24 results about "Error correction model" patented technology

This invention relates to an intelligent processing system and method for energy conversion, belonging to the field of energy conversion and control technology. It involves real-time acquisition of operational data through a sensor array; edge computing nodes preprocessing data using a sliding window averaging method and a cross-correlation delay estimation method to extract equipment health, energy output efficiency, and environmental coupling characteristics; a digital twin model employing a mechanistic model and an LSTM error correction model in parallel computation to output a sequence of predicted future states; a central decision server using a deep deterministic policy gradient algorithm to solve for the optimal control action with a multi-objective function encompassing efficiency, safety, and cost; a coordinating controller issuing commands to adjust the actuators; and repeating the above steps to form a closed-loop control. This invention reduces system response time and improves energy conversion efficiency through a collaborative architecture of digital twin prediction, reinforcement learning decision-making, and knowledge graph enhancement.

The application discloses a carbon footprint tracing system fusing a knowledge graph and space-time attention, which comprises a multi-source data fusion module, a power grid, product and auxiliary data are used to obtain space-time characteristics and a knowledge graph original data set; an electric carbon factor dynamic prediction module, preliminary prediction is carried out through a fusion model, and a dynamic error correction model is used to calibrate to obtain a final electric carbon factor; a knowledge graph construction module, based on preset entities and relations and attributes, extracts and fuses knowledge elements from original data, and constructs a special knowledge graph; a carbon footprint accurate accounting module, based on the correlation of the special knowledge graph, automatically matches the final electric carbon factor with the power consumption of each production process of the product, combines non-power carbon emission data, and obtains the total amount of product carbon footprint through an accounting model; and a carbon footprint tracing module, based on the special knowledge graph, traces the production link of the total amount of product carbon footprint. The application improves the product carbon footprint accounting accuracy and full-link tracing capability.

The application provides a high-precision calibration method, device and equipment for mapping camera interior orientation elements and a storage medium. The method comprises the following steps: firstly, using a collimator and an electronic theodolite to independently calibrate the rotary table angle positioning error and the parallel light tube optical axis pointing deviation, respectively, to generate an error correction model and a space angle matrix, and converting the device system error into a known quantity that can be compensated; secondly, using a positive and negative mirror center-finding method to iteratively converge to determine the accurate intersection point of the sight axis and the target surface as the main point initial observation value and the sight axis reference angle, and eliminating the uncertainty of the main point; finally, starting from the reference angle to perform full field scanning and collection, and jointly substituting the corrected angle data, the angle matrix and the main point observation value into the collinearity condition equation for overall adjustment solution, so that the interior orientation elements and the distortion parameters purely reflecting the characteristics of the camera itself are solved under the premise that various system errors have been stripped one by one.

This invention discloses a training-free method for improving GUI localization capabilities based on an operation chain, relating to the fields of human-computer interaction and intelligent software agent technology. The method includes: acquiring a target interface image and natural language commands; obtaining initial candidate regions through a basic localization model; performing local cropping followed by secondary prediction and mapping back to the original image to update the region; applying a dynamic mask to the image and repeatedly predicting to generate spatially mutually exclusive candidate boxes; constructing a labeled comparison image; combining the input error correction model with commands and rule-based prompts; and selecting the optimal candidate box as the final localization result. This invention significantly reduces accuracy and ambiguity biases in localization by performing a series of interpretable inference time-series operations on the input screenshot and natural language commands, thereby improving the executable localization capabilities and stability of desktop intelligent agents on edge devices such as computers and mobile phones.

The application discloses a misalignment error correction model construction method, a misalignment error correction method and a system, relates to the technical field of optical element interference measurement, and has the technical scheme as follows: a sample fringe pattern without misalignment error is acquired; an optical path difference of different types of misalignment errors is applied to the sample fringe pattern, error quantity and error direction of the sample fringe pattern are adjusted, a plurality of sample interference patterns are obtained, and each sample interference pattern and corresponding error quantity, error direction and misalignment error type are one-to-one corresponding, so that a training data set is obtained; wherein, one sample interference pattern corresponds to one type of misalignment error; a neural network to be trained is trained by using the training data set, and when a loss function converges, a misalignment error correction model is obtained. The misalignment error correction model provided by the application improves the precision of misalignment error correction of a large-NA spherical optical element.

The application provides a regional crop yield estimation method, system, device and storage medium, adopts a random forest model to simulate the yield of crops in a region, calibrates the model error based on a rotation correction algorithm, and predicts the yield of crops in the region through the random forest model after correction. The rotation correction algorithm can effectively calibrate the general error of the existing crop yield estimation model in the high / low yield area. Compared with the existing error regression correction algorithm and machine learning correction algorithm, the rotation correction algorithm further corrects the error regression model of linear fitting to improve the correction effect in the low / high value area. Compared with superimposing a machine learning error correction model, the rotation correction algorithm has stronger generalization ability, avoids overfitting phenomenon caused by using a too complex machine learning model during training, and has better correction effect and wider applicability.

The application discloses a kind of complex terrain area air pollution multi-source heterogeneous data fusion and space-time alignment method.For the problem that multi-source data is inconsistent in time scale, spatial scale is not matched and alignment error is large under complex terrain, first, the application obtains multi-source data and pre-processes;Then realize time alignment by using the method based on time distance and double weight of data quality;In spatial alignment, introduce the terrain factor of elevation, slope, aspect for nonlinear modulation;Then get collaborative expression by terrain constraint fusion and multi-source feature fusion;Finally, construct ensemble learning error correction model to correct space-time deviation, output unified dataset.The application significantly improves the spatial alignment accuracy of complex terrain area, reduces space-time alignment error, and provides reliable data basis for air pollution mechanism analysis and prediction.

This invention belongs to the field of medium- and long-term power generation in new energy, and discloses a medium- and long-term power forecasting method and related equipment based on ERA5 reanalysis data. This method uses multi-year ERA5 reanalysis data to perform climate model analysis on the area where the power plant is located, obtains the climatological average irradiance, and constructs an error correction model by combining it with historical measured data from the power plant. This allows for medium- and long-term power forecasting using historical data from the same period of the power plant's climatological output. By using multi-year ERA5 reanalysis data to perform climate model analysis on the area where the power plant is located, this method can accurately capture the climatological average irradiance of the area, effectively solving the forecasting problem caused by the lack of historical data for newly built or short-term operating photovoltaic power plants. Using this method is of crucial practical significance for power plants to reasonably declare medium- and long-term trading volumes in electricity transactions and reduce revenue losses caused by positive and negative volume deviations, significantly enhancing the competitiveness of power plants in the electricity market.

The application discloses a document specification review method and system, comprising: obtaining a to-be-reviewed document and preprocessing, calling a pre-constructed professional vocabulary error correction model of a corresponding field to perform error word identification, and obtaining an error word identification result; judging whether the pre-processed to-be-reviewed document contains a reference standard, if the reference standard is not identified, generating a reference standard missing prompt; if the reference standard is identified, extracting information of the reference standard, comparing the identified reference standard information with a pre-constructed standard specification database, and judging whether the reference standard in the to-be-reviewed document is a valid version; if the reference standard in the to-be-reviewed document is a valid version, performing content consistency comparison between the to-be-reviewed document and the pre-constructed standard specification knowledge base, and judging whether there is a description error. Through the method, error words and standard specifications can be identified, and the identification accuracy and efficiency are improved.

The application discloses an arch dam valley width measurement method, device, system and storage medium. The method comprises the following steps: acquiring meteorological data of a plurality of reference points, instrument installation points, elevation highest points and elevation lowest points, acquiring elevation information and measurement values of a plurality of valley width endpoints, including slant distance, vertical angle and horizontal angle; based on the meteorological data of the elevation highest points and the elevation lowest points and the elevation information of the valley width endpoints, meteorological data at the valley width endpoints is obtained by inversion; an error correction model representing the mapping relationship between the meteorological data and the measurement error values is constructed by using a data-driven error modeling algorithm, the measurement values of the valley width endpoints are corrected based on the error correction model, and corrected measurement values are obtained; and the valley width length is calculated based on the corrected measurement values. The application significantly reduces the labor and instrument use cost of high arch dam valley width measurement, and improves the measurement precision and engineering applicability.

This invention discloses a Chinese spelling error correction method. The method increases training difficulty by replacing common words and characters in the original dataset with easily confused words and characters, thereby increasing the accuracy of model predictions. Simultaneously, the error correction model provided by this invention is optimized by integrating glyph and pinyin information based on the Macbert model, resulting in a more accurate error detection score and thus a more accurate identification of the corrected character. When the error correction model cannot obtain a suitable corrected character for a specific position in a sentence, this invention introduces a binary classification model to predict each candidate character at that specific position to obtain the corrected character, thereby enabling more accurate and efficient error correction of Chinese spelling.

The application provides a multi-pose pedestrian indoor positioning method fused with visual information, comprising the following steps: S1. establishing a mobile phone holding pose recognition model; S2. determining the mobile phone holding pose by using the mobile phone holding pose recognition model, and then compensating the heading angle of the PDR algorithm; S3. introducing a judgment algorithm to remove the meaningless compensation heading angle in the pose conversion process and redetermine the compensation heading angle; S4. based on the compensation heading angle redetermined in step S3, constructing a multi-pose PDR and outputting a positioning result; and S5. designing a PDR error correction model based on visual position recognition, eliminating cumulative errors and further improving the positioning accuracy of the multi-pose PDR. The multi-pose pedestrian indoor positioning method fused with visual information can adapt to the pose transformation of pedestrians and can correct the cumulative errors in real time, can enhance the precision and robustness of the pedestrian indoor positioning in the case of reducing the burden of the positioning system on the pedestrians and compressing the positioning cost to the greatest extent.

The application relates to the computer technical field and is particularly applied to a motion control scene in an artificial intelligence system in a production field. A simulation machine actuating cylinder motor power regulation method and system, an electronic device and a storage medium are disclosed, which comprise the following steps: setting basic parameters and establishing a multi-dimensional sensor error correction model; collecting and calibrating multi-source data such as motors, cockpit attitudes, loads, environments and abnormal sounds in real time; calculating real-time power requirements and power adaptation degrees based on a coupling model of operating actions, loads and power; fusing a tensor self-encoder and a gradient boosting tree to perform power deficiency risk prediction and abnormal sound grading early warning; adopting a multi-modal power regulation mechanism according to the prediction and determination results to implement temperature rise, overload and frequent regulation protection; collecting data after regulation to close loop correct the model and perform incremental updating. The scheme suppresses abnormal sound, improves power matching and system stability, and supports rapid adaptation of different types of simulators.

This invention discloses a model structure error correction method that couples data-driven approaches with physical constraints. The invention constructs a corrected groundwater model, comprising a groundwater numerical model, a structure error correction model, and observation errors. Markov chain Monte Carlo simulation is used to identify the posterior distributions of the parameters of the groundwater numerical model and the hyperparameters of the structure error correction model. Based on these posterior distributions, the prediction results of the corrected groundwater model are obtained. The prediction performance of the corrected groundwater model is then evaluated. This invention constructs a novel combinatorial likelihood function, explicitly integrating physical mechanism constraints into the construction process of the error correction model. This effectively constrains the structure error model while correcting structural deviations in the groundwater model, thus significantly improving the performance of the corrected prediction results in conforming to physical mechanisms.

This invention discloses a training method for a non-standard text error correction model. It involves constructing a non-standard text error correction training set, which includes multiple error correction training instructions and corresponding error correction training text pairs for each instruction. Each error correction training text pair includes both non-standard and standard training texts. Using this training set, a large language classroom model is fine-tuned to obtain a target large language teacher model. Based on the target large language teacher model and the non-standard text error correction training set, a lightweight student model is trained using knowledge distillation until convergence, resulting in a trained target lightweight student model. This target lightweight student model is used to obtain the standard text result of the non-standard text to be corrected. This method improves the accuracy of the standard text result of the non-standard text to be corrected, effectively achieving error correction processing of non-standard text in scenarios with high real-time requirements or limited resources.

This invention relates to the field of radio navigation technology, and more particularly to a scene-adaptive integrated positioning method and system. The method includes: when the unmanned vehicle's GNSS receiver is working normally, using inertial navigation and GNSS integrated navigation to output high-precision position results; calculating high-precision ranging values ​​between every two unmanned vehicles, and using these values ​​along with the data link ranging values ​​between every two unmanned vehicles as a training dataset, obtaining a data link ranging error correction model through machine learning; when the unmanned vehicle's GNSS is not working normally, calculating high-precision ranging values ​​between every two unmanned vehicles other than the unmanned vehicle itself, obtaining the data link ranging error of the unmanned vehicle based on the data link ranging error correction model, correcting the unmanned vehicle's data link ranging value, and obtaining a high-precision data link ranging value; and obtaining a high-precision position based on the high-precision data link ranging value using the inertial navigation and data link integrated navigation method. This invention improves the reliability and accuracy of navigation information.

The application discloses a text error correction method and device, electronic equipment and a storage medium. The method comprises the following steps: obtaining a text to be corrected; correcting the text to be corrected based on a trained error correction model to obtain a corrected text; wherein the trained error correction model is obtained by training an initial error correction model using training text; the training text is generated based on a text processed from a first initial text and a text processed from a first modified text; and the first modified text is a text modified from the first initial text. Thus, the trained error correction model can be used to intelligently correct the text to be corrected, thereby improving the accuracy of the text.

The application relates to the technical field of image data processing, in particular to an enterprise financial tax data processing method based on image recognition, which comprises the following steps: performing quality evaluation on financial tax bill images through a lightweight neural network and dynamically scheduling the images to different processing channels; accurately positioning and extracting a key information region by using a target detection and attention mechanism segmentation network; constructing a multi-modal error correction model, combining original text recognized by a character recognition engine with image layout features and financial tax field semantic logic to perform collaborative verification and intelligent error correction; automatically filling corrected information into a structured template, and performing verification on a hooking relationship of an embedded rule base to output data results or mark abnormalities for manual review. Therefore, the problems of rigid enterprise financial tax data processing flow, single template information and poor adaptability to complex scenarios are solved, and the overall efficiency, accuracy and robustness of financial tax bill processing are improved.

This application provides a method, system, equipment, and medium for ultra-short-term forecasting of wind power clusters under extreme weather conditions based on Beaufort wind scale, relating to the field of wind power prediction technology. The method includes: dividing wind speed into low, medium, and extreme wind speed ranges based on Beaufort wind scale and wind turbine operating parameters; matching corresponding correlation metrics to each of the three wind speed ranges to perform correlation filtering on wind power station data, obtaining filtered wind power station data carrying wind speed type identifiers; inputting the filtered wind power station data into the corresponding prediction branch according to the wind speed type to obtain initial power prediction features for each range; inputting the wind power features into the corresponding error correction model according to the wind speed type to obtain error correction features for each range; and adding and fusing the initial power prediction features and the corresponding error correction features to obtain the final predicted power. This application can improve the accuracy and stability of ultra-short-term power prediction for wind power clusters under extreme weather conditions.

The present application relates to a kind of silicon oil fan clutch slip rate prediction method, belong to vehicle thermal management and clutch state prediction technical field.The method obtains driving end speed, fan speed, target fan speed, silicon oil temperature and ambient temperature;According to silicon oil temperature, calculate silicon oil viscosity correction coefficient, calculate target slip rate and actual slip rate;Establish transmission torque estimation model, and solve balanced fan speed in combination with fan load curve;According to balanced fan speed output prediction slip rate;Again, the error correction model parameter between actual slip rate and prediction slip rate is utilized.The method can couple silicon oil viscosity-temperature characteristic, fan load curve and driving end-from driving end speed relationship, improve the accuracy and adaptability of slip rate prediction under different conditions.

The application relates to the technical field of high-temperature gas flow field measurement, in particular to a thermal resistance network thermocouple rake measurement temperature correction method, wherein the method comprises the following steps: establishing a computational fluid dynamics model, performing numerical simulation under a plurality of preset incoming flow conditions, and obtaining initial data of the computational fluid dynamics; obtaining a corrected numerical method and a turbulent flow model according to a first error correction model of calibration test measurement data and the initial data; constructing a thermal resistance relationship network and a thermal resistance network model; adjusting nodes and modules in the thermal resistance relationship network according to a second error between simulation data and output data of the thermal resistance network model, so as to obtain a thermal resistance relationship network matched with the simulation data and a correction coefficient of a heat transfer process correlation formula; and calculating a measurement temperature correction value of the thermocouple rake by using the thermal resistance network model. Therefore, the problems that related technologies only rely on a computational fluid dynamics correction scheme with high time cost, only rely on an empirical correlation formula with low correction precision, only rely on test calibration with high economic cost and the like are solved.

A text error correction method, comprising: acquiring target text to be corrected (S101); and further inputting said target text into a target error correction model, so as to obtain target corrected text (S102), wherein the target error correction model is obtained by training a student model on the basis of a teacher model and sample data, and the sample data comprises sample text to be corrected, sample correct text corresponding to said sample text, and context information corresponding to the sample correct text.

The application discloses a search association word generation method and device, an electronic device and a computer readable storage medium. The method comprises the following steps: inputting a plurality of search words in a search word set into a trained error correction model to obtain error correction words corresponding to the search words and error correction probabilities; clustering the search words corresponding to the same error correction word by taking the error correction word as a clustering center; generating a vector representation corresponding to the search words, and calculating a similarity score corresponding to the search words according to the similarity between the vector representation of the search words and the vector representation of the clustering center of the category to which the search words belong; inputting the error correction probability and the similarity score corresponding to the search words into a trained discrimination model to discriminate whether the search words belong to the corresponding category, and removing the search words not belonging to the corresponding category to obtain a final clustering result; obtaining a target search word, and generating search association words corresponding to the target search word according to the final clustering result. The application improves the accuracy of generating search association words.