30results about How to "Computing speed advantage" patented technology

The invention relates to a method for acquiring parameters of electric power harmonic waves by using Hanning window function continuous frequency spectrum interpolation, applicable to harmonic analysis and monitoring on voltage and current of a power network. Firstly, chirp Z transform (CZT) is utilized to accurately extract fundamental wave signal parameters (amplitude, frequency and phase) froman electric signal containing harmonic waves. Then the fundamental wave signal is subtracted from the analyzed electric signal and a Hanning window function is added to cut off the electric signal, and FFT (fast Fourier transformation) is utilized to calculate frequency spectrum of the residual signal. And then frequency values of all the harmonic waves are accurately calculated according to the frequency of a fundamental wave. And finally interpolation is carried out on the Hanning window function in the frequency domain according to each harmonic wave frequency, and the parameter of each electric power harmonic wave is accurately calculated. The method provided by the invention has estimated value accuracy basically equivalent to that of an addition Hanning window FFT bispectrum line interpolation fitting analysis method of electric power harmonic waves, and calculated amount of the method provided by the invention is about 1 / 2 of that of the addition Hanning window FFT bispectrum line interpolation fitting analysis method of electric power harmonic waves.

The invention discloses a method for testing the water content and distribution of cellular levels in fruit and vegetable tissues based on the Raman spectrum. The method comprises: carrying out samplepretreatment; carrying out imaging spectrum acquisition and pretreatment; carrying out Gaussian peak separation fitting on the imaging spectrum and carrying out pseudo color imaging based on a fitting result, and carrying out visualization on the water content and combination state distribution of the cellular levels. According to the invention, the visualization of the distribution of the watercontent and the combination state of the cellular levels in the fruit and vegetable tissues is completed for the first time; and relatively reliable water content quantitative analysis results of different combination states are obtained by visual imaging. The disclosed method can be used as a novel testing method for the water contents of the cellular levels in fruit and vegetable tissues. Therefore, a problem that the water changes of the cellular levels can not be detected in the existing fruit and vegetable processing process is solved; and thus the method has the broad prospects in studying the fruit and vegetable processing.

The present disclosure provides a method for optimal allocation of water resources based on a genetic algorithm. The scheme normalizes economic goals, social goals and ecological goals, so that each goal is weighted and calculated under the same dimension condition, and normalized After optimization, the weight of each objective has more practical significance; at the same time, the water resource optimization model constructed has the characteristics of multi-variable, multi-objective, and multi-constraint. The solution method has great advantages in calculation speed, accuracy, convergence, adaptability and portability.

The invention discloses a method for searching for a sliding window based on layered segmentation. According to the method, firstly, superpixel segmentation is conducted on an image; secondly, a layered segmentation algorithm of the image is run on the basis of a superpixel segmentation result of the image and a definition of region similarity; thirdly, initial windows are generated according to the layered segmentation result of the image, and the windows are screened according to the priori knowledge of a target; lastly, target candidate windows are collected densely around the remaining windows. By means of the method, on the premise that the target is not missed, only a small number of candidate target positions are generated, and the problem of a large calculated amount of feature extraction and classifier judgment in a target detection task can be effectively relieved.

The invention relates to a biomechanics modeling method for subcutaneous adipose tissues based on linear elasticity and superelasticity models. According to the biomechanics modeling method, the problems that an existing subcutaneous adipose tissue model is low in simulation accuracy and speed are solved. The biomechanics modeling method comprises the following steps: (1) establishing a geometric model, and dividing grids; (2) establishing a linear elasticity biomechanics model; (3) establishing a superelasticity biomechanics model; (4) carrying out finite element simulation on the superelasticity biomechanics model according to the step (1) and the step (2); (5) carrying out finite element calculation on the linear elasticity model; and (6) carrying out finite element simulation on a mixed linear elasticity and superelasticity model. By virtue of comparison on a simulation result and an experimental result, the accuracy of the biomechanics modeling method is verified. By virtue of comparison on simulation time of the single linear model and superelasticity model, high efficiency of a simulation algorithm of the mixed linear elasticity and superelasticity model is verified. The biomechanics modeling method is applied to the field of biomedical engineering.

The invention discloses a relay protection distance setting value check method based on measured impedance, comprising the following steps: determining a line to be checked and setting fault points; obtaining the measured impedance of protection related to the line after a minimum short-circuit fault occurs to the fault points of the line, and comparing the measured impedance with the setting value of each section of protection to get the action result of each section of protection; testing the selectivity of protection according to the action result of each section of protection and the time setting value thereof; setting fault points at the head and tail ends of the line to be checked, and checking the sensitivity of protection according to the measured impedance of protection after the minimum short-circuit fault occurs to the fault points of the line; calculating the minimum load impedance under the current operation mode, and testing whether the setting value of section-III protection can avoid the minimum load impedance value; and selecting next line to be checked, and repeating the steps. Compared with the common method at present, the proposed scheme has certain advantages in speed of calculation and accuracy of the check result, and can be better used in distance setting value online checking of a large grid.

The invention discloses a medical CT image segmentation method based on a 3D confrontation network, comprising the following steps: collecting medical CT image samples, performing standardized preprocessing; establishing a final segmentation network; putting any given medical CT image to be identified into the final In the segmentation network, the final segmentation result is obtained. In the present invention, the structural form of the 3D confrontation network is designed as a whole, and the segmentation network is used as the generation network G of the 3D confrontation network, and the pre-trained VGG11 is used as the discrimination network D of the 3D confrontation network. The invention uses the confrontation network to increase the labeling data, and through similarity calculation and cyclic random screening, the credibility of the image is enhanced to generate high-quality labeling image collections in batches, and finally the segmentation is updated through the confrontation training of the discriminant network D and the segmentation network Network parameters and discriminant network D parameters, so as to optimize the segmentation network and improve the segmentation accuracy. The method of the invention is simple in design and easy to implement.

The invention discloses a renewable energy output power probability modeling method based on orthogonal series. The method includes the following steps of S1, obtaining a renewable energy output power value through a monitoring system, and selecting a group of standard orthogonal bases to write a function in the form of the orthogonal series; S2, selecting a shrinkage coefficient for the orthogonal series form in the step S1 through a trimming estimator method, minimizing the deviation and variance of a risk function and a balance function to obtain a cut-off point, and finally determining a probability density function; S3, judging whether the probability density function can reflect the true distribution of renewable energy output power or not through test of goodness of fit. The selection of a bandwidth value does not need to be considered, the model calculation process is simple, and the calculation speed is higher; compared with the prior art, the difference of model output value and actual-measurement data is the smallest. Meanwhile, the method is not restrained by time and space conditions and has the advantages of being high in fitting precision, stability and applicability.

The present invention relates to a single-phase hybrid power flow calculation method based on node current injection. The single-three-phase hybrid model constructed in 1 for the power flow calculation of the distribution network adopts the node current injection method to calculate the power flow of the distribution network. The present invention simultaneously considers the effects of zero-sequence and negative-sequence components on the single-phase model, and ensures the accuracy of the calculation while taking into account the calculation speed.

The invention discloses a logging method of an ultrasonic logger system. An underground gauge of the ultrasonic logger system is of a cylinder shape, m ultrasonic transducers with included angles of theta are evenly distributed at the bottom of the underground gauge, the underground gauge is perpendicularly hung in a well by a cable, the cable is connected with a logging winch by a depth gauge, the logging winch is connected with a well workstation, and the well workstation is connected with a printer. The underground gauge is dropped perpendicularly by the depth gauge, the underground gauge measures the well diameter, the depth gauge measures the dropping length of the cable to obtain the well depth data, and the well diameter data and the well depth data are transmitted to the well workstation. The logging method includes that the well workstation processes the received well diameter data by means of quadratic Bezier curve, or rational quadratic Bezier curve fitting horizontal section algorithm to obtain a well diameter which can correctly reflect the actual shape of a horizontal section.

The invention discloses a Shack-Hartmann wavefront detector based on deep learning, by redesigning the wavefront reconstruction algorithm of the traditional Shack-Hartmann wavefront detector, using the convolutional neural network algorithm to replace the traditional algorithm, and using strong noise data to The integrated neural network training can not only accurately reconstruct the wavefront under strong noise, but also has obvious advantages in operation speed compared with traditional methods because the algorithm is end-to-end wavefront detection. After meeting these two design requirements, the present invention can solve the problem that in the two-photon microscopy system, as the imaging depth inside the biological tissue deepens, the strong noise generated by the tissue makes the wavefront unable to be accurately detected, and can meet the requirements of the two-photon microscopy system. The requirements of the microscopic system for the detection speed of the wavefront, and these two points are the factors that have always limited the application of the wavefront detector in the two-photon microscopic system. significance.

The invention discloses a transient modeling method for a hydrogen circulating pump in a fuel cell system. The model is fitted with a polynomial regression method to obtain the relationship function of the hydrogen circulating pump flow, rotational speed and outlet pressure, and then coupling the inertial link of a driving motor to construct a complete The transient centrifugal hydrogen circulating pump model is based on the model, and finally the control voltage of the driving motor is controlled by PID. The three models are coupled to calculate the transient response of the hydrogen circulating pump, and the hydrogen circulating pump is controlled to meet the exhaust gas recovery requirements of the fuel cell stack. This achieves the purpose of controlling the hydrogen circulation pump. Under the premise of ensuring the accuracy of simulation, the model significantly improves the calculation rate. It can not only be used to calculate the fuel cell stack under steady-state conditions, but also to simulate the real-time calculation of the stack under load changes under actual road conditions. The transient response can be better simulated with the fuel cell system, which is of great significance for optimizing the control strategy of the fuel cell system.

The invention discloses a network reconstruction method and system oriented to a cyber-physical system. The PC algorithm is used to initially construct the causal relationship of each node in the network, and the MCI algorithm is combined to screen the reconstructed connection edges in the PC algorithm, thereby Improve the authenticity and reliability of the reconstructed network.