54 results about "Nonlinear evolution" patented technology

The invention discloses a geological disaster monitoring system based on multi-modal data, and relates to the technical field of geological disaster monitoring, and the system comprises a crack analysis module which carries out the time-space correlation mining of the crack propagation rate of a monitoring region, and analyzes the nonlinear evolution characteristics and spatial differentiation rules of rock mass fracture; and the critical identification module is used for performing wavelet packet energy spectrum analysis on the inclination angle change rate of the geologic body, identifying a critical turning point of rigidity attenuation of the geologic structure in combination with a preset algorithm, judging whether the overall stability enters an instability acceleration stage or not, and performing multi-parameter collaborative detection, crack evolution cross validation and dynamic trend stability verification to obtain the stability of the geologic structure. The accuracy and reliability of geological structure rigidity attenuation critical turning point recognition are remarkably improved, a more accurate rigidity attenuation stage judgment basis is provided for an early warning module, and the capturing capacity of a multi-modal data geological disaster monitoring system for structure instability precursor is enhanced.

The invention discloses a storm surge water increase prediction method and device, electronic equipment and a storage medium, and relates to the technical field of seawater monitoring, and the method comprises the steps: obtaining multi-source observation data of a target sea area, the multi-source observation data comprising satellite remote sensing data, near-shore monitoring data, meteorological mode data and drainage basin data; performing space-time alignment and exception processing on the multi-source observation data to generate standardized space-time grid data; performing hybrid prediction modeling on the standardized space-time grid data through a physical data dual-drive modeling layer to obtain a target probability water increasing field; and dynamically correcting the target probability water increasing field by using real-time observation data to obtain a storm surge water increasing predicted value. A physical mechanism model and a probability generation model are combined, limitation of a single model is overcome, and the nonlinear evolution process of the storm surge is effectively captured; the time-space dual-drive architecture is adapted to complex coast terrains and changeable meteorological conditions, and the risk of missing report or false report is reduced.

The invention discloses an environmental data processing method and system based on ocean engineering, and relates to the technical field of data processing, and the method comprises the steps: receiving an original observation data flow through a multi-source data preprocessing module, and carrying out the dynamic noise filtering and abnormal value adaptive detection; fusing the multi-source heterogeneous data through a multi-scale data fusion module, and embedding the fused multi-source heterogeneous data into a marine kinetic equation as a soft constraint; non-linear evolution features are extracted from the fusion data through a feature extraction and state representation module, and a high-dimensional environment state vector is constructed; real-time prediction of model parameters is executed through online learning and an inference engine; and performing uncertainty propagation calculation on the processing flow through a confidence evaluation module and generating a final environment state report. According to the method, the adaptive capacity of data preprocessing can be remarkably improved, the physical consistency of multi-source data fusion is improved, the nonlinear evolution law of ocean phenomena is accurately captured, and continuous online optimization and edge side low-delay response of model parameters are achieved.

The invention relates to the field of roadway surrounding rock deformation analysis, in particular to a roadway surrounding rock deformation intelligent monitoring method and device based on three-dimensional laser scanning. The method comprises the following steps: performing three-dimensional laser scanning on a roadway to generate equalized laser point cloud data; performing point cloud relative deformation detection and surrounding rock deformation dynamic modeling on the equalized laser point cloud data to construct a dynamic surrounding rock deformation model; performing space grid division on the dynamic surrounding rock deformation model, and performing differential change trend analysis to generate a multi-unit deformation trend; performing nonlinear evolution analysis and deformation situation prediction in a future time period on the multi-unit deformation trend to generate a deformation prediction result; and performing deformation situation interference simulation on the deformation prediction result, and extracting various interference simulation results. According to the invention, surrounding rock deformation data is accurately obtained in real time, and the deformation trend is dynamically analyzed and predicted, so that early warning is performed in advance, the safety risk is effectively reduced, and the safety and monitoring efficiency of underground engineering are improved.

The invention provides a radar rainfall real-time estimation system based on double-method fusion and ground verification, belongs to the technical field of meteorological radar rainfall estimation, and adopts double-method fusion of a physical model estimation module and a driving model estimation module to cooperate with a causal inference fusion engine to improve the precision and adaptability of rainfall estimation and improve the rainfall estimation accuracy. The physical model module adaptively selects a Z-R, ZDR-R or KDP-R relation according to the quality of radar-based data and rainfall scenes, drives the model module to capture space-time nonlinear evolution characteristics of rainfall echoes and predicts short-time rainfall dynamics, and the causal inference engine identifies differentiated characteristics of two estimation fields, inferes hail, vertical airflow and other physical hybrid factors, and performs real-time prediction on the hail, the vertical airflow and other physical hybrid factors. A comprehensive rainfall estimation field is generated, a continuous error correction field converted from ground station observation data is corrected through a space-time error atlas network, and finally rainfall estimation information is generated for real-time rainfall estimation, so that the adaptability to a complex weather field and the precision of the estimation field are improved.

The invention relates to a coal mine roof micro-fracture mobile monitoring and early warning method and system, and belongs to the technical field of coal mine safety monitoring. Denoising the vibration signal to obtain a denoised vibration signal; calculating the position of a fracture source according to the denoised vibration signal; clustering all fracture source positions to obtain clustered fracture sources; feature indexes are extracted from the clustered fracture sources; combining all the feature indexes to obtain a feature vector, and training an LSTM model by using the feature vector to obtain a roof instability probability prediction model; and monitoring the target coal mine roof by using the roof instability probability prediction model. According to the method, by introducing the LSTM deep learning model, a complex nonlinear evolution rule can be learned from high-dimensional time series data, the quantitative instability probability is output, upgrading from threshold alarm to intelligent prediction is achieved, and the false alarm rate and the missing report rate are remarkably reduced.

The invention discloses a transformer area high-frequency acquisition data quality improvement method, system and device and a medium, and the method comprises the steps: constructing a multi-level rule field, converting a rule into a continuous potential energy function with directivity and range control, and forming a regular force field in a semantic space; defining a field state function, and generating a plurality of field completion candidate values through rule-driven nonlinear evolution reasoning; a candidate value path diagram is constructed, nodes of the path diagram represent candidate values, and edge weights are determined according to semantic and rule association features among the candidate values; calculating the global path participation degree of each node based on the path diagram, and selecting an optimal complement value according to the global path participation degree; and executing multi-dimensional consistency verification and high-confidence judgment, and marking and supplementing the credibility level according to a verification result, thereby realizing closed-loop improvement of data quality. According to the method, the accuracy, the logic consistency and the high confidence of missing field supplementation of the high-frequency acquisition data of the transformer area in a complex interference environment are remarkably improved.

The invention belongs to the technical field of image processing, and relates to an egg surface microcrack detection method and system based on machine vision. The method comprises the following steps: acquiring a transmission image of an egg, performing spatial calibration, determining a geometric center and a long axis scale of an egg region, and establishing a polar coordinate mapping relation of each pixel point; according to the linear distance from each pixel point to the geometric center and the included angle relative to the long axis direction, the geometric sensitivity weight of each pixel point is evaluated, and the optical path distortion degree of the edge high-curvature area is represented; extracting contrast basic energy of the pixel points by using a local space window, and carrying out nonlinear evolution on a local gray gradient by combining with a geometric sensitivity weight to obtain an adaptive contrast energy evolution result; and coupling the local background variance and the geometric sensitivity weight, and calculating the crack identification confidence of each pixel point so as to carry out identification decision of the egg surface microcracks. According to the invention, extremely fine crack signals can be captured, and accurate detection of the microcracks on the surface of the egg is realized.

The invention discloses an optical fiber link physical parameter estimation method based on a multichannel feature input neural operator, and belongs to the technical field of optical fiber communication. The method comprises the following steps: acquiring signals of a transmitting end and a receiving end under different channel parameters, and respectively calculating multiple groups of weighting and observation data corresponding to loss, dispersion and nonlinear effect according to a nonlinear evolution equation for describing optical signal transmission; splicing the observation data and processing the observation data into a high-dimensional complex feature sequence; a frequency domain operator block comprising a parallel complex frequency spectrum and a time domain convolution branch is designed to extract and fuse the features, and aggregation representation of the channel features is obtained through Fourier transform and frequency domain attention pooling; and outputting estimated dispersion and nonlinear real number parameters through a plurality of multi-layer perceptron and a linear projection layer. According to the method, physical priori knowledge of optical fiber transmission is integrated into the deep neural operator, so that the parameter estimation capability and the model generalization performance of long-distance, strong-dispersion and strong-nonlinearity optical fiber links are improved.

The present application relates to the field of underground fluid intelligent control, and discloses a method for intervention and extraction of ion type ore body solute migration, comprising: establishing a feedback monitoring network for a fluid injection distribution system, obtaining pressure parameters and flow parameters generated by interaction between a controlled branch and a controlled medium; calculating dynamic impedance parameters of internal seepage resistance state of the controlled medium; extracting a follow-up change rate of the dynamic impedance parameters, identifying nonlinear evolution of pore structure induced by flow resistance mutation; comparing the follow-up change rate with a preset threshold value to determine bypass overflow risk; generating a flow dynamic distribution instruction and adjusting valve opening to make the local flow field stable within a preset flow resistance gradient range, wherein the present application solves the problem of sensing loss induced by formation porosity evolution through a dynamic flow resistance feedback intervention mechanism, drives fluid to penetrate into unconsolidated blind areas, and constructs a continuous and dense seepage interception interface.

The invention discloses an ultra-parallel microwave photon Isin machine system based on wavelength division multiplexing and an optimization method, and belongs to the technical field of photon calculation and combination optimization. The Isin machine system comprises a multi-wavelength light source module, a microwave modulation and wavelength division multiplexing coding module, a nonlinear calculation and evolution module and a demultiplexing and dynamic feedback module, all the modules are optically and electrically connected to form a closed-loop system, and different optimization problems can be coded to independent wavelength channels to achieve parallel solving. The optimization method is applied to the Isin machine system, and closed-loop optimization is completed through the steps of coupling matrix mapping, microwave modulation loading, nonlinear evolution, dynamic feedback control and the like. The method supports synchronous processing of 1024 optimization problems, and energy consumption of a single problem is lt; the method is mainly applied to the fields of logistics scheduling, financial modeling, communication network optimization and the like, and an efficient physical calculation solution is provided for a large-scale combination optimization problem.

The present application relates to the technical field of probe card signal optimization, in particular to a high-frequency probe card signal optimization method based on dynamic impedance matching and related devices. The method steps include: obtaining real-time high-frequency signal reflection parameters of the probe card test channel to provide real-time detection basis for dynamic impedance matching. The present application relies on the real-time acquisition of the high-frequency signal reflection parameters of the probe card test channel to complete the quantitative tracing of the dynamic impedance deviation, and through the closed-loop feedback adjustment, the integrated adjustable impedance network of the driving probe is used to complete the real-time correction of the output impedance. From the root level of transmission line impedance matching, the present application breaks through the inherent limitation of the traditional static matching scheme which can only adapt to the nominal working condition, and can effectively offset the dynamic impedance mismatch introduced by the contact state difference, the probe structure micro-deformation, the device individual parasitic parameter dispersion and the nonlinear evolution of the parasitic parameters under high frequency in the actual test process, and solves the reflection residual problem that cannot be eliminated by static matching.

This invention discloses an online tool health monitoring method and system based on workpiece machining quality data fusion. First, the deviation between the actual groove width and the standard groove width of the workpiece is calculated, constructing a time-series sequence of groove width deviation. Then, a sliding time window mechanism is used to dynamically extract multi-dimensional time-series features, including mean, standard deviation, slope, out-of-tolerance count, and previous values, comprehensively characterizing the nonlinear evolution trend during tool wear. Further, the fused feature vector can be input into a random forest regression model to achieve high-precision prediction of the current groove width. Finally, based on the deviation between the predicted groove width and the standard groove width, the tool degradation state is evaluated in real time. This invention directly utilizes groove width data from the machining process, eliminating the need for additional sensor deployment. It offers advantages such as low cost, high reliability, and strong real-time performance, reducing scrap rate and improving tool life without increasing hardware investment.

The present disclosure provides an optical information processing system and a training method and device, and relates to the fields of optical computing, artificial intelligence and optoelectronic hybrid computing. The system can include an encoder, a mode selection transmission module and a decoder; the encoder is configured to modulate an input light field to generate a continuous complex light field; the mode selection transmission module is configured to perform discrete mapping of the continuous complex light field in a low-dimensional latent space, and obtain an output light field through nonlinear evolution; and the decoder is configured to generate a task processing result corresponding to the input light field according to the output light field. The scheme disclosed in the present disclosure can improve the nonlinear representation capability of the system.

The present application relates to the technical field of general control system of nonlinear feedback and precision process control, disclose a kind of construction and accurate water distribution execution system of soil moisture nonlinear response model, including characteristic perception module, infiltration damping observation module, pulse width modulation module and actuator module, module calculates the derivative term of porous medium controlled object characteristic parameter, extracts the instantaneous slope of trajectory vector in infiltration dynamic phase plane, and utilizes determination logic to trigger duty cycle constraint when second derivative term is negative and instantaneous slope absolute value exceeds threshold, and the medium flow rate is adjusted by actuator, the present application utilizes trajectory curvature feature to characterize the nonlinear evolution of internal damping of controlled object, realizes the pre-judgment to infiltration stall point, eliminates the physical phase lag of control loop, suppresses the adjustment overshoot, and ensures the operation of system under complex damping working condition.

The application discloses a kind of vertical deformation coordination control and pre-offset construction method and system of super-high special-shaped steel bridge tower, method includes: according to the time-varying deformation difference of steel and concrete, construct vertical difference deformation analysis model, calculate the vertical deformation difference of bridge tower full cycle, provide basis for pre-offset calculation;Combined with deformation difference, establish the pre-offset calculation system with the minimum deformation difference as the goal, and the improved nonlinear iterative calculation method is used to determine the construction reverse pre-offset amount in reverse direction.The iterative method takes the minimum sum of deformation difference as the goal, combines time effect and deformation data, adjusts step length and updates pre-offset amount to meet convergence requirements, while dividing stages according to construction nodes, establishing adaptive fitting function to ensure connection, reflecting nonlinear evolution of deformation.Pre-offset amount is segmented and constructed synchronously, and the posture is calibrated, linear is monitored in real time, and the difference deformation is offset by pre-offset deformation.

The invention discloses a beam damage identification method based on a Koopman auto-encoder neural network, and the method comprises the steps: collecting the dynamic response data of a structure under the impact load effect based on the finite element simulation analysis of a cantilever beam; then, the collected data is preprocessed, and a data set used for training a Koopman auto-encoder neural network is constructed; training a Koopman auto-encoder neural network by using the data set so as to learn a nonlinear evolution law of the beam structure power system and obtain a linear lifting form of the beam structure power system; and based on the trained network, calculating Koopman modals of a health state and a to-be-detected state, and constructing a damage index by comparing the difference between the health state and the to-be-detected state, thereby realizing accurate identification and positioning of the beam structure damage. According to the method, the characteristic that a nonlinear system is globally linearized according to the Koopman theory is utilized, the defect that a traditional method is sensitive to structural nonlinearity and environmental noise is overcome, and the method has the advantages of being high in recognition precision, high in anti-noise capacity and the like and is suitable for beam structure health monitoring under complex working conditions.

The invention relates to the technical field of data processing, in particular to a PM2.5 data calibration method and system for a pollution peak value. According to the method, the logic of the dynamic interval is constructed based on model operation fine particulate matter fluctuation characteristics, so that a nonlinear evolution rule can be captured and static threshold limitation can be eliminated, and core geometric constraint conditions are effectively fused in a mode of constructing a planar topology network for suspected peak values and measuring and calculating spatial diffusion gradient values; a directional flux value is derived based on a triangular interior angle sine constant, a dispersion scalar is generated to participate in a mechanism of reference comparison, pure concentration comparison is converted into a conveying strength quantitative evaluation process with meteorological physical support, and a real extreme value pushed by meteorological external causes and false burrs caused by equipment faults are accurately distinguished. And replacement and retention operations are executed in a targeted manner according to the discrimination result, the technical barrier that the authenticity of the peak value is easily confused by the conventional means is overcome, and the physical real attribute and traceability analysis precision of the data calibration sequence are practically guaranteed.

The invention discloses a large equipment hoisting risk intelligent sensing system, and belongs to the technical field of hoisting risk identification. The method is used for solving the technical problem that an overload or out-of-control risk cannot be early warned in time due to the fact that online state monitoring of key devices such as a torque limiter and an anti-falling safety device is not integrated in an existing scheme. Through fusion of a fiber bragg grating, a millimeter wave radar and a torque sensor, full-scale monitoring from material damage to overall dynamic behaviors is achieved, an improved Transform model is combined with a space-time self-attention mechanism, dynamic updating of the eccentric distance is implemented based on torque-strain combined calculation, and the real-time control requirement of hoisting operation is met; according to the method, a three-dimensional risk tensor fusing structure defect-load offset-environment interference is constructed, the three-dimensional risk tensor is converted into a Lorenz system initial condition through state mapping, a future trajectory is predicted through a pre-training model, and end-to-end modeling of dynamic risk and nonlinear evolution is realized.

The invention discloses a spanning frame structure size optimization method and system based on a firefly algorithm, and relates to the technical field of electric power engineering construction, and the method comprises the steps: firstly building a parameterized finite element model, implementing normalized hierarchical loading, obtaining a tangent stiffness matrix evolved along with a load, calculating the modal participation degree of a component, and constructing a dynamic participation degree track; and according to the quantitative trajectory smoothing index, the modal dominance and the key component suppression index, constructing a brightness function with coupled rigidity and trajectory features. The firefly attraction degree is corrected by means of trajectory similarity based on modal energy fingerprints, and nonlinear search is executed through a characteristic resonance mechanism. The method solves the problem that traditional static optimization neglects geometric nonlinear evolution to cause modal abrupt change, and the dynamic stability and evolution robustness of the spanning frame in the full stress process are remarkably improved while the structural light weight is achieved.

The application discloses a small Reynolds number liquid film flow simulation method, which is based on a flow physical model, derives normal stress and shear stress balance conditions and motion boundary conditions of a liquid film surface under the action of gravity, airflow shear force and evaporation effect, and carries out non-dimensionalization and simplification on wall surface non-slip boundary conditions, flow mass equation and momentum equation and liquid film surface boundary conditions; aiming at small Reynolds number flow, the method is based on long wave theory to derive a differential equation of liquid film thickness evolution with time and space under the synergistic influence of gravity, airflow shear force and evaporation effect; the differential equation is subjected to weak nonlinear evolution analysis of wave, the change relationship of wave instability growth rate with gravity, airflow shear force and evaporation effect is obtained, and a change image of the influence of airflow shear force and evaporation effect on the liquid film wave instability growth rate is given.

The application provides a multi-element data acquisition and fusion analysis method and system for structural health monitoring, comprising: acquiring target building or structure monitoring data by using an integrated multi-element sensor and transmitting; processing the monitoring data; establishing an LSTM neural network modal parameter prediction model and an LSTM neural network correction inclination prediction model and training; determining a modal parameter sequence and a correction inclination sequence for prediction, inputting the modal parameter sequence and the correction inclination sequence into the LSTM neural network modal parameter prediction model and the LSTM neural network correction inclination prediction model respectively, obtaining a modal parameter prediction sequence and an inclination prediction sequence, and performing trend analysis and early warning. The application effectively eliminates the influence of environmental parameters on the measured structure parameters and the prediction results, and the LSTM neural network is used to train and fit the measured data of buildings or structures under different conditions, so that the nonlinear evolution characteristics of the measured structure time series data can be expressed with high precision.

The application discloses a controlled Lengyel-Epstein brain image classification method, and belongs to the technical field of brain tumor auxiliary diagnosis and medical image processing. The method comprises the following steps: firstly, acquiring a brain MRI image to be identified, and extracting a nonlinear index of a pretreated image to map to a high-dimensional calculation region; then, constructing a bivariate feature evolution field, acquiring a nonlinear evolution instruction set for an image classification task of a brain MRI data set, and applying the nonlinear evolution instruction set to the calculation region; monitoring a time evolution residual error of a feature vector potential energy component in the bivariate feature evolution field in real time, extracting a final stable polarity distribution of the feature vector potential energy component as a processing result, and realizing brain MRI image classification. The application solves the technical problem that a traditional reaction diffusion model cannot converge to a stable decision surface within a limited step, improves the brain MRI image classification accuracy, and has high robustness and interpretability.

The invention relates to the technical field of medical image processing, in particular to an ultrasonic image processing system based on image enhancement, which comprises a field construction step: constructing a pixel-level structure tensor matrix, and mapping original ultrasonic image data into tensor field data; a flow direction locking step: performing feature space decomposition on the tensor field data to generate a diffusion guide graph containing a main feature vector; a nonlinear evolution step: constructing an anisotropic diffusion tensor based on characteristic value information, and executing time step iterative solution based on a partial differential equation; a reunion output step: performing boundary continuity reunion on the enhanced image data; according to the method, the direction of the tissue is accurately locked through the Riemannian metric tensor field, smooth texture smoothing and inverse texture sharpening are achieved, speckle noise is greatly restrained, and meanwhile the microstructure features are reserved.

The application relates to the field of intelligent construction and man-machine interaction technology, in particular to an intelligent construction site auxiliary decision system based on augmented reality; the system comprises the following: a data acquisition module: through a wearable sensing device and augmented reality glasses, real-time acquisition of physiological and behavioral data of operators; a first processing module: generation of a cooperation trust index of a quantitative man-machine cooperation state based on the acquired data; a second processing module: combination of the cooperation trust index and a preset basic risk, evaluation and output of a current overall failure risk through a nonlinear evolution model; an adaptive decision module: determination and output of a final decision scheme according to the overall failure risk, combination of efficiency priority and resilience priority two utility functions; the application converts the physiological and behavioral data of the operators into quantitative state indexes by introducing the cooperation trust index, solves the problems that the man-machine cooperation state is fuzzy and difficult to accurately measure in the prior art, and provides a data basis for risk assessment and adaptive decision.

The invention provides an electroencephalogram micro-state recognition and sleep-aiding guiding method for a sleep transition period, and relates to the technical field of intelligent sleep aiding. The method comprises the following steps: synchronously acquiring environmental disturbance and electroencephalogram physiological signals in a sleep transition period, calculating two interference indexes of noise-electroencephalogram coupling and illumination-melatonin regulation, and carrying out normalized coupling pretreatment; a stable environment interference coupling strength index is obtained through iterative feedback, a nonlinear evolution function is matched according to interference grade branches, and a dimensionless sleep state index is generated; and dynamically adjusting an electroencephalogram micro-state matching threshold value, adapting to sleep-aiding guide parameters such as audio and light, and circularly checking the effect until the standard is reached. According to the method, the problems of inaccurate electroencephalogram micro-state recognition and poor sleep-aiding guide adaptation under bedroom low-frequency noise and illumination coupling interference are solved, the recognition precision and the sleep-aiding adaptation are improved, the sleep transition period is shortened, and the method is suitable for indoor complex interference sleep scenes.

The invention discloses a grape wine flavor nonlinear evolution characterization method and system based on sensory data, and relates to the technical field of grape wine and fruit wine flavor chemistry, and the method comprises the following steps: collecting sensory characteristic quantitative data of a plurality of years of grape wine samples, analyzing a flavor characteristic dynamic weight by adopting a principal component analysis (PCA), establishing a nonlinear evolution model in combination with Gaussian process regression; and dynamically characterizing the sensory quality of the wine in different storage periods through the nonlinear evolution model. According to the method, through combined innovation of dynamic weight calculation and Gaussian process regression, visual characterization of the sensory quality along with wine age evolution is achieved, and the method has important significance on wine process technology optimization, storage period management and differentiated market promotion strategies.

The application discloses a forestry remote sensing image generation method and system based on multi-modal data, and relates to the technical field of computer vision.The steps of the method include: synchronously acquiring multi-modal data of a target forest area to generate a plurality of compressed encoding arrays; constructing a forestry base matrix through spatiotemporal spectral decoding reconstruction, constructing a compensation weight, constructing a nonlinear evolution surface based on meteorological factors, and generating a first remote sensing image; adjusting a spectral band range, generating a forestry fusion matrix through a deep learning network; performing clustering analysis, determining a decoding order of spatiotemporal spectral reconstruction, dynamically allocating corresponding weight coefficients, combining a three-band feature map, and inputting the three-band feature map into a preset deep learning network again to generate a second remote sensing image.The application realizes low-cost migration from an RGB image domain to a three-band image domain, models a remote sensing image in a data-driven manner, and improves spectral and spatial texture adaptability.

The invention discloses a method for predicting optical soliton time domain evolution in an optical fiber based on an improved width learning system. The method comprises the following steps: constructing a width learning system, wherein the width learning system adopts a flattened network structure to realize structure horizontal expansion of feature nodes, enhanced nodes and the like; improving a width learning system, wherein the improvement comprises improving an incremental learning algorithm action range to synchronously adjust multiple layers of nodes so as to enable the system to learn more data features; the training system is suitable for optical soliton prediction. According to the method, the network structure can be dynamically updated in the training process, complex retraining is avoided, and the training efficiency and the prediction accuracy are effectively improved. In addition, the method utilizes a ridge regression algorithm to quickly solve pseudo-inverse from a hidden layer to an output layer, so that the training time is remarkably shortened, and quick and accurate prediction is realized. The method can be applied to rapid prediction of other optical nonlinear evolution processes, and a new thought is provided for application of the flattening network in physical field modeling and the optical field.

The present application relates to the technical field of data processing, in particular to a PM2.5 data calibration method and system for pollution peak value. In the present application, the logic of constructing a dynamic interval based on model operation fine particulate matter fluctuation characteristics can capture the nonlinear evolution law and thus get rid of the static threshold limit, the way of constructing a plane topological network for suspected peak value and measuring the spatial diffusion gradient value effectively fuses the core geometric constraint condition, the mechanism of deriving the directional flux value relying on the sine constant of the internal angle of a triangle and generating the dispersity scalar to participate in the benchmark comparison, the process of converting the simple concentration comparison into the quantitative evaluation process of the transport intensity with meteorological physical support, the accurate distinction between the real extreme value driven by the meteorological external cause and the false burr caused by the equipment failure, the targeted replacement and retention operation according to the identification result, the breakthrough of the technical barrier that the conventional means is easy to confuse the true and false of the peak value, and the physical real property and the traceability analysis accuracy of the data calibration sequence are effectively ensured.