal counterfactual instance; and implements the interpretation of the deep knowledge tracking model based on the optimal counterfactual instance. The present invention provides a post-hoc interpretation strategy based on the causal inference theory for the deep knowledge tracking model, and proposes a counseling counterfactual interpretation idea and corresponding method suitable for the knowledge tracking problem context.

The utility model discloses a polytetrafluoroethylene sleeve of a polycrystalline silicon reduction furnace, which comprises a polytetrafluoroethylene sleeve body, the polytetrafluoroethylene sleeve body comprises a top part and a rod part, the top part is arranged on the rod part, and the cross section area of the top part is larger than that of the rod part; and the height of the top is smaller than that of the top of the teflon sleeve in the prior art. According to the technical scheme, the structural size of the polytetrafluoroethylene sleeve is improved, and through the improved polytetrafluoroethylene sleeve, the problems of aging, insulation reduction and even pollution of polycrystalline silicon caused by discharge during use of the polytetrafluoroethylene sleeve are solved.

The invention relates to the technical field of education, in particular to an intelligent adjustment method, system and device for student learning and a storage medium. In order to solve the problems that in the prior art, electronic education equipment is difficult to meet individual requirements of students and is difficult to improve learning quality, the method comprises the following steps of: selecting whether to execute repeated learning operation or not according to stage attention in real time in a learning stage; in the test stage, corresponding consolidation learning operation is selected to be executed based on the answering time and accuracy of each test question; besides, after all the questions are answered, chapter and section absorption degrees are obtained, a graph structure is constructed with chapter and section contents and student personal information, student learning preference information is extracted, and learning comprehensive features are obtained; based on the learning comprehensive features, the student personal features and the to-be-recommended learning method features, obtaining a recommendation value; and finally, updating the learning content form of the next chapter based on the to-be-recommended learning method of which the recommendation value is greater than the recommendation threshold so as to meet the individual learning requirements of the students and improve the learning quality of the students.

The invention discloses a pipeline gas sampling and calibration system which comprises a sampling device, the other end, away from a gas pipeline, of the sampling device is detachably connected with a gas detection sensor, the gas detection sensor is provided with a gas sample connector, the sampling device comprises an electric control piston valve assembly, and a gas sample channel is correspondingly arranged on the electric control piston valve assembly. The gas sample passage is in butt joint with the gas sample interface; the electronic control piston valve assembly is further provided with a calibration gas circuit, the calibration gas circuit is communicated with the sliding way, a calibration exhaust port is formed in the sliding way, the electronic control piston valve assembly further comprises a control system, and the control system comprises a control module, an operation module and a data acquisition module. When gas sampling or calibration is carried out, a control instruction is sent to the control module through the operation module so as to control the gas detection sensor, the electric control piston valve assembly and the first gas valve to carry out gas sampling or calibration, and the data acquisition module can acquire detection data transmitted by the gas detection sensor in real time in the gas detection process.

The application discloses an emergency scene-oriented vertical field large model training and retrieval enhancement method, system, device and storage medium, relates to the technical field of intelligent decision support for emergency scenes, and comprises the following steps: collecting emergency field data and uniformly converting the text, completing segmentation and block division according to semantic breakpoints, and removing duplicates to form a pre-training data set, and constructing a supervised fine-tuning data set on the pre-training data set; after expert sorting and screening according to rules, dividing into reward and punishment models, DPO and PPO training subsets, and selecting a basic model to sequentially perform LoRA continuous pre-training, LoRA supervised fine-tuning and DPO / PPO preference alignment training, and outputting an emergency vertical field large model; constructing a multi-dimensional emergency field benchmark test system through the emergency vertical field large model, and customizing and adapting the emergency vertical field large model according to the functional requirements of different use scenarios. The method disclosed by the application fills the blank of intelligent application in the current emergency field.

The invention provides a method, a system, equipment and a medium for evaluating real-time operation conditions of a new energy power grid, and the method comprises the steps: building a new energy power grid evaluation index system composed of multiple stages of evaluation indexes according to the operation conditions, including the static safety level, the balance adjustment capability, the new energy consumption level and the carbon emission level, of the new energy power grid; and calculation and evaluation result division are carried out on each evaluation index, and finally, a hierarchical progressive method is adopted to carry out hierarchical comprehensive evaluation on the real-time operation condition of the new energy power grid. According to the method, the real-time operation level of the power grid can be visually, rapidly, accurately and comprehensively evaluated, the perception capability of a dispatcher on power grid abnormity is improved, the comprehensive mastering capability on the overall operation situation of the large power grid is improved, and the reliability and stability of power grid operation are further improved.

The utility model discloses a simulation catheter teaching simulation appliance, which comprises a rubber arm, a rubber hand and a fixing ring, cavities are arranged in the rubber arm and the rubber hand, the fixing ring is fixedly connected onto the rubber arm for fixing and assisting, a circulation mechanism is arranged on the fixing ring, and the circulation mechanism is connected with the rubber arm. The circulating mechanism comprises a water injection bottle, a double-end catheter, a drainage tube and an adjusting knob; the water injection bottle is communicated with the rubber arm and the rubber hand through the double-end catheter; by means of the mechanism, a user can adjust the flow according to teaching requirements and simulate the blood flow speed in different physiological states, students can visually feel the impact force of the blood flow on the catheter and the stability of the catheter in the blood flow, more real and dynamic operation experience is provided for the students, the mastering degree of the students on catheter operation skills is effectively improved, and the teaching efficiency is improved. Understanding of the vascular structure and catheter operation is enhanced, and the ability to deal with real operation scenes is improved.

The invention provides a training examination system and method and electronic equipment, and relates to the technical field of computers. The training examination system comprises a generation management module used for obtaining training information of a target post; generating a training task and an assessment task corresponding to the target post based on the training information; the training management module is used for sending the training task or the assessment task to the trainee; based on the interactive operation of the trainee, executing the training task or the assessment task; generating a training record corresponding to the training task and an assessment result of the trainee in the assessment task; and the information management module is used for managing system data in the training examination system, and the system data comprises the training information, the training record and the examination result. According to the invention, unified management of the training examination process is realized, and the management efficiency is improved.

The invention belongs to the technical field of teaching experiment devices, and particularly relates to an advanced control algorithm experiment device for a variable pitch system of a wind generating set, which comprises a 2-meter-high wind generating set and a variable pitch system control algorithm device, and is characterized in that the 2-meter-high wind generating set consists of a tower, a wind indicator, an anemometer, a cabin, a rotor blade and a hub; a brake mechanism, a low-speed shaft, a gearbox, a variable-pitch system, a high-speed shaft and a generator are arranged in the cabin. The problems that an existing wind turbine generator master control and variable pitch controller is high in cost, and an advanced control algorithm is difficult to achieve are solved. Data acquisition and algorithm writing are realized through the sensor board, the industrial control all-in-one machine and other parts, variable pitch control, program correction and automatic scoring can be demonstrated, students can master the operation process and advanced control algorithm of the wind turbine generator, the teaching cost is reduced, and the teaching effect and the innovation ability of the students are improved.

A real-time detection method, system, storage medium and program product for an automobile emergency power supply, in which the resistance value of the emergency power supply detected by a multimeter is obtained; it is determined whether the resistance value is equal to a preset safety value; if the resistance value is equal to the preset safety value, it is determined whether the remaining power is greater than a preset minimum power threshold; if the remaining power is greater than the preset minimum power threshold, the historical use device is determined; the historical average power consumption is determined according to the total power and the historical average use time; it is determined whether the historical average power consumption is greater than the remaining power; if the historical average power consumption is greater than the remaining power, a first message is generated and sent to a mobile terminal bound to the emergency power supply; if the historical average power consumption is not greater than the remaining power, an early warning message is generated and sent to the mobile terminal. When it does not meet the use requirements, an early warning is given, so that the owner's grasp of the real status of the emergency power supply is improved, and the impact caused by the inaccurate status of the emergency power supply is reduced.

The invention discloses an interactive teaching method and system based on a teaching video, and relates to the technical field of information, and the method comprises the following steps: 1, collecting the surface behavior data of a student watching the teaching video for at least three times recently, and the corresponding knowledge evaluation data; 2, analyzing the collected behavior data and knowledge evaluation data through a machine learning algorithm to obtain a concentration state distribution diagram of students in each section of video class, and judging the knowledge absorption ability of the students in combination with the corresponding knowledge evaluation data; step 3, performing targeted optimization processing on teaching video contents in the video library according to the knowledge absorption capability of the students, and inserting an interaction task set adaptive to the students into corresponding time nodes in the teaching videos; and step 4, pushing the optimized teaching video to a student terminal, receiving task completion data of the student in real time, and continuing to optimize the interaction task according to the task completion data to complete video teaching.

The utility model discloses an intrahepatic bile duct puncture practice model which comprises a base body used for simulating a liver structure, a plurality of network pipelines used for simulating a bile duct network structure are arranged in the base body, a plurality of expansion balloons located at different positions are arranged in the network pipelines, and the expansion balloons are used for simulating biliary obstruction. Each dilation balloon is further connected with a first dilation catheter, the other end of each first dilation catheter penetrates out of the base body, and the first dilation catheters are used for providing dilation media for the dilation balloons. According to the intrahepatic bile duct puncture practice model provided by the technical scheme of the utility model, biliary obstruction at different degrees and positions can be simulated according to training requirements, so that the applicability and the flexibility of the model are greatly improved. By means of the design, medical practicers and doctors can operate and practice according to more diversified actual conditions, so that the ability of adapting to different complex conditions is improved, and the surgical skills and operation precision of the medical practicers and the doctors can be enhanced.