66 results about "Nonlinear methods" patented technology

Signal “peak” and “valley” removal properties of mathematical morphology operators are exploited in a method and apparatus for detecting, removing, or improving fidelity of pacemaker signal components of the Electrocardiogram (ECG) at sampling rates well below the pacemaker signal Nyquist rate. The method works for the wide variability in pacemaker signal amplitude, width, firing frequency, and other characteristics encountered in practice, and it works for malfunctioning pacemakers that may fire at any point relative to the QRS complex of the ECG signal. Filtering operations require minimal digital storage and are computationally inexpensive (no multiplications), mainly involving “maximum” and “minimum” type signal detections over a finite time history of the input signal. This implies that the method can be inexpensively implemented on small instruments in either hardware or software. The method may also be used to estimate the height and polarity of the pacemaker voltage spike, even though the base sampling period may be longer than the width of the spike.

Provided is a standard 12-lead ecg signal reconstruction method. Limb lead signals are reconstructed through formula computing, chest lead signals are reconstructed through a nonlinear method of a BP neural network optimized based on a genetic algorithm; according to the nonlinear method, firstly, the optimal initial weight value and threshold value solution space of the BP neural network is looked for through a genetic algorithm, and serves initial setting of the BP neural network; next, the lead signals used for reconstruction serve as input of the BP network, reconstruction target lead serves as output for training, a known lead set is input into a trained network, and an unknown lead construction result can be obtained. The standard 12-lead ecg signal reconstruction method is high in accuracy.

The invention uses genetic algorithms to realize dynamic channel allocation (DMA). By contrast, a nonlinear method is more suitable for nonlinear characteristics of the reality and for parallel arithmetic, can be flexibly transferred to fit various wireless channel allocation technologies; the invention is additionally characterized that channel performance is collectively estimated and channel resources are equally distributed so as to comparatively improve the comprehensive utilization ratio of the channel resource. The invention adopts carrier frequency, time slot, spread spectrum codes and the angle of wave beam as genomes to realize chromosome coding; the operating procedure is that a sliding window is used for filtering a species group to form a vector X which equals (CF3, C, F2, CF1, CF0, TS2, TS1, TS0, SF3 ,SF2, SF1, SF0, BS2, BS1, BS0), and then the genetic algorithms, such as hybridization, variation, etc., are selected to realize the dynamic allocation and optimization of the channel.

The invention discloses a coal seam floor water-inrush risk evaluation method. The evaluation method comprises a first step of building fuzzy subjection functions of each influencing factor, a second step of applying a step-by-step discrimination method to distinguish influencing capabilities of the influencing factors, a third step of building the coal seam floor water-inrush risk evaluation method on the basis of a nonlinear discriminant analysis model and on the basis of a reproducing kernel theory and a discriminant analysis theory according to the coal seam floor water-inrush influencing factors screened out in the second step, and a fourth step of performing water-inrush risk evaluation on a working face with the floor water-inrush risk by combination of an intuitive discrimination method according to the built model. The influencing capabilities of the coal seam floor water-inrush influencing factors can be analyzed objectively, the influencing factors obvious in influence are effectively screened out, interference to evaluation results caused by the factors which are weak in influencing capabilities is reduced, the nonlinear discriminant analysis model is built on the basis of existing water-inrush sample data, and the nonlinear method is scientific and practical in comparison with previous methods.

The invention relates to a digital automatic gain control method for a narrow-band signal after broadband system channelization. The method comprises the following steps: calculating average power of a current data frame and a reference power difference, and performing stepped correction and adjustment on a power gain coefficient according to the difference; and converting the adjusted power gain coefficient into an amplitude gain value, and performing multiplication according to the gain coefficient to the last frame of data to obtain a final signal, and outputting the signal. The method provided by the invention is simple and easy to implement, the power difference is calculated, and the stepped correction is performed on the power gain coefficient by using a quasi parabolic nonlinear method; and the power gain coefficient is converted into the gain value, and the multiplication is carried out through the gain coefficient to the last frame of data, thereby effectively solving the quantization error problem in the conversion of a floating point into a fixed point and improving the hearing judging effect of the signal, and the digital automatic gain control method has the advantages of easy implementation, high reliability and high practical value.

The invention belongs to the field of spacecraft orbit dynamics and control. The invention discloses a spatial non-cooperative target space-based search method based on nonlinear deviation evolution,which comprises the following steps: 1) according to ground initial orbit determination data, obtaining the number of orbits of a space-based observation platform, and an initial relative state mean value and a covariance matrix of a spatial non-cooperative target relative to the space-based observation platform; 2) calculating a state transition tensor of relative state prediction and transitionto an angle measurement space by considering a second-order nonlinear term, and splicing the two sections of transition tensors by adopting a tensor product; 3) nonlinearly converting the initial relative state mean value and the covariance matrix into a mean value and a covariance matrix of the angle measurement quantity; 4) adopting a spiral moment camera to point to search strategy, searching atarget until the target is captured or the task is ended. According to the method, a nonlinear method is adopted to analyze and calculate the mean value and the covariance matrix of the angle measurement quantity, the dynamic change of the target position in the searching process is considered, the calculation precision and efficiency are high, and the applicability to actual tasks is good.

The invention discloses a multi-source sensor joint calibration method based on a three-dimensional regular octagonal structure. The method comprises the steps of calibrating internal parameters of acamera; constructing a three-dimensional laser point cloud model; correcting the three-dimensional laser point cloud based on inertial data; optimizing a laser jitter error superposition mean value; optimizing the laser stepping angle error increment; and constructing a multi-line-segment joint external parameter calibration model, and solving an external parameter optimal solution through a nonlinear method. The method can effectively improve the joint calibration precision and the data fusion effect of the multi-source sensor in a complex agriculture and forestry environment.

The invention discloses a comprehensive evaluation and prevention method for a seepage field of an underground water seal oil storage in an operation period, and the method combines engineering geology and hydrogeological reports, and obtains hydrological survey data in a survey period. The daily water inflow of the reservoir site area is counted by combining the construction period with the actual situation, and the drilling water level change value is periodically observed and recorded. And underground water distribution of an unexcavated surrounding rock area is detected through a geological advanced prediction method, and whether a water curtain system is additionally arranged or not is judged. The pumping capacity of the water pump station and the osmotic pressure value of the oil storage cavern surrounding rock during operation are counted. A time sequence analysis method, a finite element analysis method and a nonlinear method BP neural network are combined, comprehensive analysis and evaluation are conducted on the seepage field of the underground oil depot in the operation period, the benefit of a water curtain system is evaluated, and an adjustment scheme is given.

Systems and devices for processing image or other data using non-linear methods to compensate for localized slopes are described. In one implementation, the slope of the sample values in an image or other dataset is estimated in one or more directions using a non-linear filter, such as a median filter. The values of at least some of the samples of interest are compensated using the estimated slope values to remove the effects of the slope. The compensated values may then be processed to determine if the target is present in the samples of interest, or for any other purpose.

The present invention disclosed an iterative analyzing method, which can detect the lesion in the image quickly. In brief, the iterative analysis method of the medical image disclosed by the present invention is roughly as follows: first, the original spectral image cube is expanded into a spectral image cube by a method of nonlinear dimensional-expansion, and then detecting the target's subpixel by the method of constrained energy minimization to produce an abundance image; the abundance image is fed back to the spectral image cube for create another spectral image cube by the nonlinear method. Furthermore, the abundance image is only used for detecting the subpixel of the target and does not include any spatial information, so, in order to obtain the spatial information of spectral image, it obtains the spatial information around the subpixel by using a blurring tool such as a Gaussian filter. After the spatial information is fed back to the spectral image cube, the subpixel target detection is repeatedly performed until a predetermined termination condition is satisfied.

The invention discloses a nonlinear method for computing open and close times of a breaker, which relates to the technical field of numerical processing of operating data of primary equipment in a power supply system and solves the technical problem of computing effective open and close times of a breaker. The method comprises the steps of: (1) building a parameter model of the breaker; (2) setting the operating state of the breaker into a fault state, a normally close state and a normally open state; (3) respectively building three state models according to the three operating states of the breaker; (4) matching the real-time operating state of the breaker to the corresponding state model for solving, and then carrying out weighting summation to get the effective open and close times of the breaker; (5) obtaining the current operating condition of the breaker according to the parameter model of the breaker and the effective open and close times of the breaker, which are obtained in the step (4), and making overhaul plane and overhaul time. The method can accurately and quickly compute the effective open and close times of the breaker in real time.

The embodiment of the invention relates to the technical field of communication, in particular to a secret communication method and device, which are used for more accurately constructing a driving-response system conforming to the reality. The method comprises the steps that a response system receives a superposed signal sent by a driving system, wherein the superposed signal is obtained by modulating the driving system through a random parameter fractional order nonlinear method, and the superposed signal comprises a chaotic signal and a useful signal; and the response system demodulates thesuperposed signal by using a synchronous controller and a parameter identification rule to obtain the useful signal, wherein the synchronous controller and the parameter identification rule are determined in the process of realizing synchronous confirmation of the driving system and the response system.

The invention discloses a power equipment discharge signal separation and classification method based on kernel principal component analysis. The method comprises the steps of: (1) obtaining a power equipment discharge signal characteristic quantity as a sample data set; (2) processing the sample data set by using kernel principal component analysis to obtain a principal component variable as a new characteristic quantity; (3) clustering the discharge signals through a K-means clustering method to obtain type tags of different types of discharge signals; (4) selecting a principal component variable which can best reflect different discharge signal differences, and drawing a two-dimensional scatter diagram and a three-dimensional scatter diagram; and (5) carrying out gate circling among different types of discharge signals according to the type tags to finish automatic classification of the discharge signals. According to the method, principal component characteristics are extracted byadopting a nonlinear method, an appropriate kernel function is selected to extract the principal component of the discharge signal from the original discharge pulse signal, and automatic separation and classification of multiple discharge source pulses and multiple interference source pulses are realized, so that discharge type identification of power equipment is realized.

The invention discloses a nonlinear method for describing an urban surface landscape structure. The method comprises the steps of 1, performing temperature retrieval: calculating a radiation energy value, a radiation brightness temperature and a ground temperature; 2, for a temperature difference of measure selection regions in thermal field pattern study, assuming that an image has MXN pixels, performing coverage ((M/epsilon)X(N/epsilon) square boxes are required) on a whole image region by using a group of square boxes of different dimensions epsilon, calculating a temperature difference delta T epsilon (delta T epsilon = (Tmax-Tmin)/epsilon) in each box, imagining delta T epsilon as the height of an image curve, then performing coverage by using (delta T epsilon + 1) boxes, and calculating regional measure by using a formula: mu (epsilon) = (delta T epsilon + 1)/sigma (delta T epsilon + 1); 3, enabling q to be greater than -10 and less than 10, calculating a partition function xq(epsilon), and calculating a quality index tau (q) in a double-logarithmic coordinate system; 4, calculating alpha (q) by using a formula: alpha (q) = (tau (q+1) - tau (q-1))/2; and 5, calculating f(alpha) by using a formula: f (alpha) = q alpha (q) - alpha (q).

The invention discloses a point cloud dynamic region graph convolution method, and a point cloud dynamic region graph classification method and a point cloud dynamic region graph segmentation method using the point cloud dynamic region graph convolution method. According to the method, a new convolution operation form for point clouds is adopted, point feature information of a plurality of different neighborhoods is aggregated through a nonlinear method according to a constructed point cloud picture structure, so that neurons can adaptively select the size of a region. Compared with existing technical schemes such as PointNet for analysis on a single point, the method has the advantages that a plurality of different local neighborhood graph structures are constructed, each neuron can adaptively select a proper neighborhood receptive field, and then similar convolution operation is performed by using the relation between each point and a neighborhood point to obtain local features, the surrounding neighborhood information can be better combined, the local geometric information can be more effectively extracted, and finally the classification or segmentation accuracy of the point cloud data is improved.

The invention discloses a novel composite tracking control method for a chemical intermittent process. The method comprises the following steps of establishing a discrete model of the chemical intermittent process by utilizing a local nonlinear method, and then introducing a tracking control law, defining the state error of the system and the tracking output error to obtain an expanded process control system, adopting a compensation control algorithm to design an updating rule, obtaining the gain of the controller according to a closed-loop system model and the stability condition to further obtain the updating rule and a control rule. Compared with a traditional control strategy, the novel composite tracking strategy provided by the invention has the advantages that the good tracking characteristics are achieved by adopting tracking feedback control of tracking control along the process direction, and better control performance of the system is guaranteed.

The invention provides a time-frequency interleaving preprocessing combined method for resisting narrowband interference and impulse noise in an OFDM system. The method specifically comprises the following steps of carrying out LDPC coding and corresponding constellation labeling on a signal in the OFDM system, optimizing parameters of a time domain interleaver, and fully whitening the noise; after OFDM modulation is carried out at a sending end, utilizing cyclic shift operation of a sub-matrix of a block interleaver in a frequency domain to carry out frequency domain interleaving on OFDM sub-carriers; and carrying out noise preprocessing through an amplitude limiting method of a receiving end, and nonlinearly removing the noise and the interference which are still higher than thresholds after full interleaving. According to the method, the characteristics of the interleaver and the advantages of noise processing of a nonlinear method are fully combined, the resistance performance of the OFDM system to the narrowband interference and the impulse noise can be effectively enhanced, and compared with a traditional block interleaving method, the system transmission performance is further improved.

The invention discloses an ozone generation contribution rate analysis method and device. The method comprises the following steps: acquiring observation data comprising different elements; analyzing the observation data of each element in the observation data, and respectively judging whether the observation data are in the global threshold range of the corresponding element; when the data are within the global threshold range, judging that the data are normal data; and inputting the element observation data judged as normal data into the trained TabNet for calculation to obtain the ozone generation contribution rate of the corresponding element. The device comprises an observation data acquisition unit, a data quality control unit and a calculation unit. According to the method and the device, the definition of ozone generation contribution rate is provided; VOCs observation data, meteorological observation data and conventional pollutant observation data are directly applied to ozone generation contribution analysis; an application strategy of applying the deep table learning neural network to ozone generation contribution rate calculation is proposed; a nonlinear method capable of analyzing ozone generation contribution rate of observation data at a single time is provided.

The invention discloses an intelligent high-speed traffic flow prediction method based on a continuous monitor, and belongs to the technical field of traffic flow prediction. An expressway portal frame collects real-time traffic flow data, a feature project is established for historical flow data to obtain multi-dimensional features, feature selection is performed through linear and nonlinear methods, a TCN + GBDT-based network model is established, and the real-time traffic flow data is obtained. Historical big data is used for training, and the model can predict the traffic flow at the next time interval. Compared with a traditional model driving method, a pure sequence feature-based neural network method, a single RNN-type neural network-based method and the like, the method has the advantage that the prediction precision is improved.

The invention discloses a hierarchical nonlinear subspace dictionary learning method for remote sensing image scene classification. HNSDL projects a data sample into a series of subspaces by using a hierarchical nonlinear method and realizes data dimension reduction; in order to improve the identification capability of the model, a local structure constraint term of sparse coding is introduced in the learning process. The method has the following technical progress that: (1) different from the existing linear method of a linear transformation matrix, the proposed method maps data to a series of projection spaces through a series of nonlinear functions; compared with a linear transformation matrix, the nonlinear projection can better represent the nonlinear structure of the remote sensing image. And (2) by utilizing a local structure constraint item with label information, dictionary learning not only retains a nonlinear local structure of data, but also enables intra-class samples to be compact, and the class separability of the dictionary is enhanced. And (3) projection space learning and dictionary learning are simultaneously learned in one model, and an alternative iterative optimization method is used for parameter optimization, so that the parameters of the model can simultaneously obtain optimal solutions.