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

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 discloses a wind power output typical scene generation method based on BIRCH clustering and a Wasserstein distance, and the method comprises the steps: firstly obtaining an optimal quantile of a wind power probability through the minimization of the Wasserstein distance, and discretizing a continuous probability density function into a plurality of probability density points; performing iterative reduction, splicing and re-reduction on the initial scene set by applying BIRCH clustering to obtain a typical scene set. According to the method, the wind power output scene can be quickly and accurately reduced, and compared with other algorithms, the algorithm has great advantages in calculation time and storage scale.

The invention relates to the technical field of power grid transient simulation calculation and specifically relates to a motor load transient process calculation method suitable for electromagnetic transient simulation. The method can quickly calculate load transient characteristics of an induction motor under symmetrical and asymmetrical fault voltage sag of a power system, and is characterizedby (1) firstly, solving a first-order mechanical transient equation of a rotor to obtain changes in speed and slip of the rotor after disturbance; (2) solving a transient potential differential equation in the form of a rotor phasor to obtain change in transient potential; and (3) obtaining current and power response of the motor after the disturbance. Compared with PSCAD/EMTDC electromagnetic transient simulation results, the result of the calculation method verifies effectiveness of the algorithm; and compared with PSD/BPA electromechanical transient simulation results, the result of the calculation method shows the advantages of the algorithm under asymmetric voltage sag. An analytical method explicitly gives computation expressions of mechanical and electrical parameters of the induction motor after the disturbance; and in addition to being used to quickly evaluate mutual influence of the induction motor and the voltage sag, the method can also be used in stability analysis of thepower system.

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 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 single image defogging method and a single image defogging device based on bilateral filtering and median filtering. The single image defogging method comprises the steps of:estimating an atmospheric light value of an original foggy image based on dark channel priori; adopting bilateral filtering for estimating ambient light roughly according to the atmospheric light value; performing median filtering on the ambient light to obtain improved ambient light; and restoring the original foggy image based on the atmospheric light value and the improved ambient light. The single image defogging method combines the advantages of bilateral filtering and median filtering, can fully avoid color oversaturation of the image and a halo effect of image edges, and can effectively deal with the foggy image impurities at the edges and corners.

The invention provides a water resource optimal allocation method based on a genetic algorithm, and the method achieves the normalization of an economic target, a social target and an ecological target, enables all targets to be subjected to weighted calculation under the same dimensional condition, and enables the weights of all targets to have more practical significance after normalization. Meanwhile, the constructed water resource optimization model has the characteristics of multiple variables, multiple targets, multiple constraints and the like, when the model is solved, a genetic algorithm is adopted, and compared with a traditional solving method, the provided method has great advantages in the aspects of calculation speed, precision, convergence, adaptability and mobility.

The invention relates to a node current injection based single three-phase hybrid tide calculation method. The technical feature of the method is that the method includes steps of 1, constructing a single three-phase hybrid model used for power distribution network tide calculation; 2, performing power distribution network calculation by adopting a node current injection method based on the singlethree-phase hybrid model used for power distribution network tide calculation constructed in step 1. According to the invention, influence due to zero sequence and negative sequence components are taken into account, and calculation precision is ensured at the same time of ensuring the calculation speed.

The invention discloses a method for simulating and optimizing the echo signal intensity of a moving surface ship, which comprises the following steps of: establishing a three-dimensional ship body model below the water surface according to a ship model line graph, subdividing a triangular surface grid, solving a moving posture of the three-dimensional ship body model and a two-dimensional rough sea surface complex reflection coefficient in combination with different sea condition parameters, and according to the established four-path echo model of the water surface ship and four-path echo signals of the moving water surface ship, superposing the echo signal intensity of each path, and simulating the echo signal intensity of the moving water surface ship. The method disclosed by the invention has a remarkable advantage in calculation speed, and the average error between the theoretical simulation result and the actual measurement result of the echo signal intensity is less than 2dB, so that the echo signal intensity of the moving surface ship can be accurately and quickly simulated.

The invention discloses an image fuzzy classification method applying image correlation. The method comprises the steps of 1, for to-be-classified images, calculating fractal dimension characteristicsand energy characteristics of the images, and according to the values of the fractal dimension characteristics and the energy characteristics, classifying the images into three types; 2, marking thetotal image quantities in the three types of the images as n1, n2 and n3 respectively, calculating texture characteristic mean values of the characteristics of the three types of the images, and marking the texture characteristic mean values as mean1, mean2 and mean3 respectively; 3, calculating differences between each image and the texture characteristic values of the three types of the images;4, calculating the correlations between each image and the three types of the images; 5, calculating a mean value of the correlations of the three types of the images in the respective types; and 6, determining whether the images belong to the current type of the images or not.

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.