22 results about "Real error" patented technology

The invention discloses a dynamic error real-time compensation method and system for a heavy-load vertical machining center, and belongs to the technical field of high-end numerical control equipment, precision manufacturing and intelligent control. Inputting the state vector into a dynamic error model to solve a three-dimensional space dynamic error vector; processing the error vector to generate a real-time compensation instruction; a compensation instruction is injected into the numerical control system to correct the machining track online; and obtaining a real error to update the dynamic error model on line. According to the method, the technology of combining multi-physics field data fusion and a neural network agent model is adopted, frequency decoupling and dual-channel compensation injection are performed on errors, and an online model self-optimization feedback closed loop is established, so that real-time and high-precision compensation on multi-source coupling dynamic errors such as thermal-induced and force-induced multi-source coupling dynamic errors can be realized; and the limit machining precision and stability under the heavy-load machining condition are remarkably improved.

The invention discloses a time-varying precision customized fine self-healing method based on virtual-real homeomorphic evolution, and relates to the technical field of machine tool precision control, and the method comprises the steps: constructing a homogeneous transformation matrix according to the position relation of all parts of a machine tool, and obtaining an error unified mathematical model based on a small error theory; then, a measuring device is used for obtaining small errors of movement and rotation angles and reverse clearance errors of a translation shaft; and then constructing a virtual-real homeomorphic digital twin platform, and integrating the model and the error to realize virtual-real error homeomorphic evolution. Acquiring real-time sensing data, predicting and evaluating precision time variation in combination with a time sequence, and driving virtual-real synchronous evolution to generate a self-healing path; finally, positioning deviation is determined according to path start and end error changes and an error model, and the path is transmitted when a preset value is exceeded; and the time delay defect of traditional real-time compensation is relieved through advanced prediction / deduction. According to the method, unified error modeling of the multi-axis machine tool is achieved, the defect of reverse gap neglect is overcome, errors caused by compensation lag and blind compensation are solved, and the machining precision is guaranteed.

The invention discloses an adversarial agent-based number-real consistency error adaptive prediction method and system, and belongs to the technical field of information physical system and artificial intelligence crossing. According to the method, a destructive agent for maximizing a number-real consistency error and a prediction agent based on a pre-training language model are constructed for continuous confrontation, in the interaction process, the destructive agent intelligently generates disturbance through reinforcement learning so as to explore a weak link of a system, and the prediction agent predicts a disturbance result; after the system obtains the real error, on one hand, a strategy of destroying the agent is awarded and updated, and on the other hand, the real error is utilized to carry out supervised learning type increment fine adjustment on the prediction agent. Through the closed-loop mechanism of double-agent coevolution, the prediction precision and generalization ability of the data-real consistency error can be significantly improved, and the system is endowed with efficient and online adaptive learning ability, so that the reliability and safety of the data-real fusion system can be comprehensively improved.

The invention relates to a supervised learning method based on an improved wind field direction loss function in deep learning, and the method comprises the steps: redefining a wind direction error between a correction result and reference data, and controlling the range of the wind direction error between-180 degrees and 180 degrees; and meanwhile, the wind direction data is normalized before the model is input, and the normalized error range is redefined to be between [-1, 1]. The loss function is used for calculating the redefinition error of the normalized data. Experiments show that the improved loss function can better reflect the real error between the correction data and the reference data, thereby ensuring that the correction result of training quickly approaches the reference data.

The application discloses a kind of based on statistical error analysis's gate group timing flow abnormal diagnosis method, comprising the following steps, S1, flow data pre-processing based on filtering method;S2, flow error distribution analysis;S3, based on k up / down Reasonable flow error range determination of σ rule;S4, over gate flow coefficient relationship curve fitting;S5, theoretical gate station flow calculation;S6, abnormal flow data diagnosis and correction.The advantage is: the method can make full use of the specific monitoring error distribution characteristics and monitoring variable mapping relationship of different gate stations, reduce the missed detection or false detection rate.Through introducing the real error distribution of flow monitoring data in the abnormal data diagnosis model, the specific monitoring error range of specific project is determined, and the reliability of flow abnormal diagnosis process is improved.

The present invention relates to the field of chip verification technology, and in particular to a method, electronic device, and storage medium for error injection. The method initializes each target model, including: the target model obtains a model instance of the error injection model; the target model registers all of the target model's own hardware error types to the error injection model through the model instance; the stimulus end injects some or all of all registered hardware error types into the error injection model and saves the information; when the target model is simulated or tested and reaches an error occurrence point, the hardware error type saved in the error injection model is read, and if the reading is successful, it is determined that a corresponding hardware error has occurred; otherwise, no corresponding hardware error has occurred. This achieves the purpose of simulating real errors and ensuring the functional integrity of the target model.

The application discloses a local differential privacy publishing method of data statistical result under continuous monitoring, and the method is used for aggregating and statistically processing the data provided by a data provider through a heuristic budget allocation method of fully utilizing user budget allocation mode, remaining budget taking half and budget absorption, so as to significantly improve the utility of the data statistical result under continuous monitoring. A two-stage heuristic algorithm provides key information of publishing strategy selection by evaluating whether the real statistical result of the current timestamp data is significantly fluctuated compared with the last timestamp, and uses the statistical variance of the agreed protocol to evaluate the real error based on the characteristics of the real data of the data provider which is invisible to the data collector, so as to reasonably estimate the error of the current aggregation statistical result in the scene where the data collector cannot access the real data, compare the error with the lazy publishing error, select the most suitable publishing strategy, significantly improve the utility of the data under the condition of ensuring the privacy protection of the data, and reduce the communication cost and overhead.

The invention relates to the technical field of data processing, and discloses an error log screening method, device and equipment and a computer readable storage medium. The method comprises the following steps: in a process of testing a memory chip, generating an error log, and obtaining an error data bit and an actual error address corresponding to the error log; obtaining a target data bit, a target address mask and a target error address; performing comparison operation on the target data bit and the error data bit to obtain a comparison result, and performing matching operation on the target address mask, the target error address and the actual error address to obtain a matching result; and according to the comparison result and the matching result, determining whether an error log is reserved. By comparing the target data bit with the error data bit and matching the target address mask, the target error address and the actual error address, whether the error log is reserved or not is determined, and the error log screening accuracy is improved.

The embodiment of the invention provides a spacecraft flight attitude angle correction method and device and electronic equipment. The method comprises the following steps: processing a real-time state space formed by a real-time error of an inertial navigation system and a real-time error of a starlight guidance system by using a target correction model to obtain a compensation value of a mathematical platform misalignment angle, correcting the mathematical platform misalignment angle by using the compensation value, and correcting the flight attitude of a spacecraft by using the corrected mathematical platform misalignment angle. Wherein the real-time state space directly takes real error data as input and does not need to depend on a preset fixed model, so that when noise dynamically changes or model errors exist, the system state can be captured through real-time error feedback, and filtering divergence is reduced. In a high dynamic scene, errors caused by linearization can be reduced. Furthermore, the compensation value corresponds to the real-time state space, and the compensation value is dynamically changed, so that the error can be always in a controllable range by reducing an error accumulation chain.

Embodiments of the present application provide a spacecraft flight attitude angle correction method and device and electronic equipment. The method comprises: using a target correction model to process a real-time state space composed of real-time errors of an inertial navigation system and real-time errors of a starlight guidance system to obtain a compensation value of a mathematical platform misalignment angle, correcting the mathematical platform misalignment angle using the compensation value, and correcting the flight attitude of the spacecraft using the corrected mathematical platform misalignment angle. The real-time state space directly uses real error data as input and does not need to rely on a preset fixed model, so when the noise dynamically changes or the model error, the system state can be captured through real-time error feedback, and filter divergence is reduced. In a high dynamic scene, the error caused by linearization can be reduced. Further, the compensation value corresponds to the real-time state space, and the compensation value is dynamically changing, so that the error accumulation chain can be reduced, and the error can always be within a controllable range.

The invention relates to the technical field of airport analog simulation, in particular to an adaptive optimization method for airport dynamic simulation parameters based on SAC reinforcement learning, and the method comprises the steps: combining an SAC algorithm with an improved Transform model for multi-modal spatial-temporal feature extraction; static configuration and multi-time-frame real-time states of various business objects such as airport flights, airport positions, runways, guarantee vehicles and personnel are used as input, and parameters such as release intervals, sliding path priorities and resource allocation modes required by simulation are output through a strategy network; a strategy is continuously optimized in a simulation-real error closed loop by using a double-Q network and an entropy regularization mechanism, and key control parameters such as runway distribution, sliding path priority, release interval and guarantee vehicle and personnel scheduling are learned and adjusted online in real time from the simulation-real error closed loop; therefore, the simulation precision and the system robustness are ensured in a complex dynamic environment.

The application discloses a time-varying precision customized fine self-recovery method based on virtual-real isomorphism evolution, relates to the technical field of machine tool precision control, and comprises the following steps: firstly, according to the positional relationship of each component of a machine tool, a homogeneous transformation matrix is constructed, and a unified mathematical model of errors is obtained based on a small error theory; secondly, a measuring device is used to obtain small errors of movement and rotation angle and reverse gap error of a translational axis; then, a virtual-real isomorphism digital twin platform is constructed, and a model and errors are integrated to realize virtual-real error isomorphism evolution; real-time sensing data is acquired, time series prediction is combined to evaluate precision time variation, a self-recovery path is generated by driving virtual-real synchronous evolution; finally, positioning deviation is determined according to error variation at the beginning and end of the path and the error model, and the path is transmitted when the error exceeds a preset value; in addition, the time lag defect of traditional real-time compensation is relieved through advanced prediction / deduction. The method realizes unified error modeling of a multi-axis machine tool, makes up for the defect of ignoring reverse gap, solves errors caused by compensation lag and blind compensation, and guarantees machining precision.

A computing device can assign possible error codes to subscore bins. The subscore bins can include a sent rate bin, a compliance rate bin, a fraud bin, a latency bin, and an engagement bin. The computing device can transmit network traffic messages and receive actual error codes associated with the network traffic messages. The actual error codes can overlap with and include some or all of the possible error codes. A subscore can be calculated for each subscore bin based on the actual error codes received. A deliverability score can be calculated based on the subscore for each subscore bin.

The invention discloses a multi-laser-head lineation error calculation method, device and equipment and a storage medium, and the error calculation method comprises the steps: carrying out the random error filtering of a first interval data set composed of the intervals of a plurality of lineation lines on a battery, and then calculating the real y-axis coordinate and the target y-axis coordinate of each lineation line, the method comprises the following steps: firstly, marking a plurality of laser heads, calculating to obtain a real error data set of a plurality of marking positions, then, obtaining a real x-axis coordinate and a real y-axis coordinate of each laser head, and calculating a real y-axis coordinate of each laser head during each marking to obtain a third coordinate data set, and finally, calculating a polynomial coefficient of a polynomial model to obtain a third coordinate data set. And the target y-axis coordinate of each lineation and the polynomial coefficient are utilized to predict the error of the positions of the plurality of lineation lines through the polynomial model, and the lineation positions of the plurality of laser heads are compensated, so that the lineation precision of the plurality of laser heads and the uniformity of lineation spacing can be effectively improved, and the error accumulation of the lineation of the plurality of laser heads is avoided.

The application relates to the technical field of data processing, and discloses an error log screening method, device, equipment and computer readable storage medium. The method comprises the following steps: in the process of testing a memory chip, generating an error log, and acquiring error data bits and an actual error address corresponding to the error log; acquiring target data bits, a target address mask and a target error address; performing a comparison operation on the target data bits and the error data bits to obtain a comparison result, and performing a matching operation on the target address mask, the target error address and the actual error address to obtain a matching result; and determining whether to retain the error log according to the comparison result and the matching result. By performing the comparison operation on the target data bits and the error data bits, and performing the matching operation on the target address mask, the target error address and the actual error address, whether to retain the error log is determined, and the accuracy of error log screening is improved.

The application discloses a kind of based on artificial intelligence's document text error identification control method and system, it is related to artificial intelligence technical field, the present application is by obtaining the document text error data in historical document set and dividing error type to construct historical error feature dataset;According to document type, split historical error feature dataset and train recurrent neural network, obtain the error identification baseline model matched with different types;When user writes, the similarity of real word and keyword is pushed error risk early warning;After user uploads document, match document type and call corresponding submodel to identify error, push alarm and calculate actual error rate;Estimate error rate is calculated in combination with historical average error rate, document complex coefficient and the like parameter, according to the deviation of actual error rate and estimate error rate, adjust error identification baseline model parameter, realize the accurate, dynamic identification and control of document text error, improve the pertinence, timeliness and precision of error identification.

The invention belongs to the technical field of electric meter measurement detection, and particularly relates to an electric energy meter aging process temperature and measurement accuracy anomaly detection method and system, and the method comprises the steps: obtaining source field data and target field data; performing field adaptive pre-training on the feature extractor and the error regression device according to a self-defined global loss function to obtain a pre-training model; generating a personalized fine tuning factor for the target ammeter according to the complexity, reflected by the target domain data, of the environment where the target ammeter is located; and on the basis of the personalized fine tuning factor, performing online fine tuning on the pre-training model deployed on the target ammeter to obtain a real error prediction value of the ammeter to be detected, and judging whether the aging process temperature or the metering accuracy of the ammeter is abnormal or not according to the real error prediction value. According to the invention, domain deviation can be eliminated from the pre-training stage, and adaptive online fine tuning is realized, so that the accuracy of metering error correction in the full life cycle of the electric meter is improved.

A computing device can assign possible error codes to subscore bins. The subscore bins can include a sent rate bin, a compliance rate bin, a fraud bin, a latency bin, and an engagement bin. The computing device can transmit network traffic messages and receive actual error codes associated with the network traffic messages. The actual error codes can overlap with and include some or all of the possible error codes. A subscore can be calculated for each subscore bin based on the actual error codes received. A deliverability score can be calculated based on the subscore for each subscore bin.

According to the tolerance digital twinning-based PSF simulation system and method provided by the invention, on the basis of the diffraction theory, a full-system optical transfer model based on the Huygens-Fresnel principle is constructed, and simplified processing of high-order aberration and diffraction terms by traditional geometric ray tracing is avoided; and then system error parameters are introduced as digital twin modeling variables, so that simulation models are in one-to-one correspondence with actual optical system parameters and have the capability of representing error ranges, and the limitation that in an existing method, a PSF model is separated from a real structure of an optical system and cannot cover a real error source of the system is overcome; high-fidelity PSF modeling of a specific optical system can be realized, and the method can be applied to high-precision imaging simulation and debugging verification.

The invention discloses a document text error recognition control method and system based on artificial intelligence, and relates to the technical field of artificial intelligence. Document text error data in a historical document set are obtained, and error types are divided to construct a historical error feature data set; splitting a historical error feature data set according to document types and training a recurrent neural network to obtain error recognition baseline models matched with different types; when a user writes, an error risk early warning is pushed through the similarity between notional words and keywords; after a user uploads a document, matching the document type, calling a corresponding sub-model to identify an error, pushing an alarm and calculating an actual error rate; an estimated error rate is calculated in combination with parameters such as a historical average error rate and a document complex coefficient, and parameters of an error recognition baseline model are adjusted according to the deviation between the actual error rate and the estimated error rate, so that accurate and dynamic recognition and control of document text errors are realized, and the pertinence, timeliness and precision of error recognition are improved.

This invention relates to the field of intelligent manufacturing technology and discloses a digital twin simulation design and verification method for improving mold accuracy. The method includes: discretely sampling the three-dimensional structure of the mold and constructing neighborhood relationships; extracting key measurement points to characterize structural changes; establishing a mapping between the theoretical and actual distances of measurement point pairs to achieve dynamic tracking of accuracy during multiple molding processes; further identifying unstable regions and performing predictive compensation through accuracy evolution trajectory and trend analysis; mapping the compensation results to a digital model to achieve synchronous updates between virtual and real data; subsequently constructing a local and global collaborative adjustment mechanism based on virtual and real error feedback; implementing adaptive adjustments to the control unit; and utilizing historical data to extract optimization strategies to adapt to environmental changes; and finally, through error field analysis and closed-loop correction, gradually converging the system to the target accuracy, achieving high-precision and intelligent control of the mold manufacturing process.

The invention discloses a multi-bit error positioning and correcting method and system, a computer and a storage medium, and the method comprises the steps: presetting two different generator polynomials, and respectively constructing optimization lookup tables; a syndrome corresponding to the original data is calculated, the optimization lookup table is queried according to the syndrome, and a single error position candidate set of the two channels is generated; a forced error bit is introduced and iterated, the optimization lookup table is queried according to an update syndrome corresponding to the forced error bit, and a double-error candidate set of the two channels is generated; presetting a mandatory error bit set, and carrying out the operation of the previous step on each mandatory error bit to generate a three-error candidate set of two channels; and obtaining an intersection of the single-error position candidate set, the double-error candidate set or the three-error candidate set of the two channels to obtain an actual error position. According to the method, two different generator polynomials are combined with a pre-constructed optimization lookup table technology to position errors in parallel, so that accurate positioning and correction of at most 3-bit random distribution errors are realized.