389 results about "Linear fitting" patented technology

The invention relates to the technical field of 3D printing error compensation and precision control, and particularly discloses a printing comprehensive precision measurement method for 3D printing equipment, which comprises the following steps: designing a standard test piece containing various typical geometrical characteristics, and combining non-contact three-dimensional optical scanning and contact probe measurement means to measure the printing comprehensive precision of the 3D printing equipment. Spatial distribution data of printing errors are obtained in a controlled environment, a three-dimensional error field model with direction and amplitude information is constructed through point cloud registration and feature comparison, and based on error distribution features, the local resolution is dynamically adjusted by adopting a multi-level self-adaptive grid division strategy, so that a three-dimensional error field model with direction and amplitude information is constructed. A local error compensation model is established by combining tightly-supported radial basis function interpolation and a linear fitting method, a boundary consistency matching algorithm is introduced to realize multi-level model fusion, a unified global error compensation field is formed, an error offset vector field is embedded into a slicing process, point-by-point dynamic correction of a nozzle path is realized by using a trilinear interpolation algorithm, and the nozzle path is dynamically corrected. And the forming precision and the surface quality of the key structure are improved.

The invention provides a production line equipment fault prediction method and system, and relates to the technical field of fault prediction.The method comprises the steps that multi-source time sequence signals are synchronously collected, a multi-dimensional health feature sequence is extracted, the comprehensive sensing capacity covering multiple states of equipment is established, and further, the fault prediction efficiency is improved. By selecting a key health feature sequence for each device and performing linear fitting to quantify the performance degradation severity, accurate description and early recognition of the performance degradation trend of the device are realized, and in addition, the causal influence intensity between the devices is quantified by adopting a transfer entropy algorithm, so that the accuracy of the performance degradation trend is improved. Scientific modeling and visual analysis of a fault propagation path and a linkage effect in a production line are realized, and a comprehensive risk index is generated through weighted fusion of performance degradation severity and a global impact factor, so that a comprehensive risk assessment index is formed. Based on the cumulative failure probability function, the comprehensive risk index is converted into a fault probability curve changing along with time, a visual mathematical expression of the risk trend is formed, and a scientific basis is provided for decision making.

The invention discloses a pulse eddy current signal late period slope feature extraction method and system based on segmentation fitting, and belongs to the technical field of nondestructive testing. The method comprises the following steps: performing logarithmic transformation and filtering processing on a to-be-detected eddy current signal; intercepting a characteristic signal segment based on the filtering signal; dividing the feature signal segment into a plurality of sub-segments, performing piecewise linear fitting, determining a linear segment, and extracting a slope as a feature value; repeating the steps to obtain characteristic values of the test pieces with different thicknesses, performing curve fitting according to the characteristic values of the test pieces with different thicknesses and the corresponding thicknesses, and calculating a fitting error; and finally, iteratively optimizing the division number of the sub-segments by taking the minimum fitting error as a target. Through adaptive filtering, segmentation fitting and an intelligent optimization algorithm, the anti-interference performance, the linearity and the measurement precision of pulse eddy current signal feature extraction are remarkably improved, and efficient and reliable thickness detection of the high-temperature and high-pressure industrial pipeline under the condition that a heat preservation layer does not need to be disassembled is achieved.

The invention discloses a high-energy-efficiency logarithm floating point multiplier based on truncation error compensation. The system comprises a compensation parameter modeling unit and a floating point multiplier module. The compensation parameter modeling unit obtains an optimal compensation value k through linear fitting and parameter optimization, and realizes regional self-adaptive compensation in combination with block modeling so as to reduce truncation errors; the floating point multiplier module comprises a special value detection module, a symbol exclusive or module, an exponential addition module and a mantissa truncation compensation module. In a logarithm domain, multiplication is replaced by addition, and compensation logic is introduced to correct errors. The multiplier supports various floating point precisions and configurable truncation bit widths, provides three representative structures to give consideration to both the precision and the energy efficiency, and is suitable for high-energy-efficiency calculation scenes such as neural networks and image processing.

The invention discloses a method and system for detecting abnormal welding spots in the spot welding process. The method comprises the steps that a time sequence data sequence collected in real time in the welding process is obtained; calculating a first-order difference sequence of the time sequence data sequence; screening a target data segment representing a stable trend based on the first-order difference sequence; performing linear fitting on the first K data points of the target data segment to obtain a first trend direction; performing linear fitting on the last K data points of the target data segment to obtain a second trend direction; calculating the variable quantity of the first trend direction and the second trend direction; and when the variable quantity exceeds a preset threshold value, it is judged that welding is abnormal, and an alarm is triggered. The stable trend section is screened through the differential sequence, noise interference is effectively eliminated, and it is ensured that the quality criterion is based on reliable data. And whether the welding spot is abnormal or not is quickly and accurately judged, so that real-time and non-destructive spot welding quality on-line monitoring is realized.

The invention relates to a hitch angle estimation device, a hitch angle estimation method, and a program product. A hitch angle estimation device according to the present disclosure performs conversion from an observation coordinate system to a world coordinate system on a point sequence representing a tow bar and a lower end portion of a trailer in an image captured by a camera mounted on a vehicle that tows the trailer via the tow bar. In the present invention, linear fitting of a point row representing a tow bar and a lower end of a trailer after the execution of the transformation is performed is performed, and a hitch angle of the trailer in a world coordinate system is calculated on the basis of a straight line representing the tow bar in the world coordinate system and a straight line representing the lower end of the trailer in the world coordinate system obtained by the linear fitting. And executing Kalman filtering processing on the hitching angle of the trailer in the world coordinate system.

The invention discloses a disastrous big wave occurrence probability and wave height combined prediction method, device and equipment, a medium and a product, and relates to the technical field of meteorological ocean forecasting. The method comprises the following steps: acquiring an effective wave height measured value and environmental factor data of a to-be-predicted region in a historical time period; key environment factors are screened through spatial correlation analysis and linear fitting verification, a first training data set and a second training data set are constructed, and a multiple linear regression model and a logic regression model are trained respectively to obtain a wave height prediction model and a big wave occurrence probability prediction model; and inputting the key environmental factors of the current time period of the to-be-predicted region into the model to obtain the predicted occurrence probability of the disastrous waves and the predicted value of the significant wave height in the predicted time period. According to the method, the key environmental factors are screened and the two models are used for respective prediction, the influence of the various environmental factors on the big waves is considered, the occurrence probability and the wave height of the disastrous big waves can be predicted more accurately, and the ocean safety is effectively guaranteed.

The invention provides an IGBT (Insulated Gate Bipolar Translator) turn-off time measuring method, system, equipment and medium, and belongs to the field of power semiconductor device testing technology and state monitoring, and the method comprises the following steps: firstly, selecting 20% Vdc and 80% Vdc as turn-off initial measuring points, secondly, carrying out first error correction by adopting a linear estimation method, and finally, carrying out second error correction by adopting a linear estimation method; performing linear fitting on the voltage difference Vce data through a least square method to obtain initial turn-off time; and finally, aiming at systematic deviation caused by inherent concavity and convexity characteristics of the voltage curve, performing secondary error correction by adopting a convex hull algorithm, acquiring a lower envelope line of the ic-tOFF relation curve through an Andriw algorithm, and identifying and eliminating the problem of Overestimation of the turn-off time based on a convex hull accurate point set. According to the invention, accurate measurement of the turn-off time of the IGBT under the condition of low sampling rate is realized, the hardware cost is significantly reduced, and a reliable technical basis is provided for IGBT state monitoring and life prediction.

An intelligent monitoring and early warning method for damage failure of a filling body beside a gob-side entry retaining roadway comprises the steps that a high-water-content material standard sample is prepared, uniaxial compression testing and acoustic emission monitoring are carried out, and mechanical property data and acoustic emission waveform signals are synchronously obtained; carrying out failure stage division based on the maximum stress value and a sliding linear fitting method, and carrying out automatic classification and labeling operation; solving variational mode decomposition key parameters through an intelligent optimization algorithm to obtain decomposed intrinsic mode components; constructing a sample data set through a data enhancement process; training to obtain a waveform-based destruction stage classification model; elastic wave signals are collected in real time, variational mode decomposition and noise reduction processing are carried out, intrinsic mode components highly related to the damage degree are obtained and serve as input data, and a classification result is output through a waveform-based damage stage classification model; and performing graded early warning. According to the method, the real-time dynamic monitoring of the damage failure of the roadside filling body can be realized, and the early warning of the damage failure of the high-water material roadside filling body can be realized.

The invention provides a DEM-based reservoir return water inundation range analysis method, device and equipment and a medium, and relates to the technical field of water conservancy projects. Historical water level data of a plurality of test points in a reservoir are obtained, linear fitting is carried out according to a preset direction, a corresponding water level distribution curve is obtained, and a reservoir model corresponding to the reservoir is divided into a plurality of reservoir model sections with water level equal difference distribution according to a water level equal difference principle and a preset fixed value. Calculating a water level elevation value of a central point of each reservoir model section, constructing a corresponding digital elevation model (DEM), determining a boundary contour of a submerged area, and performing smooth connection processing to obtain a reservoir backwater submerged model, namely a reservoir backwater submerged range; the water level change can be quickly responded, the inundation prediction result can be updated in real time, the abnormal data recognition rate is increased, the inundation analysis accuracy is improved, the high-resolution DEM data is combined, meter-scale-precision inundation range delimitation is achieved, and the method is suitable for the reservoir area with the complex terrain.

The invention belongs to the technical field of metal material science, and particularly relates to an alloy fracture toughness analysis method and system based on a nanoindentation technology, and the method comprises the following steps: S1, carrying out nanoindentation continuous rigidity test on an alloy sample by adopting a Berkovich indenter to obtain an indentation load-indentation depth curve; performing a tensile property test on the alloy sample to obtain the elongation at break and the intrinsic elastic modulus of the material; s2, linear fitting parameters are determined according to the logarithmic linear relation between the effective elastic modulus and the indentation depth; s3, calculating a critical effective elastic modulus according to a continuous damage mechanics theory; S4, calculating the critical indentation energy of the unit contact area of the target alloy through a P-h curve; and S5, calculating a fracture toughness index according to continuum mechanics and a Griffith fracture theory. The method solves the problems of low efficiency and macro-micro data disjunction of a traditional method, and provides a systematic solution for fracture toughness analysis of plastic materials including but not limited to plastic metals such as zinc alloy, aluminum alloy, magnesium alloy and the like.

The invention discloses a plane alignment element segmentation and node identification method and system based on curvature characteristics, and the method comprises the steps: carrying out the smoothing of the curvature of point cloud data of a road center line through a local weighted regression algorithm, and obtaining a smooth curvature; traversing the smoothed curvature by adopting a sliding window, calculating a variable coefficient of the curvature in each window, judging the consistency of the curvature and the linear element type according to the size of the variable coefficient to obtain a preliminary candidate node, respectively extracting the curvatures of the left side and the right side of the candidate node by taking the node as a center, and carrying out least square method linear fitting to obtain a final candidate node; and obtaining an intersection point of the fitting straight lines, and determining a final node position. The method can effectively solve the problems of large curvature fluctuation of the road center line and high node identification difficulty, has the advantages of high node positioning accuracy, strong adaptability and high calculation efficiency, and can be widely applied to the fields of road digital modeling, traffic engineering design and maintenance and the like.

The invention provides a laser radar attitude compensation wind measurement method and system, and relates to the technical field of laser radar wind measurement, and the method comprises the steps: obtaining the six-degree-of-freedom motion data of a carrier; driving the two-axis stabilized platform to compensate based on the three-axis rotation angle so as to stabilize the direction of the laser beam, and scanning to obtain initial radial wind speed data; then synchronizing the translation speed with the wind speed data by using a linear fitting algorithm to obtain a synchronous data set; a laser radar scanning mode is determined, and a corresponding attitude compensation strategy is selected according to the mode; inputting the synchronous data set and the strategy into a pre-training machine learning model to generate a compensation parameter, and processing the initial data through Kalman filtering based on the parameter to obtain target radial wind speed data; and finally, performing wind field inversion calculation based on the target data to generate a wind field measurement result. According to the invention, the wind field measurement precision of the laser radar on the motion platform under the complex sea condition is improved.

The invention discloses a berthing state estimation method based on shipborne laser radar target detection, and the method comprises the steps: constructing a ship three-dimensional target detection network model through an improved PointPill coding network, and introducing an average pool layer and an attention pooling layer into the coding network to carry out the fusion of the global perception information and the local perception information of a target, and performing model training according to the constructed model loss function, detecting and confirming whether an anchored ship exists in a preset berth through an optimal detection network model, and respectively acquiring whether the anchored ship exists or does not exist in the preset berth by introducing an RANSAC plane fitting algorithm. A berthing speed used for ship berthing state estimation is obtained by combining the required berthing front edge plane or the outer contour surface of the target ship, namely a berthing berthing boundary surface and the position of the shipborne laser radar; the defect that the berthing state estimation accuracy is poor due to the fact that a berthing line model obtained through a traditional method based on linear fitting serves as the berthing state estimation reference is solved.

A design method of high-capacity hydrogen storage alloy, the specific method comprises the following steps: obtaining data, data preprocessing, data set division, establishing a prediction model, determining the optimal prediction model and alloy composition design. By comparing various machine learning algorithms, the optimal prediction model is selected and determined. Compared with the alloy composition design method of a single prediction model, the high-precision prediction of the hydrogen storage capacity of the hydrogen storage alloy can be further improved, and the efficiency is higher than that of the traditional empirical criterion method and the linear fitting method. By using Xgboost algorithm as the fitness function, the average and maximum fitness function values of the population are converged, and after the preparation of the alloy, the relative error between the actual hydrogen storage capacity and the predicted hydrogen storage capacity is as low as 0.54%, that is, the high-precision prediction is achieved.

The invention discloses a bearing performance degradation type identification method based on a vibration signal tail attenuation rate, belongs to the technical field of mechanical equipment state monitoring and fault diagnosis, and aims to improve the prediction precision of the remaining service life of a rolling bearing and improve the applicability of an intelligent prediction model by effectively identifying the bearing degradation type. The system comprises a data acquisition and preprocessing module, a tail attenuation rate calculation module, a degradation point detection module and a degradation type identification module. Firstly, feature extraction and segmentation processing are carried out on an original vibration signal of the rolling bearing; then, histogram distribution of vibration signal amplitudes is constructed, a right tail area sample is extracted, and a tail attenuation rate index representing signal extreme fluctuation attenuation behaviors is calculated through logarithmic transformation and linear fitting processing; and finally, constructing a dual judgment criterion of ratio threshold value judgment and adjacent point difference value comparison, and realizing automatic identification and classification of three typical degradation modes of bell shape, horn shape and sudden deformation.

The present disclosure relates to the technical field of model construction, and provides a river water dynamics model construction method based on multi-source remote sensing, which comprises the following steps: acquiring a river remote sensing image to obtain a water edge line, acquiring water level data by means of an altimeter satellite, and giving an elevation value to a water edge line corresponding point according to the water edge line and the water level data; for a straight river section, a value is taken according to a specific rule, and for a curved river section, the value is determined according to linear fitting of multiple elevation measuring point values on a river section; based on the elevation values of the water edge lines of the straight and curved river sections, a river terrain digital raster model is further constructed by interpolation along the water edge line, the river water dynamics model is constructed by using the river terrain digital raster model, the simulation effect of the water dynamics model is determined by comparing the water level consistency or the water surface vector form consistency, and the calibration and verification of the river water dynamics model are realized. The present disclosure realizes accurate simulation of the river water dynamics process. The present disclosure can reduce the cost of model parameter acquisition.

The invention relates to the technical field of network acceleration, and discloses an edge-end large language model reasoning acceleration method and accelerator, and the method comprises the following steps: reconstructing a calculation process of a decoding stage, and carrying out the deep fusion of a multi-head attention mechanism and the calculation operation of a feedforward network; the weight and key value data are stored in HBM, and the coefficient and the accumulated attention score are stored in DDR; for linear matrix calculation, a unified matrix calculation unit is used for executing multi-precision matrix operation; for nonlinear function calculation, a mathematical transformation and linear fitting method is adopted, a Softmax function is converted into operation with 2 as the bottom through a bottom conversion formula, and truncation and third-order linear fitting are conducted on a Sigmoid function; and constructing a key value screening algorithm based on the accumulated attention score, dynamically adjusting a key value storage position, maintaining a recent key value cache region and an important key value cache region in a limited cache space, and realizing key value efficient cache in long text reasoning.

The invention discloses a method for determining a parameter lookup table of a neural network activation function, a method and system for calculating the neural network activation function based on the parameter lookup table and a storage medium, and belongs to the technical field of neural network accelerators, and the method comprises the steps: dividing the neural network activation function into a linear region and a nonlinear interpolation region; dividing the nonlinear interpolation region into a plurality of sub interpolation regions; performing linear fitting on the multiple segments of sub-interpolation regions to obtain slopes and intercept of multiple linear functions in one-to-one correspondence with the multiple segments of sub-interpolation regions; and determining a parameter lookup table according to the input variable value range of the neural network activation function corresponding to the multiple segments of sub-interpolation regions and the corresponding relationship between the slopes and intercept of the multiple linear functions, and storing the parameter lookup table in the memory. According to the method, the activation function is divided into the linear region and the nonlinear interpolation region, the activation function of each partition is approximately calculated through the linear function, the parameter lookup table is determined, and the activation function calculation is realized.

The invention discloses a power system forced oscillation source positioning method based on ICEEMDAN-ITEO, and belongs to the technical field of power system stability analysis and control. In a signal adaptive decomposition and modal extraction part, an ICEEMDAN algorithm is utilized to decompose a wide-area measurement signal, suppress modal aliasing and noise interference and obtain a modal component with physical significance; the energy weight of each modal component is calculated by adopting ITEO, the identification capability of the dominant forced oscillation component is enhanced through an adjustable resolution parameter, and the time domain waveform of the dominant forced oscillation component is reconstructed; in the oscillation source judgment part based on the dissipation mode energy flow, the dissipation mode energy flow of each generator port is calculated according to the dominant oscillation mode component, and accurate judgment of a disturbance source is realized through linear fitting of slope characteristics of the dissipation mode energy flow; in the verification and application part of the oscillation source positioning result, the accuracy and robustness of the method in different noise and disturbance scenes are verified through a test system comprising a network-forming fan.

The invention belongs to the technical field of municipal pipeline detection, and provides a municipal pipeline detection method and system based on nondestructive detection and artificial intelligence, and the method comprises the steps: carrying out the AI reconstruction through the pipe diameter point cloud data of different pipe sections, generating a continuous profile map, and recognizing the pipe sections with reduced pipe diameters; generating a slope curve by using the slope detection data to identify a slope change pipe section; calculating the overlapping mileage proportion of the pipe diameter reduction pipe section and the gradient change pipe section, performing linear fitting on the gradient curve of the overlapping pipe section, and judging the pipe diameter misjudgment risk pipe section according to the decision coefficient; determining a coupling relationship between gradient change and pipe diameter reduction based on historical data of pipe sections of the same type, establishing a coupling model, and judging whether the actual pipe diameter reduction amount and the theoretical pipe diameter reduction amount of the pipe section with the pipe diameter misjudgment risk conform to the coupling relationship or not; if not, the diameter of the oblique section is converted into the real inner diameter according to the gradient inclination angle by combining the RANSAC algorithm, the real inner diameter is compared with the designed inner diameter again to determine whether the pipe diameter is reduced or not, and it is ensured that the pipe diameter detection result is accurate.

The invention discloses a rock anisotropy index determination method based on indoor drilling and multi-parameter fusion, and relates to the technical field of geotechnical engineering. The method comprises the following steps: S1, carrying out digital drilling experiments on a rock sample in different directions to obtain drilling parameters of the rock in different directions; s2, feed force and torque force are calculated according to the drilling parameters, and the ratio of the feed force and the torque force is calibrated based on a linear fitting equation of the feed force and the torque force; s3, calculating internal friction angles of the rock in different drilling directions on the basis of the force balance analysis model; s4, calculating unit grinding energy of the rock in different drilling directions; and S5, calculating the comprehensive anisotropy index of the rock based on the variable coefficient. The rock anisotropy index can effectively quantify the directivity difference of the rock, can directly serve rock mass stability analysis, drilling scheme optimization and support design decision of an engineering site, and effectively connects indoor test and field engineering application.

The invention relates to the field of industrial automatic detection, in particular to an industrial component point cloud identification method and system. The method comprises the following steps: collecting multi-frame point cloud data, extracting a mold table region point set and fitting a reference plane; identifying an effective point set based on the reference plane, establishing a local coordinate system and calculating a sensor rotation matrix; generating a correction point cloud through the rotation matrix calibration point cloud; dividing the corrected point cloud into grids, and detecting and expanding a contour point set; performing straight line fitting on the contour point set to determine boundary lines and angular points; comparing the angular point forming region with a preset feature to eliminate anomalies; the connecting corner points form a closed contour. The method solves the problems that an existing point cloud processing technology depends on a static environment and is weak in dynamic interference resistance, contour extraction is greatly influenced by noise, and key point connection is prone to errors.

The invention discloses a shield tunneling machine soil pressure balance control method and system based on multi-parameter dynamic optimization, and relates to the technical field of shield tunneling machine control. The method comprises the following steps: calculating a support force required in a shield tunneling process under a current working condition and a theoretical shield soil bin pressure observation value, and establishing a relationship through a linear fitting mode to obtain a soil bin pressure transfer coefficient under a current stratum condition; calculating the theoretical support force of the tunnel face in the tunneling process, and calculating the preset value of the shield soil bin pressure in combination with the soil bin pressure transfer coefficient; and acquiring a real-time shield soil bin pressure observation value, and performing multi-parameter collaborative optimization on an error between the preset value of the shield soil bin pressure and the real-time shield soil bin pressure observation value by using a particle swarm optimization algorithm to obtain an optimized control parameter. According to the method, the pressure of the soil bin can be adjusted more accurately according to actual conditions, so that dynamic balance of an excavation face in the construction process is ensured, and safety and stability of underground tunnel construction are guaranteed.

The invention discloses a multi-region linear fitting error compensation method and system for an air pressure sensor. The method comprises the steps that original data of the air pressure sensor are acquired and preprocessed; identifying the preprocessed data based on an inflection point detection algorithm of a second derivative, and dividing intervals according to identified inflection points; performing linear fitting in each divided interval by applying a weighted least square method, calculating fitting parameters and goodness of fit of each interval, removing abnormal points in the interval, and performing fitting again; judging whether the adjacent intervals meet a merging condition or not, and performing merging and fitting optimization on the adjacent intervals meeting the condition; generating a calibration parameter table, outputting an SVG fitting curve graph according to the calibration parameter table, and correcting an error of a measured value of the air pressure sensor; by analyzing distribution characteristics of error data, intervals with similar error characteristics are divided in a self-adaptive mode, a linear fitting model is independently constructed for each interval, and averaging errors of a single model on a non-linear area are avoided.

The invention discloses a piglet nest average weight estimation method based on instance segmentation. The method comprises the following steps: S1, data acquisition; s2, data preprocessing; s3, piglet target detection; carrying out piglet target detection; s4, constructing an instance segmentation model ODSeGAN, and outputting an instance segmentation mask of each piglet; s5, counting the number of pixels in the mask area as an input feature of the weight estimation model; constructing a linear regression model by adopting a least square method so as to realize optimal mapping between the pixel area and the body weight; s6, nest average weight calculation and multi-frame fusion estimation: calculating a piglet pixel area average value based on a segmentation result in S4; in the estimation stage, the system counts the number of all segmented piglet instances in the column in each frame of image, and averages the estimated weights of the piglet instances to obtain the nest average weight estimated value of the current frame. According to the method, the piglets can be accurately segmented under a complex shielding condition, and then the average weight of the nest is estimated by combining pixel regression and a linear fitting technology.

The present application relates to a kind of GH4706 alloy long time heat exposure organizational performance evaluation method, belong to high-temperature alloy performance evaluation technical field, solve the problem of not comprehensive of the performance evaluation of alloy by existing method.The method includes selecting alloy forge piece to carry out long term heat exposure test, obtains its organizational performance data and / or organizational parameter under different temperature and different heat exposure time, forms measured data set;Establish the relationship p=T(lgt+C) / 1000 between time-temperature uniform parameter and heat exposure time and heat exposure temperature, formula is substituted into by experimental data and different C value, calculate p;Through linear fitting, establish the preliminary fitting model between p and organizational performance data and / or organizational parameter;Determine the best C value;Establish the final fitting model between p and organizational performance data and / or organizational parameter;The equivalence of model is verified using p value, and performance is evaluated using final fitting model.The method of the present application is comprehensive to the performance evaluation of alloy.

The application relates to a calculation method for realizing a Tanh excitation function based on FPGA, which comprises the following steps: obtaining the quantization bit number during FPGA realization, determining the decimal bit quantization bit number, determining the accuracy of input and output values of the Tanh excitation function and the range of the output values based on the decimal bit quantization bit number, processing the values greater than 0 based on four linear fittings, one quadratic fitting and two table storage respectively, and mapping and processing the values less than 0 based on the odd symmetry of the Tanh function. The application can realize the Tanh function with high accuracy under the condition that the FPGA resource is limited.

This invention provides a method for identifying noise intervals in pulsed eddy current (PDC) detection signals, belonging to the field of nondestructive testing (NDT) technology. It can at least partially solve the problems of reduced signal-to-noise ratio and the impact of noise data on the accuracy of quantitative analysis in the later stages of PDC detection. Based on the magnetic field attenuation rate signal acquired during PDC detection, the method performs a logarithmic transformation on the signal amplitude, then performs inverse linear fitting starting from the last sampling time corresponding to the signal, calculating the slope sequence point by point, and applying multi-level mean filtering to the slope sequence. By determining the time position at which the smoothed slope sequence first reaches or crosses a preset threshold, the starting position of the invalid signal interval, dominated by noise, is determined. Using the identification results, invalid signal intervals can be removed from subsequent data processing, thereby improving the stability and reliability of PDC detection in the detection of clad metal structures.

The present application belongs to the technical field of high-pressure grouting process optimization, and particularly relates to a high-pressure grouting process optimization method for roadway water inrush cracks. The method first improves the Retinex and dark channel priori combination for image preprocessing; based on the attention-enhanced U-Net, the crack is accurately segmented, and the length, width and depth are quantified; the distributed water pressure sensor is arranged in the crack influence area, the Gaussian smoothing, sliding window linear fitting and correlation analysis are used to remove the abnormal measuring points and calculate the weighted average water pressure; taking the crack geometry and effective average water pressure as the input, a double-branch prediction model is constructed to output the preliminary range of slurry-water ratio, grouting pressure and segmented grouting length; further, a neural network prediction model is constructed from the similar working conditions selected from historical samples, and the improved particle swarm optimization is combined to repair the longitudinal wave velocity of the post-rock mass close to the normal value as the target, and the three types of process parameters are optimized to obtain the optimal grouting scheme.