56results about How to "Improve response accuracy" patented technology

The invention relates to the field of robot attitude control, in particular to a robot attitude control method and system based on intelligent recognition. The method comprises the following steps: continuously acquiring a real-time sandbox scene video stream based on a high-definition camera; and carrying out time sequence frame segmentation on the real-time sandbox scene video stream, and carrying out global contrast enhancement processing to obtain a global contrast optimization image sequence. Carrying out dynamic object identification on the global contrast optimization image sequence, and carrying out continuous frame optical flow motion tracking to generate a dynamic object optical flow change track; optimizing the image sequence according to the global contrast to extract scene ground and scene obstacle data; performing dynamic trajectory rendering on the scene ground and scene obstacle data based on the dynamic object optical flow change trajectory so as to construct a dynamic scene twinborn model; robot position posture parameters and user operation instructions are recognized. According to the invention, timely attitude adjustment response and accurate movement position control are realized.

The invention discloses a method for generating question reply through retrieval enhancement based on an iteration mode. The method comprises the following steps: S1, taking user input as a retrieval text for retrieval; s2, summarizing the large model to obtain a primary information summary; s3, sorting the content retrieved and returned in the step S1 into a recall fragment set; s4, summarizing the initial information in the S2 as a retrieval text to perform the retrieval; s5, the big model perfects the previous information summary by using the content returned by the current retrieval to obtain the current information summary; s6, sorting the content returned by the retrieval into a recall fragment set, identifying whether a new fragment exists or not, if not, entering step S8, and otherwise, entering step S7; s7, performing current retrieval by taking the current information summary as a retrieval text, and entering S5; and S8, replying user input according to the information summary. The RAG accuracy is improved based on an iteration mode, more perfect information is obtained through iteration retrieval, and the model reply quality is improved.

The invention belongs to the technical field of oil-gas exploration and development, and relates to a transient electromagnetic wave through casing resistivity measurement method, which comprises the following steps of: establishing a cased well stratum model by considering a borehole, a casing and a stratum and combining transient electromagnetic wave logging characteristics; in the cased well stratum model, resistivity is measured by adopting an apparent resistivity extraction method based on a late signal, or an anomaly-background method based on a middle signal, or a compensation differential coil structure; the compensation differential type coil structure comprises a transmitting coil and at least one compensation coil system, the compensation coil system at least comprises a first receiving coil and a second receiving coil which are sequentially arranged, the transmitting coil and the second receiving coil are wound in the forward direction, and the first receiving coil serves as the compensation coil and is wound in the reverse direction; and the transmitting coil and the compensating coil system are sequentially arranged from front to back. According to the method, the transient electromagnetic wave through casing logging response precision can be improved, and formation resistivity extraction under the influence of a high-conductivity casing is achieved.

The invention discloses an intelligent frequency modulation method and system for a photovoltaic inverter and a storage medium, and relates to the technical field of power systems. The intelligent frequency modulation method for the photovoltaic inverter comprises the following steps: voltage monitoring coincidence evaluation; performing frequency monitoring coincidence evaluation; and evaluating power monitoring coincidence. According to the invention, the voltage monitoring circuit is evaluated to obtain a voltage monitoring coincidence evaluation index and determine whether voltage frequency modulation is carried out, and then the frequency monitoring circuit is evaluated to obtain a frequency monitoring coincidence evaluation index and determine whether frequency modulation is carried out. And finally, evaluating the power detection circuit to obtain a power monitoring coincidence evaluation index, and judging whether to perform power frequency modulation, thereby achieving the effect of improving the response accuracy in the intelligent frequency modulation process of the photovoltaic inverter, and solving the problem of insufficient response in the intelligent frequency modulation process of the photovoltaic inverter in the prior art.

One or more embodiments of the invention provide a training sample generation method and device for a large language model, and the method comprises the steps: obtaining medical data containing disease features and diagnosed diseases, and obtaining various diseases and corresponding disease features; generating a training sample set and a label thereof based on the medical data, various diseases and corresponding disease features; the training sample comprises data used for indicating whether disease features corresponding to a target disease exist or not, and the label of the training sample is used for indicating whether the target disease is diagnosed or not; performing cross validation on the training sample set based on a machine learning model to obtain a prediction result of the machine learning model for the training sample set; based on the label and the prediction result, screening out a high-quality training sample matched with the prediction result and the label from the training sample set; the high-quality training sample is used for performing supervised fine tuning on the large language model.

The invention relates to an AMT (Automated Mechanical Transmission), in particular to an AMT system and a control method thereof, which are used for solving the problems that the existing AMT system has certain defects in the aspects of clutch control, accurate control in a gear shifting process, stability of a self-locking device and comprehensive application of a sensor, so that the gear shifting precision, the response speed and the stability are insufficient. According to the AMT system, the integrated gear selecting and shifting executing mechanism is adopted, and compared with an externally-hung control system, the integrated gear selecting and shifting executing mechanism can improve the gear taking-off and gear shifting response speed of the system, reduce electric control wire harnesses and gas circuit arrangement and greatly reduce the system failure rate.

The invention discloses a knowledge-enhanced question-answering method based on a multi-dimensional mixed retrieval strategy, and relates to the technical field of computer processing, S01, setting a plurality of question-answering scenes based on preset literature materials, and inputting the plurality of question-answering scenes as training data into a BERT model for training to obtain a fine tuning model; s02, segmenting the preset document material through a granularity document block segmentation algorithm to obtain a plurality of paragraphs Sn and a plurality of sentences Cn corresponding to each paragraph Sn; s03, the multiple paragraphs Sn are imported into the BERT model to execute similarity judgment, the multiple similar paragraphs Sn are combined according to a judgment result, and a combined paragraph Sn set is obtained; and S04, key information of each paragraph Sn in the combined paragraph Sn set is extracted. According to the method, a multi-dimensional mixed retrieval strategy is adopted to create a quick lookup model about periodical literatures, so that data response is quicker.

The invention relates to a multi-type distributed energy coordination control method and device and a medium. The multi-type distributed energy coordination control method comprises the steps of collecting real-time operation data of distributed resources; calculating the adjustable range of each device based on the online condition and the real-time operation data of the distributed resource device; decomposing a power grid frequency modulation instruction through a power grid frequency modulation instruction decomposition algorithm, preferentially allocating equipment with high priority, and for equipment with the same priority, performing weighted average allocation based on rated power on a difference value between a power grid dispatching instruction and actual total power to obtain instruction power of each station; and controlling each device in the station according to the instruction power of each station through a station control algorithm. Compared with the prior art, on-line, off-line and operation conditions of each device are considered, power redistribution among resources is carried out under the condition that the aggregator power meets the power grid frequency modulation instruction, it is guaranteed that each decomposition result is reasonable, the adjustment characteristics of various devices can be met, and the response precision is improved.

The invention discloses a precision current transformer measurement and data analysis method, particularly relates to the field of current detection and dynamic signal analysis, and comprises a precision magnetic core design scheme based on a nanocrystalline strip and a dynamic magnetic performance test and frequency response analysis method. A nanocrystalline strip with high magnetic conductivity and low loss characteristics is selected as a magnetic core material, and a high-performance magnetic core suitable for multiple frequency ranges is designed; the magnetic stability and response characteristics of the magnetic core material under different working conditions are evaluated by combining high-frequency current source excitation with dynamic magnetic performance test; a signal acquisition and frequency decoding technology is adopted, an output voltage signal is analyzed in real time, and the linearity and precision of a measurement system are optimized through a mathematical model; according to the method, accurate measurement and data analysis of complex current signals under high-precision and high-frequency conditions are realized, and the applicability and reliability of the current transformer are improved.

This invention discloses a non-destructive testing method for evaluating the fatigue characteristics of long-life pavement materials, belonging to the field of pavement material performance testing technology. The method includes the following steps: mapping the detection data collected by each sensing unit in an intelligent sensing array to the same time reference, constructing a continuous and consistent time-stamped acquisition chain, and generating a time-stamped consensus band at the end of the time-stamped acquisition chain. By constructing the time-stamped acquisition chain and consensus band, this invention identifies the dominance switching moments of multi-source detection data, divides fatigue evolution stages, rearranges the data output order, constructs a relay sequence, and introduces a dynamic control mechanism to release the dominant signal segment by segment, achieving continuous expression and jump suppression of the fatigue state during the dominance switching process.

The present application relates to the technical field of ecological restoration scene modeling, in particular to an ecological restoration scene modeling method and system platform applied to land space planning, comprising the following steps: based on successful ecological restoration technical scheme cases and land units to be restored.The present application realizes modeling in data source selection granularity and expression dimension by normalizing the multi-dimensional attributes between successful ecological restoration technical scheme cases and land units to be restored, constructing source-target land feature fingerprint sets, and combining annual mean temperature, precipitation, soil type, pH value, altitude, native vegetation type and labor cost and other specific attributes.Further, by setting importance coefficients according to the influence degree of each dimension in historical restoration success cases, the present application guides feature difference weight calculation, and obtains technical migration adaptation score by combining normalized difference value calculation, thereby establishing a quantifiable source land adaptation capability evaluation index, and breaking the original limitations of relying on experience judgment.

This invention discloses a method, equipment, and medium for super high-rise construction monitoring based on digital twins, relating to the field of digital twin technology. The method includes: acquiring construction operation data from the super high-rise construction site and constructing a digital twin construction model; updating the boundary conditions of the digital twin construction model through a set of construction disturbance events to generate a disturbance-driven twin model; verifying the consistency between the virtual and real data of the model simulation state records in the construction state set and the disturbance-driven twin model to generate a twin credibility evaluation set; improving the local monitoring resolution of high-risk areas to generate twin resolution configuration results; performing local high-precision monitoring of the credible construction twin based on the twin resolution configuration results; and outputting super high-rise construction monitoring results. This invention, by improving the local monitoring resolution and generating twin resolution configuration results, avoids the waste of computational resources caused by global high-precision monitoring, thereby improving the efficiency of monitoring resource utilization and the response accuracy of high-risk areas.

The application discloses an airplane ground air conditioner energy-saving control method combining environmental parameter prediction, and relates to the technical field of airplane ground support equipment. The method comprises the following steps: collecting airport environmental parameters in real time, training a machine learning prediction model in combination with historical data, and predicting environmental change trends in advance; analyzing the thermal characteristics of an airplane according to the characteristics of the airplane, performing landing time error compensation, and accurately obtaining an environmental parameter sequence of a landing period; calculating an environmental parameter and a cabin thermal characteristic deviation coefficient, and dynamically configuring temperature and energy-saving weights; and iteratively optimizing air conditioner control parameters by using a multi-objective optimization algorithm to obtain optimal air conditioner control parameters. The application realizes the transformation from passive control to active prediction, effectively solves the problems of control lag and high energy consumption in the prior art, significantly reduces energy consumption while ensuring comfort, and improves the intelligent level and operation efficiency of an airplane ground air conditioner system.

The invention discloses an integrated simulation and optimization method for overvoltage protection of a 10kV system of a hydropower station. The method comprises the following steps: acquiring operating parameters of a circuit breaker, a transformer, a cable and a lightning arrester; establishing a circuit breaker arcing time dynamic adjustment module, a transformer iron core nonlinear characteristic dynamic simulation module, a cable frequency response optimization module, a self-adaptive lightning arrester volt-ampere characteristic control module and a resistance-capacitance absorber frequency self-adaptive adjustment module based on the simulation platform; the modules are integrated to a simulation system, and system overvoltage protection measures are optimized through simulation parameters. According to the method, the reliability and the adaptability of overvoltage protection of the 10kV system of the hydropower station are remarkably improved, and a powerful guarantee is provided for the safety and the stability of operation of the hydropower station.

The invention discloses a touch position detection method, a touch display driving device, equipment and a storage medium, and belongs to the technical field of touch. The method comprises the steps that multiple pieces of first touch control data are obtained based on touch control operation in a touch control screen, the multiple pieces of first touch control data correspond to multiple sensing areas, and at least one sensing area in the multiple sensing areas is an irregular area; performing data calibration on the plurality of first touch data to obtain a plurality of second touch data, the plurality of second touch data corresponding to a plurality of calibration areas, the plurality of calibration areas being regular areas, and the plurality of calibration areas being determined based on the plurality of sensing areas; and determining a touch position of the touch operation in the touch screen based on the multiple pieces of second touch data. On the basis, interference of irregular areas can be reduced, the accuracy of the determined touch position is improved, and then the response precision of the touch screen is improved.

The application provides an ablation-based large model example selection method, which comprises the following steps: building a user question representation model, training the user question representation model based on a database, obtaining examples related to a user question through the trained user question representation model, building a first language model and a second language model, and inputting the user question and the related examples into the first language model. The ablation-based large model example selection method provided by the application can obtain suitable examples in an existing database and help a large model generate better replies.

This disclosure provides a neural microelectrode with temperature control and temperature measurement functions, which can be applied in the field of microelectromechanical systems (MEMS) biosensor technology. The neural microelectrode includes: a substrate including a detection region; a microelectrode array disposed on the substrate and overlapping with the detection region, the microelectrode array being used to measure the electrical signal of a analyte; a temperature control element disposed on the substrate and overlapping with the detection region, used to control the temperature of the detection region; a temperature measurement element disposed on the substrate and overlapping with the detection region, used to measure the temperature of the detection region; and an insulating layer disposed on the microelectrode array, the temperature control element, and the temperature measurement element, covering the microelectrode array, the temperature control element, and the temperature measurement element. This disclosure integrates temperature control and temperature measurement functions with the neural microelectrode, realizing temperature control and temperature measurement of neurons, while simultaneously recording the electrical signals of neurons.

The application discloses a parallel hole damping-spiral tube inerter underwater damping device and belongs to the technical field of underwater vibration control. The device comprises an upper shell, a lower shell, a damping spring, an open small hole damping and an open fluid inerter. The upper shell and the lower shell are two parts in relative sliding. The damping spring and the open small hole damping are assembled in parallel to form an integral body, which together constitutes an open spring damping. The open spring damping and the open fluid inerter are connected in parallel, and the upper and lower ends of the two are respectively fixed to the upper shell and the lower shell. The open small hole damping and the open fluid inerter are both installed in the upper shell and the lower shell. The application adopts a parallel topology structure of inerter, damping and spring. The double spiral pipes and the symmetrical piston rods realize the fluid inerter effect in two directions. The small holes on the damping piston realize the speed-dependent throttling damping. The device has the characteristics of wide frequency vibration isolation, strong static load capacity, symmetrical positive and negative characteristics and deep water environment.

This disclosure provides a distance determination method, apparatus, and storage medium. The method includes: determining whether data acquired by an infrared distance sensor of an electronic device is greater than a distance threshold and less than a first proximity threshold; if so, acquiring light-sensing data; determining whether the light-sensing data is greater than a preset light-sensing value; if so, determining a second proximity threshold; and determining the positional state between an obstruction and the screen of the electronic device based on the second proximity threshold and the data acquired by the infrared distance sensor. This disclosure, based on data acquired by the infrared distance sensor, the first proximity threshold, the distance threshold, the light-sensing data, and an oil stain algorithm, performs multi-level precise differentiation of the positional state between the obstruction and the screen of the electronic device, effectively avoiding false alarms caused by misinterpreting obstructions as oil stains for optimization, and improving the response accuracy of the electronic device's screen.

This invention relates to the field of marine water sample detection and control technology, and discloses an automated control system for stratified marine water sample collection, including parameter acquisition, feature recognition, spatial retrieval, and sampling control modules. The feature recognition module uses pressure feedback to convert the medium property sequence into differential features and extracts dynamic control boundaries to offset the heave and oscillation of the controlled detection components. The spatial retrieval module constructs a three-dimensional parametric surface based on the dynamic control boundaries and discretizes it to generate an octree topology space. It establishes a hash map index table that corresponds to the trigger state and the memory addressing base address. By addressing the base address with a single instruction, the algebraic calculation process is eliminated, and the trigger delay jitter caused by the calculation is eliminated. This invention realizes the transformation from fixed depth triggering to active physical and chemical boundary tracking control mode, eliminates instruction lag and decouples spatial fluctuation interference, and ensures that the sampling action is accurately locked to the physical and chemical boundary.

The utility model discloses an air conditioner connecting line processing and shearing device, a tension adjusting device comprises a lifting operation table and a lifting air cylinder for driving the lifting operation table to lift, and an adjusting pulley is rotatably mounted on an adjusting seat; the adjusting cylinder drives the adjusting seat to be close to or away from the corresponding fixed pulley so as to maintain the tension of the wire rod to be stable. The pulley block is integrated on the liftable lifting operation table, and the adjusting seat driven by the adjusting cylinder is combined, so that the pulley block can dynamically follow the wire height change caused by the release point change of the unwinding disc, the pulley block and the wire path are always kept on the same horizontal plane, and the inclination or bending of the wire due to the height difference is effectively avoided; the friction damage risk is reduced, the response precision and stability of tension adjustment are remarkably improved, it is ensured that tension is uniform and stable in the wire conveying process, meanwhile, equipment abrasion is reduced, and the machining efficiency and the product quality consistency are improved.

The application discloses a cyanobacterial community succession prediction method and device fusing ecological mechanism and causal enhancement. The method comprises the following steps: collecting multi-source high-frequency monitoring data of a target water area, and constructing an environmental driving factor characteristic matrix; constructing a causal weight matrix through causal relationship aggregation analysis, and outputting a causal enhancement characteristic matrix; constructing a species ecological strategy embedding matrix; performing multi-scale decomposition on algal density time series data; constructing a multi-species interaction prediction model based on a multi-head self-attention mechanism, fusing environmental characteristics, causal enhancement characteristics, ecological strategy embedding and multi-scale time series characteristics, and predicting the biomass and relative abundance of each dominant cyanobacterial species; and outputting a cyanobacterial community succession process and a water bloom risk grade. Through deep fusion of causal enhancement, ecological strategy embedding and multi-head attention mechanism, the precision, stability and ecological interpretability of cyanobacterial community succession prediction are significantly improved, and the application is suitable for water bloom early warning and ecological management of lakes, reservoirs and rivers.

This application provides a data transmission method and apparatus, to improve transmission accuracy of a response message. The method includes: receiving, by a network device, uplink data sent by a terminal device based on a grant-free manner; and sending, by the network device, a response message for the uplink data to the terminal device according to a type of the response message for the uplink data, where the type of the response message for the uplink data is a first response type or a second response type, the first response type is that each time the terminal device sends the uplink data, the network device sends the response message for the uplink data to the terminal device once, and the second response type is that after the terminal device sends the uplink data for a plurality of times, the network device sends, to the terminal device once, the response message for the uplink data that is sent for the plurality of times.

The invention relates to a drug interaction type prediction method fusing a multi-source feature cross-graph structure, and belongs to the technical field of deep learning. The method comprises the following steps: acquiring drug data to obtain a drug data set, and preprocessing the drug data set; performing Jaccard similarity calculation on the drug data to obtain chemical structure, biological target and enzyme action multi-dimensional data representation of the drug, and further obtaining initial feature representation of the drug; updating the initial feature representation of the medicine through a heterogeneous graph to obtain a medicine representation fused with a topological structure; the drug representation of the fusion topological structure fuses drug characteristics of different sources through an attention mechanism to obtain a final drug pair representation; the drug representations of the fused topological structures corresponding to the two drugs in the drug pair pass through a feature interleaving layer to obtain a final comprehensive representation; and the final drug pair representation and the final comprehensive representation are processed by a prediction module to obtain a drug interaction prediction probability vector. According to the invention, the accuracy and efficiency of drug interaction prediction are optimized.

This invention provides a method and apparatus for deep inference of an archipelago through computing power on an intelligent computing center cloud platform. It relates to the fields of intelligent computing centers, smart computing centers, computing power infrastructure, and smart computing cloud technologies. The method includes: Step S1, generating N initial response data corresponding to an initial problem, where N is a positive integer greater than 1; Step S2, dividing the N initial response data into M islands, each island including K initial response data, where M and K are positive integers greater than 1; Step S3, evolving the K initial response data included in each island to obtain K evolved response data included in each island; Step S4, generating first inference response data based on the K evolved response data of each of the M islands; Step S5, outputting the first inference response data. This invention can significantly improve the accuracy of the response.

The invention discloses a transaction processing method and device, a storage medium and electronic device.The method comprises the steps that an initial transaction processing large model is created based on a basic large model, a sample user query prompt is input into the initial transaction processing large model for model training, and in the forward propagation model training process, a user query prompt is input into the initial transaction processing large model; the method comprises the following steps: performing transaction preprocessing on a sample user query prompt through an initial transaction processing large model to obtain a reference processing result and a sample reference query prompt, and performing transaction processing by adopting a cross attention mechanism based on the sample user query prompt and the sample reference query prompt to obtain a prediction processing result; and determining model loss based on the prediction processing result, the reference processing result, the sample user query prompt and the sample reference query prompt, and performing model parameter adjustment on the initial transaction processing large model based on the model loss to obtain the transaction processing large model.

This invention discloses a low-altitude intelligent target tracking method based on multimodal fusion, comprising the following steps: S1, constructing a multimodal input sequence to generate a fused input dataset; S2, inputting a semantic structure feature map into an improved TimeSformer model to output a predicted structure map; S3, constructing a consistency scoring matrix; S4, extracting candidate target samples whose trajectory distances meet a set similarity threshold and generating anchor point feature vectors; S5, extracting target feature vectors and historical identity feature vectors to generate an identity similarity score; S6, performing identity regression judgment based on the identity similarity score; S7, recording the fusion output result and identity label status of each frame to generate a target tracking trajectory sequence and an identity record set. This invention achieves accurate tracking under complex occlusion and multimodal interference conditions in low-altitude scenes, possessing advantages such as high tracking robustness, strong identity recovery accuracy, and excellent anomaly response time.

This invention relates to a distributed energy collaborative optimization method based on a cloud-edge-device architecture, belonging to the field of smart microgrid technology. Technical solution: S1. Define a four-layer cloud-edge-device architecture; S2. Information transmission stage: Information is reported upwards and circulated and transmitted at the same level; S3. Demand optimization stage: Demand is decomposed downwards and optimized collaboratively at the same level; S4. Energy flow stage: Energy is aggregated upwards and interacts and communicates at the same level. This invention, through its designed four-layer cloud-edge-device architecture, enables distributed energy to shift from centralized control to active participation in demand response. Based on this, a distributed energy collaborative optimization method based on the cloud-edge-device architecture is further designed, achieving full-domain collaborative optimization of distributed energy. Firstly, it reduces the response error of distributed energy; secondly, it effectively reduces the response time of distributed energy, solving the problem of excessive computing power in traditional centralized control; and thirdly, it reduces the operating cost of distributed energy, addressing the high operating cost of distributed energy.

This utility model belongs to the field of low-altitude flight technology and provides an electric pitch control mechanism for aircraft propellers. It includes a propeller assembly, a propeller hub, a drive assembly, and a pitch control module. The propeller assembly is connected to the propeller hub, and the pitch control module is installed inside the hub. A sensor assembly is mounted on the pitch control module to collect real-time changes in the propeller's angle or distance. The propeller assembly is connected to the pitch control module and uses the sensor assembly to collect and provide real-time feedback on these changes, thereby controlling the pitch angle. This patent uses sensors to collect the rotation angle of the lead screw shaft, the displacement of the pitch slider, and directly collect the pitch angle, enabling real-time feedback and control of the pitch angle. Its structure is simple, and it achieves stepless control of the pitch angle. Adjusting washers are used to adjust the clearance of the lead screw shaft within the propeller hub, improving the response speed of the lead screw shaft and the response accuracy of the propeller's pitch angle.