26results about How to "Accurate feedback" patented technology

The invention discloses a judgment document element extraction method based on a large model, a storage medium and equipment, and the method comprises the steps: collecting a large number of judgment documents, carrying out the cleaning and word segmentation of each judgment document, and marking key element information; taking the judgment document as the input of the large language model, outputting a key element extraction result of the judgment document, setting a reward function based on the marked key element information, and performing feedback optimization on the large language model in combination with a dominant actor-commentator algorithm; and inputting a newly obtained judgment document into the optimized large language model, and extracting key element information. The judgment document element information can be accurately extracted, the analysis efficiency and accuracy of the judgment document are effectively improved, and manual intervention is reduced.

This invention discloses a method and system for diagnosing and assessing teacher competence based on multimodal data fusion, applied in the field of data analysis. The method includes acquiring teaching video data of the target teacher and its corresponding lesson plan text data; then adaptively segmenting the lesson plan text into multiple text blocks, with the segmentation boundaries determined based on semantic density distribution; fusing corresponding sentiment enhancement vectors into each text block and extracting audio track prosodic features from the teaching video; establishing a precise correspondence between each text block and audio track segment using spatiotemporal alignment technology, and then calculating the first and second matching degrees; adjusting the contribution weight of semantic density to each matching degree to generate the final diagnostic assessment result of teacher teaching competence. The teacher competence diagnostic assessment method based on multimodal data fusion provided by this invention effectively improves the objectivity, granularity, and educational guidance value of teacher teaching competence assessment.

The intelligent twist lock for the container ship comprises a lock body master control box, an upper lock head driving box and a lower lock head driving box are arranged on the upper portion and the lower portion of the lock body master control box respectively, an upper lock head is arranged on the top of the upper lock head driving box in a sliding mode, and a lower lock head is arranged at the bottom of the lower lock head driving box in a sliding mode; an upper lock head driving assembly is arranged in the upper lock head driving box, and a lower lock head driving assembly is arranged in the lower lock head driving box. A lock body information chip convenient for a ship remote control platform to remotely operate the twist lock is arranged on the front side of the upper lock head driving box; the lock body information chip and the control module are additionally arranged on the twist lock, and the twist lock is in communication connection with the ship remote control platform, so that key data basis and equipment support are provided for intelligent stowage and yard automatic management of the container ship; the driving motor is matched with the precise spiral lifting and rotating mechanism, so that the whole process automation of locking and unlocking of the twist lock is realized, the working efficiency is greatly improved, and the labor intensity is reduced.

The present application relates to a kind of cardiopulmonary resuscitation feedback device, equipment and storage medium.It includes: first determination module, for obtaining the electrocardiosignal of organism in preset time, and according to the electrocardiosignal determination the change rule of the electrocardiosignal;Second determination module is used to determine the cardiopulmonary state of the organism according to the change rule of the electrocardiosignal;The cardiopulmonary state includes the first state of the change rule meeting preset requirement and the second state of the change rule not meeting preset requirement;Feedback module is used to feedback the suggestion information of the chest compression to the organism in the cardiopulmonary state of the organism is second state.The cardiopulmonary resuscitation feedback device, equipment and storage medium provided by the present application enable ordinary user to obtain accurate, comprehensive guidance and feedback when carrying out cardiopulmonary resuscitation.

The present application relates to a kind of for anastomat pressure resistance performance test instrument, cantilever frame is fixed on host computer, first height adjusting component is installed on host computer, water tank and first height adjusting component slip connection, horizontal adjusting component and cantilever frame are connected, second height adjusting component and horizontal adjusting component are connected, sample connector and second height adjusting component are connected;Cantilever frame slip connection has pressure sensor;Water tank and sample connector are connected by circulation pipe, water tank is connected by backwater pipe and pressure sensor, water tank is connected by test water pipe and pressure sensor;The highest water level line of water tank is connected with overflow pipe, pressure sensor and sample connector are connected by connecting pipe, connecting pipe is connected with drain pipe, overflow pipe, connecting pipe and drain pipe are all equipped with control valve.The present application also relates to a kind of for anastomat pressure resistance performance test method.The present application can accurately detect anastomat pressure resistance performance in real time, belongs to the technical field of anastomat performance test.

The invention relates to an intelligent extractor hood which comprises a machine shell provided with an oil smoke channel. The interdigital electrode array is arranged in the oil smoke channel, the interdigital electrode array comprises two interdigital electrodes, and the electrodes of the two interdigital electrodes are arranged in a crossed manner; the vortex induction part is arranged in the oil smoke channel and located on the upstream of the interdigital electrode array along the fluid flowing path; the controller is configured to obtain the humidity and the oil smoke concentration in the oil smoke channel according to the electrode capacitance value of the interdigital electrode array; the electrode capacitance fluctuation dominant frequency of the interdigital electrode array is determined according to the vortex fluctuation caused by the vortex induction piece, and the airflow flowing speed in the oil smoke channel is obtained. According to the intelligent range hood, the humidity, the oil smoke concentration and the airflow speed can be measured at the same time through simple and low-cost devices, so that the conditions of water vapor and oil smoke in the oil smoke channel are distinguished. The invention further relates to an intelligent control method of the range hood, automatic control over different cooking scenes is achieved, and the fan gear of the intelligent range hood is dynamically adjusted.

The invention discloses an airflow circulation optimization structure and an energy-saving control method of a large walk-in constant temperature and humidity laboratory, the airflow circulation optimization structure comprises a laboratory box body and an airflow circulation system arranged in the box body, and the airflow circulation system comprises two groups of air supply units and a group of air return units which are symmetrically arranged along the length direction of the box body. Through the combination of the symmetrically arranged air supply units and the rotational flow air supply outlets, a gear linkage adjusting mechanism and a particle swarm algorithm are matched, accurate and synchronous optimization of the angles of the air supply outlets is achieved, rotational flow full coverage is achieved, local dead angles are eliminated, the temperature and humidity uniformity in the box is greatly improved, and the reliability of test data is guaranteed. And a flow equalizing plate and a multi-probe sensor group of the main air return channel ensure the accuracy of monitoring data and provide reliable feedback for control.

The application discloses a waterproof coiled material lap joint waterproof precision detection device and method, and relates to the field of waterproof coiled material detection. The waterproof coiled material lap joint waterproof precision detection device comprises a detection table, a detection cavity is formed in the detection table, the upper end of the detection cavity is of an open structure, a positioning groove is formed in the upper position of the detection table and corresponds to the detection cavity, a leakage pressing disc is arranged in the positioning groove, a pressing plate is detachably connected to the positioning groove, a plurality of groups of detection cavities are equidistantly arranged along the length direction of the detection table, and the positioning groove, the leakage pressing disc and the pressing plate are respectively and correspondingly arranged in cooperation with each group of detection cavities. Through the arrangement of the plurality of groups of detection cavities, the electrical cooperation of the switchable mode boosting unit, the induction unit and the prompt lamp, the synchronous or single independent detection of the plurality of groups of coiled materials is realized, the detection efficiency and precision are improved, and the waterproof performance of the lap joint can be directly determined.

This invention provides a thermochromic tape for packaging. The thermochromic tape is a long strip, comprising a substrate layer, a color-changing adhesive layer, and a release film layer stacked sequentially; the color-changing adhesive layer is a thermochromic adhesive layer. This invention, by mixing water-based adhesive and thermochromic materials to form a color-changing adhesive layer with color-changing function and excellent adhesion, can indicate changes in ambient temperature during the packaging and transportation of the packaged product; simultaneously, the tape's stickiness decreases after color change, making it easy to peel off without detachment.

A modular die head control system belongs to the field of thin film processing technology. It includes a host computer connected to a main controller, which is connected to the control signal input terminal of a switching device. The contacts of the switching device are connected to thermal expansion bolts. The system is characterized by: a main frame (3) being provided, within which at least one set of control modules are arranged by plug-in connection; the main controller and switching devices are located in the control circuit within the control modules; and several bolt control terminals (9) for connecting thermal expansion bolts are provided on the surface of the control modules. In the modular die head control system of this application, the control modules are installed in the main frame by plug-in connection, and bolt control terminals for connecting thermal expansion bolts are provided on the control modules, enabling centralized control of the thermal expansion bolts at the lip, so as to accurately control and provide feedback on the state of the thermal expansion bolts.

The utility model provides a roadbed supporting device capable of adjusting settlement, and relates to the technical field of road engineering. A fixed base is fixedly connected to the bottom of the supporting stand column, sliding sleeves are fixedly connected to the top end face of the outer side of the fixed base in an annular array mode, detection sliding rods are slidably connected into the sliding sleeves, sliding detectors are fixedly installed at the tops of the outer walls of the sides, facing the supporting stand column, of the sliding sleeves, and a main supporting assembly is fixedly installed in the supporting stand column. According to the roadbed supporting device capable of adjusting settlement, the detection sliding rod, the supporting spring and the sliding detector are combined, the settlement condition of a roadbed can be monitored sensitively in real time, settlement information is accurately fed back, the friction force between monitoring components is reduced through the design of the sliding friction balls, the accuracy of monitoring data is further improved, and the safety of the roadbed is improved. The hydraulic cylinder and the supporting seat are rotationally connected through the ball socket and the ball head, the height of the supporting seat can be flexibly adjusted, the settlement difference value is compensated, and the problems that stability is poor and settlement is difficult to monitor and adjust are solved.

The invention provides a method and device for detecting articles left in a vehicle and the vehicle, and relates to the technical field of automotive electronics, and the method comprises the steps: responding to a detection instruction, analyzing the detection instruction, and obtaining a target feature of a target object; in response to the detection instruction, obtaining an in-vehicle image, and performing visual feature extraction on the in-vehicle image to obtain visual features; processing the target feature and the visual feature, and detecting a target object conforming to the target feature from the in-vehicle image through an open vocabulary target detection method; and when the target object is detected, feeding back the specific position of the target object for dynamically identifying the left article according to the open vocabulary.

This invention relates to an automated circuit board scoring method, apparatus, electronic device, and storage medium. The automated circuit board scoring method of this invention includes: extracting regions of interest (ROIs) from images of both sides of the circuit board, and performing circuit line segmentation and component identification on the ROIs of the motherboard on both sides respectively to obtain a circuit segmentation diagram and a component information diagram of the motherboard; sequentially extracting and erasing numbers from the circuit segmentation diagram of the motherboard to obtain the motherboard's numbering information and an optimized circuit segmentation diagram; performing solder joint identification and geometric recognition on the optimized circuit segmentation diagram of the motherboard to obtain solder joint, length, and width information of the motherboard respectively; associating the component information diagram with the solder joint information to generate a component-circuit line topology diagram; and performing a comprehensive scoring calculation on the solder joint, length, width information, and component-circuit line topology diagram to obtain a comprehensive score for the current motherboard. The automated circuit board scoring method of this invention effectively improves the granularity and accuracy of scoring.

This invention discloses a method and system for real-time data tracking of digital steel coils on a cold rolling processing line. The method defines and configures the equipment and locations on the cold rolling processing line, including location names and tracking modes, which are divided into primary tracking mode and secondary tracking mode. It iterates through the primary and secondary tracking message data, matching each tracking location with material information. It records the material information changes at each tracking location in real time, obtaining the arrival and departure times of the material at each tracking location. Simultaneously, based on the real-time production data of the material, combined with the length and time of the steel coil head crossing that location, it calculates production data parameters, including: length calculation, maximum, minimum, and average value calculation, multiple process standard hit rate calculation, and alarm calculation. This invention can accurately reflect the actual production status, reduce the downtime failure rate of the processing line, and significantly improve product quality and production efficiency.

The invention discloses a method, a system and equipment for optimizing a steady-state operation verification point of a low-frequency system, and belongs to the technical field of power transmission. The method comprises the following steps: acquiring a plurality of power points of the low-frequency power transmission system in a normal operation process, and carrying out unified processing on the power points to form a conventional working condition; frequently occurring high-frequency working conditions are provided in the normal operation process of the low-frequency power transmission system, the adjacent working conditions of each high-frequency working condition are calculated, and the high-frequency working conditions and the corresponding adjacent working conditions are obtained; performing combination and deduplication processing on the conventional working condition, the high-frequency working condition and the corresponding adjacent working conditions to obtain a working condition of the steady-state verification point; and dynamically adjusting power points in the working conditions of the steady-state verification points, identifying working condition points causing instability of the low-frequency power transmission system, and carrying out optimization check. According to the method, the optimization problem of the operation verification point in the steady-state power checking of the low-frequency power transmission system is solved, the accuracy of operation verification is improved, and the operation verification can better feed back the operation condition of the actual system.

The invention provides a laser additive manufacturing dynamic monitoring method and system based on multi-sensor cooperative sensing, and the method comprises the steps: S1, constructing a distributed clock synchronization system, and collecting a multi-physical field signal in a laser powder feeding additive manufacturing process; s2, key process parameters in the signals are extracted, time domain alignment is carried out on the key process parameters and pose data of an integrated machine tool motion control system, and time-space associated data are obtained; s3, a space-time association database is established to store space-time association data, and an additive manufacturing three-dimensional model is obtained; s4, the key process parameters are mapped to the additive manufacturing three-dimensional model for visual rendering, and digital twin bodies are generated; and S5, dynamic monitoring and feedback are conducted on the additive manufacturing process based on the digital twin body. Through collaborative innovation of a multi-source heterogeneous data space-time alignment mechanism, a process parameter dynamic compensation algorithm and a physical field visualization engine, precise mapping and real-time twinborn feedback of full-dimensional parameters in the additive manufacturing process are achieved.

The application discloses a fault query method and a query system of a motor control system, and relates to the technical field of motor control. The method comprises the following steps: according to a fault identifier, querying pre-generated associated information to obtain a storage positioning result of a full-cycle frozen frame set corresponding to the fault identifier; wherein the full-cycle frozen frame set comprises a pre-fault data subset, a fault moment data frame and a post-fault data subset arranged in time sequence; the associated information is used to establish and map the fault identifier and the storage position and time sequence of each data frame in the full-cycle frozen frame set; according to the storage positioning result, the full-cycle frozen frame set is read from a storage medium and assembled into a serialized response message in time sequence, and the serialized response message is fed back to the external device. The method overcomes the limitations of the traditional single-frame and unordered data query mode, and realizes efficient and accurate acquisition of a continuous data set arranged in time sequence in the full cycle of a specific fault.

The application discloses a kind of automatic driving data processing method, system and equipment. Among them, the processing method of automatic driving data includes: for each automatic driving related data, the corresponding relationship of automatic driving collection data and event is established;Event is synchronized to work order system, to create the work order of the event by the work order system;The event is analyzed by the work order, to carry out label annotation;When the state change of work order is monitored, the structured association relationship of label and the event is established, and the structured association relationship is written in association table, to complete the label structuring of the event.Using the present application, the data processing efficiency is improved, the repetitive operation of manual participation is reduced, the speed of problem discovery to data available is accelerated, the analysis result can be accurately fed back to data through automatic mapping of data and event, so that label can directly participate in query, statistics, clustering and data mining, support large-scale analysis.

The invention provides an electric power operation teaching method and device based on a large language model, and the method comprises the steps: collecting a first visual angle video of a student through an image collection module, enabling a real-time video stream to change along with the visual angle of the student, being not liable to block, and being capable of capturing operation details; real-time analysis is conducted on the real-time video stream, operation logic and equipment state changes are obtained, relevant knowledge is retrieved based on an RAG knowledge enhancement module to serve as context information to be sent into a large language model together with the equipment state changes for reasoning, and the general reasoning capacity of the large language model is anchored to professional electric power specifications; and finally, professional and accurate real-time guidance and feedback are provided by utilizing the reasoning ability of the large language model.

The utility model relates to a knob valve capable of displaying gears and accurately adjusting water flow and air pressure, and belongs to the technical field of pressure adjusting structures of rotary valves of medical instruments. The utility model discloses a knob valve capable of displaying gears and accurately adjusting water flow and air pressure. The knob valve comprises an adjusting rotating structure, a potentiometer adjusting plate, a pressure adjusting valve fixing plate and a pressure adjusting valve. The potentiometer adjusting plate and the pressure adjusting valve are fixed to the two end faces of the pressure adjusting valve fixing plate respectively, and the adjusting rotating structure penetrates through the potentiometer adjusting plate and the pressure adjusting valve fixing plate to be connected with the pressure adjusting valve. The pressure adjusting cap is rotated to drive the adjusting rod to rotate, the adjusting rotating inlet in the rod body of the adjusting rod is communicated with the adjusting rotating outlet in the front section of the adjusting rod, the adjusting rotating inlet and the adjusting rotating outlet move relative to the pressure adjusting valve inlet and the pressure adjusting valve outlet in the axial direction, and meanwhile the pressure adjusting cap drives the potentiometer contact to rotate to display gears. And accurate adjustment of water pressure and air pressure is realized.

The invention discloses a method, a device and a system for realizing data access control based on a digital contract, relates to the field of data security, solves the problem of insufficient security of existing data access, and comprises the following steps: S1, acquiring an electronic contract, generating a digital contract element list, and checking; s2, generating a contract according to a consistency verification result to obtain a digital contract, and deploying the digital contract in combination with a data flow alliance chain; s3, extracting a data connection node to verify and judge the data request, and controlling the data according to a verification and judgment result; s4, performing authority control on the data user, performing data output on the data user based on the authority control, and collecting and uploading full-flow information of data output to an alliance chain for evidence storage; s5, monitoring the digital contract, and controlling the digital contract in real time; according to the invention, the security and stability of data access can be effectively improved.

This application provides a method and system for voice outbound call tag retrieval based on reinforcement learning. The method includes: identifying business targets requiring initial exploration, which have low frequency of occurrence or have not appeared in the system's historical decision-making experience; allocating a predetermined proportion of specific resources to the business target, these specific resources being unaffected by the system's predicted success rate for the business target; utilizing these specific resources to conduct initial outreach to the target group of the business target and acquiring customer interaction information generated from the initial outreach; performing intensity enhancement processing on the learning signal generated from the customer interaction information; adjusting the recommendation strategy for the business target based on the intensity-enhanced learning signal; evaluating the initial outreach effect of the business target; and adjusting the allocation proportion of the specific resources based on the initial outreach effect. This application improves the accuracy of tag retrieval and the rationality of resource allocation.

The invention discloses a channel information compression feedback method and system based on a large language model. The method comprises the steps that firstly, a channel state information compression feedback task is reconstructed into a mask lexical element prediction task aligned with the function of the large language model; then, an information theory mask selection strategy based on self-information is designed, channel state information elements with the highest self-information are recognized and screened out at a user side to serve as visible lexical elements to be fed back, elements with the low self-information serve as masked lexical elements, and therefore the prediction accuracy of the large language model is maximized; and the base station end reconstructs complete channel state information representation according to feedback visible lexical elements through context reasoning by utilizing the strong modeling capability of a large language model. According to the method, the user side only needs to deploy a lightweight network, the calculation complexity is mainly concentrated on the base station side, the limitation of a traditional small model on the channel state information reconstruction precision is effectively broken through, and the communication rate in a multi-user multi-input multi-output scene is remarkably improved.

This invention discloses a multi-agent dialogue response control method and apparatus. Relating to the field of artificial intelligence, the specific solution is as follows: In response to detecting an event that a new dialogue message has been received at the dialogue interface, the backtracking evaluation attribute of the new dialogue message relative to historical dialogue pairs in the dialogue interface is determined based on the new dialogue message and the interaction pair text. The interaction pair text is obtained by processing historical dialogue pairs in the dialogue interface; the historical dialogue pairs include a first user input message and a first agent response message, and the interaction pair text includes a processed second user input message and a second agent response message. When the backtracking evaluation attribute is greater than a preset attribute threshold, an agent response message for the new dialogue message is generated based on the first target agent to which the target interaction pair text associated with the backtracking evaluation attribute belongs. The agent response message for the new dialogue message is then displayed in the dialogue interface. This invention improves the accuracy and rationality of backtracking judgment.

The utility model relates to a roller shutter door technical field especially based on intelligent roller shutter door of lamp area instruction, include: lintel, door curtain, guide rail subassembly, indicator and along the guide rail inboard guide groove edge and the bottom of lintel surrounds the arrangement of lamp area, cooperate the color controller, data acquisition module and controller, realize the linkage of lamp area color and door curtain state and the biological activity signal before door body front, the controller according to the door curtain closing, opening, running state and the biological activity signal before door body front, control lamp area presents blue, green, yellow, red and different color, clearly feedback roller shutter door working state and surrounding environment, improve the indication intuitiveness and information transmission efficiency. The utility model gives consideration to durability, flexibility and anti-aging performance, ensure door curtain sliding stability, prolong the service life of equipment, improve the safety of logistics and production workshop.