177results about How to "Short computing time" patented technology

The invention discloses a high-sensitivity satellite navigation signal capturing method and a system. The system comprises a digital down-conversion module, an average sampling and block accumulation module, an FFT (fast Fourier transform) module, a circumference shifting module, a local PRN (pseudo random noise) code FFT conjugate memory, a complex multiplier module, an IFFT (inverse fast Fourier transform) module, a differential coherence integration module, a peak detection module and a sequential control module. The digital down-conversion module realizes digital down-conversion operation for satellite digital intermediate frequency signals; the average sampling and block accumulation module averagely samples satellite data and completes a block accumulation function; the FFT module searches code phase frequency domains; the circumference shifting module utilizes Doppler circumference shifting search to replace frequency compensation; the local PRN code FFT conjugate memory stores a local PRN code FFT conjugate result; the complex multiplier module realizes signal de-spreading; the IFFT module calculates different code phase coherence results; the differential coherence integration module accumulates differential coherence energy of de-spread satellite signals; the peak detection module realizes signal capturing output; and the sequential control module controls timing sequence of the various modules of the system. Weak signal capturing speed and sensitivity of a satellite navigation receiver are improved, and parameters can be configured flexibly.

The invention discloses a time series data cleaning method based on correlation analysis and principal component analysis, comprising the following steps: finding out hidden correlation between transformer faults and other power data by using a Pearson's correlation coefficient (PCC); reducing the dimension and noise of all relevant time series through principal component analysis (PCA); and inputting part of cleaned data as a training set into a BP neural network (BPNN) for training and learning, and taking the remaining data as a test set to verify a model. Compared with the traditional technology, the accuracy of transformer fault diagnosis is significantly improved, the accuracy of classification is improved, and the operation time is shorter for high-dimensional data.

The invention provides a pipeline FFT processor variable in the number of points, which comprises a first 1024-point variable FFT processing module, a twiddle-factor processing module, a second 1024-point variable FFT processing module and a selection-control module, wherein the four modules and an intermediate data storage module outside the processor jointly complete FFT two-dimensional processing large in the number of points; each 1024-point variable FFT processing module comprises a first 32-point variable FFT processing sub-module, a second 32-point variable FFT processing sub-module, a twiddle-factor processing sub-module, an intermediate data storage sub-module and a selection-control sub-module; variable-point FFT operation is implemented through the 32-point variable FFT processing sub-modules; the twiddle-factor processing module generates and multiplies intermediate twiddle factors by the result of FFT operation; and the selection-control module realizes control over a whole chip. The processor is suitable to be realized in single-chip FPGA or ASIC, and can simultaneously obtain high speed, low power consumption, high precision and other characteristics.

The invention relates to a CT (Computed Tomography) transmission imaging beam hardening correcting method and particularly relates to a base image TV minimized model based CT beam hardening correcting method. On the basis of establishing a beam hardening correcting model with an adjustable parameter, the base image TV model based CT beam hardening correcting method comprises the following steps: firstly, converting an original projection data through the model pretreatment so as to obtain sets of pretreated projection sequences under different adjustable parameter conditions; respectively reconstructing the pretreated projection sequences, thereby obtaining a series of corrected base images; taking a full-variation function of a target image as a cost function and obtaining an optimal solution of a weighting coefficient according to an iterative method; and lastly, summing the weights of the obtained series of base images, thereby forming a final reconstructed image. Compared with the traditional method, the method provided by the invention has the advantages that the priori knowledge, such as scanning copy material, radiation source condition, and the like, is unnecessary, the method is unlimited by external conditions and is strong in universality, and the linear combination is solved when a weighting coefficient is confirmed, so that the method has the advantages of small computation complexity and short operation time.

The invention discloses a nonlinear system state estimation method based on Kalman filtering positioning, and proposes a strongly adaptive Kalman filtering mechanism which combines a nonlinear filtering algorithm and Kalman filtering. The method comprises the steps: carrying out the simultaneous estimation of node positions and channel parameters through employing an RSSI state estimation algorithm based on square root volume Kalman filtering, and obtaining the estimation value of a state vector; employing the Kalman filtering for further processing according to the linear change of a state equation, obtaining optimal estimation, and building a strongly adaptive square root volume Kalman filtering algorithm; giving the design steps of the strongly adaptive square root volume Kalman filtering algorithm; and calculating a theoretical square root error lower bound under a state space model based on the RSSI state estimation. The method enables an estimation result to be improved, and improves the precision. The method does not need to excessively depend on improper initial conditions, can be well adapted to a highly nonlinear system, and is not liable to enable the algorithm to be divergent and ineffective.

The invention provides a method for controlling axial direction of a high-precision magnetic bearing based on an interference observer. The method consists of a controller and the interference observer, wherein the interference observer comprises a Q filter and a rationalization generalized object inverse QGn part; the controller in a control system calculates according to displacement deviation to obtain basic control quantity so as to form a position closed loop control system; an interference value obtained by observation of the interference observer is negatively fed back into the basic control quantity to compensate exterior interference; and the formed current control quantity drives power amplification to realize high precision suspension of the magnetic bearing. The method leads differences caused by both exterior interference and objective parameter variation to be equivalent to a control input end, and introduces equivalent compensation into the control quantity to realize interference inhabitation. The method can carry out online observation and effective inhabitation against exterior interference which is not modeled or known, thereby improving control precision of suspension, and contributing to stability of the system.

The invention relates to a power system load prediction method based on a Markov chain. Under the condition that a value Lt-1 is known, according to historical data, various change trends of next time t are counted, probabilities are counted, and a trend with a largest probability is taken as a final prediction result. The method has the following advantages: load prediction can be carried out with a few samples, operation speed is fast, operation time is short, and a result of probability prediction can be obtained.

The invention relates to a ground penetrating radar underground target position detection method. The method comprises the following steps: a ground penetrating radar B-Scan data are subjected to multi-scale amplitude, phase and direction characteristic component extraction; three characteristic components for the first scale are extracted, and a color image is combined by using an HSB channel representation method; a target area of interest is found out; graying is carried out on the color image in the target area of interest, and edge extraction is carried out on an amplitude image in the target area of interest; a target hyperbolas is positioned; and according to the target hyperbolas, target position determination is realized. the method of the invention does not need data training, and online detection is easily carried out; the target information can be kept in a complete mode, the target positioning precision is enhanced, and as for shallow layer detection, false alarms are not likely to be generated, and non metallic pipeline targets are not likely to be omitted; clutter suppression effects can be enhanced, and strong clutters and target echoes can be effectively distinguished; and the algorithm is simple, and target positioning can be ensured to be quickly carried out.

The invention discloses an improved k-means clustering method based on a distributed computing platform, and introduces a distributed computing platform Spark for the problem of slow processing of mass data, the Kruskal's algorithm for the problem of too many iterations and the Tanimoto distance for the problem of giving no consideration to the correlation among features of a vector. First, a minimum spanning tree is constructed for the randomly selected k points by using the Kruskal's algorithm, the corresponding weight sum is obtained, and the process is repeated for n times. Then, according to weight sums obtained within the n times, the maximum weight is selected thereform and that distance values between edges composed of the k points are not much different is ensured. In this way, the relatively uniform distribution of cluster centers can be guaranteed. Finally, a clustering algorithm is performed using k-means algorithm improved by using the Tanimoto distance.

The invention relates to a method for collision and occlusion detection between virtual and real objects. The method comprises: firstly, establishing a world coordinate system, and positioning a virtual three-dimensional model in a virtual model scene; secondly, using a physical depth camera to capture real environment depth images, and obtaining the position and azimuth of the physical depth camera by using an azimuth tracking system; thirdly, a computer obtaining three-dimensional orientation of the physical depth camera, defining the three-dimensional orientation as a first orientation; placing the virtual depth camera on a second orientation, and synthesizing a first synthetic depth image formed on the virtual depth camera by the virtual three-dimensional model according to the imagingmodel of the depth camera; calculating a depth value of each pixel point on the first synthetic depth image in the first orientation, to obtain a second synthetic depth image; finally, performing collision and occlusion detection and distance calculation of virtual and real objects by using the second synthetic depth image and the real environment depth image. The method is based on the depth image and rasterizes the depth image, and carries out collision and occlusion detection. The processing speed is fast and detection is accurate.

The invention discloses a polarized image fusion method based on discrete continuous curvelet, and belongs to the field of image processing. The method comprises the following step: firstly, performing discrete continuous curvelet to polarization strength and polarization ratio images to obtain low-frequency sub-band coefficients and each direction sub-band coefficient; then selectively fusing low-frequency sub-band coefficient for the low-frequency sub-band coefficient by using weighted average criterion, selectively fusing each direction sub-band coefficient for each direction sub-band coefficient by using area energy maximal criterion, and finally performing inverse discrete continuous curvelet to obtain a final fusion image. The discrete continuous curvelet is realized fast by adopting the Wrapping-based method, and the redundant information of the conversion result is low. Experimental results prove that the algorithm disclosed by the invention is very effective; furthermore, the edge and space grain information of the fused image are clear, the computing time of the algorithm is short, and the image information can be displayed well in real time.

The invention provides a grinding surface roughness prediction method based on an improved support vector machine algorithm. Parameters of a support vector machine algorithm are optimized by using a grasshopper algorithm, and the method specifically comprises the following steps: (1) designing grinding machining parameters and an experimental method, and collecting machining parameters in the grinding machining process; (2) dividing the data into a training set and a test set by using the thought of cross validation, and performing normalization processing; (3) enabling the grinding wheel linear speed, the workpiece speed and the grinding depth to serve as input parameters, enabling the surface roughness to serves as output parameters, and constructing a GOA-SVM prediction model for predicting the roughness of the grinding surface. Compared with other prediction grinding, the grinding surface roughness prediction method based on the improved support vector machine algorithm provided bythe invention can rapidly seek a globally optimal solution in a complex search space, has the characteristics of low cost, high precision and easiness in operation, and has a smaller error between aprediction value and a true value.

The invention discloses a preprocessing and adaptive genetic simulated annealing algorithm-based phase unwrapping method. The method includes the following steps that: adjacent dipole pair residues within a distance threshold value R are pre-processed; the residues are optimized and combined through adopting an adaptive genetic simulated annealing algorithm; and each pair of combined positive and negative residues are connected together through branch-cut lines in an interferometric phase image, and the branch-cut lines are bypassed through using a flood-fill method, so that phase unwrapping can be performed. According to the preprocessing and adaptive genetic simulated annealing algorithm-based phase unwrapping method of the invention, a part of dipole pair residues are pre-processed through a pre-processing method of the variable distance threshold value R, so that the lengths of the branch-cut lines can be formed; the residues are optimized and combined through the adaptive genetic simulated annealing algorithm, so that the total length of the branch-cut lines can be formed, and closed regions are few, and total computation time is little; and since the set branch-cut lines are short, accumulative errors around the branch-cut lines can be decreased when unwrapping is performed, and therefore, unwrapping accuracy can be improved, and the number of unwrappable islands can be decreased.

The invention relates to a Markov chain based method for accurately forecasting power system loads. The method is characterized by counting various variation trends of the next moment t according to the historical data and counting the probability of the variation trends under the condition of the given value Lt-1, finally taking the trend with the highest probability as the forecast result and obtaining the final result by correcting the sum of the probability on the two sides of the forecast result. The method has the following advantages: the loads can be forecast by only a few samples; the operating speed is high; the operation time is short; and the accurate probability forecast result can be obtained.

The invention discloses a system for testing the power of a loudspeaker. The system comprises a computer; a plurality of test groups are connected onto the computer; each test group comprises a loudspeaker to be tested, a power tester and a power amplifier, wherein the power tester is used for detecting the loudspeaker; and the power amplifier is used for driving the loudspeaker. In the system for testing the power of the loudspeaker, a loudspeaker testing instrument is simple in structure; the number of required equipment is small; and the cost is low. The invention further provides a method for testing the power and a low-frequency parameter of the loudspeaker. The method for testing the power of the loudspeaker has the advantages of being simple to operate, accurate in low-frequency parameter measurement, rapid in operation speed and short in operation time.

The invention brings forward an underground work personnel accurate positioning system and work method mixing WIFI-UWB positioning technologies. The system mainly comprises five major modules, i.e., a WIFI fixation device, a UWB fixation device, a center server, a WIFI receiving apparatus and UWB positioning label of a mobile point to be detected, and a ground monitoring center. The method comprises the following steps: S1, according to an RSS-based WIFI positioning technology employing a triangular positioning method, determining the position of a point to be detected in a triangular area; S2, the UWB fixation device in the area where the mobile point to be detected is disposed measuring the position of the point to be detected by use of a TOA-based positioning technology, and determining a circular curve taking a measurement distance as a radius; and S3, combining the obtained triangular area and the circular curve, and performing analysis according to four different conditions of zero intersection point, one intersection point, two intersection points and three intersection points to obtain accurate coordinates of the mobile point to be detected. According to the invention, the underground work personnel positioning accuracy is improved.

The invention discloses a box particle filtering based extension target CBMeMBer tracking method, belongs to the field of radar signal processing, and mainly solves the problems of long time for realizing extension target CBMeMBer tracking and easiness for crossing of measurement and division in particle filtering. The method comprises the steps of extension target measurement and generation, extension target measurement and division, conversion of a target state into an interval form, box particle based CBMeMBer prediction, updating and target number estimation. According to a principle that measurements generated by the same target are closer, measurements with a Mahalanobis distance smaller than a distance threshold are included in the same division unit, so that extension target measurement and division are completed. Then, prediction and updating are performed with a box particle filtering based CBMeMBer method, so that tracking and state estimation of a target are completed. The method has the characteristics of good filtering performance and short calculation time, and can be used in the fields of radar signal processing and the like.

The invention discloses a digital image triple interpolation amplification method by combining local direction features. According to the method, firstly, four direction feathers with low-resolution-ratio images are extracted, and then, interpolation points are divided into three types according to the position relationship, and the interpolation is respectively carried out. For different-line different-row interpolation points, distances between known points and points to be interpolated and corresponding direction feathers are firstly used as weighting coefficients for carrying out interpolation, then, the direction feathers of the known points in the adjacent region are used for judging whether sub blocks are 45-degree and 135-degree high-texture regions or not, and if so, the corresponding interpolation points are updated in the texture direction by a linear interpolation method; and when same-line same-row points to be interpolated are evaluated, interpolation points of 0-degree and 90-degree high texture regions are subjected to interpolation in the corresponding direction, and other points are reconstructed according to the traditional linear interpolation method. The method has the advantages that edge fuzzyness and sawtooth phenomena of the interpolation images can be effectively inhabited, meanwhile, the operation time is shortened, the edges of the reconstructed images are clearer, and the visual effect is sharper.

The invention requests to protect a face feature extraction method based on fusion of an improved LTP (Local Ternary Pattern) and two-dimensional bidirectional PCA (Principal Component Analysis). The method comprises the following steps that: S1: establishing a difference function equation of the weight value sum of a center pixel and a neighbourhood pixel to solve the statistical characteristics of the center pixel and the neighbourhood pixel; S2: establishing an adaptive gradient descent iterative function to calculate a weight coefficient which enable a difference equation to be minimum to define the threshold value of an IALTP (Improve Adaptive Local Ternary Pattern) operator; S3: taking the mean value and the standard difference of local area pixel weight as the coding way of three patterns of the IALTP; S4: solving a projection matrix of a face image in a line and row direction through the two-dimensional bidirectional PCA, and then, establishing joint mapping to obtain the global feature information of the face image; and S5: utilizing a way of fusing local features and global features to put forward an algorithm which fuses the IALTP with the two-dimensional bidirectional PCA. By use of the method, a high identification rate can be obtained, and the method can exhibit high robustness on illumination and random noise.

The invention belongs to the technical field of image processing, and particularly provides an automatic selection method for an optimal homogenous region of an ultrasound image and based on quad-tree decomposition. According to the automatic selection method for the optimal homogenous region of the ultrasound image and based on the quad-tree decomposition, firstly, a threshold value determined through maximum between-cluster variance binarization is used as a consistency standard for the quad-tree decomposition of the ultrasound image; secondly, all current maximum blocks are sequentially taken out from decomposition results according to an order from large to small, and optimized selection is conducted according to the judgment basis of the blocks in the optimal homogenous region; and lastly, whether an automatic selection result of the optimal homogenous region is output or the initial binarized threshold value is adjusted to repeat the above steps is determined according to a judgment result. The automatic selection method for the optimal homogenous region of the ultrasound image and based on the quad-tree decomposition is high in efficiency and short in operation time; manual intervention is not needed, and only ultrasound image data are used for driving; and the automatic selection method is quite suitable for determination of homogenous regions at the time of handling noise reduction of ultrasound image spots, selection of an area-of-interest, boundary detection, target segmentation and the like, and has a remarkable actual using effect.

A rapid cross section contour generation method suitable for a triangular mesh model comprises the following specific steps that step A, intersecting a mesh surface patch of the triangular mesh modeland a cross section to generate cross section intersection points, coordinates of the cross section intersection points are obtained, and the coordinates of the cross section intersection points are stored in a cross section intersection point set; step B, two vertexes of the edge where the cross section intersection point is located, and storing all retrieved vertexes in a vertex set in consideration of repeatability; step C, classifying the vertex sets according to the vertex retrieval times in the step B to obtain vertex sets with different vertex retrieval times; and step D, generating a model contour line according to the vertex set. The invention provides a rapid section contour generation method suitable for a triangular mesh model, which solves the defect problem in the backgroundtechnology, so that the occupied memory of a computer is smaller, and the efficiency is higher.

The invention relates to a processing method and a processing device for image zooming. The method comprises the following steps: bitmap formatted data of a source image is obtained by receiving and decoding information data of the source image; the amount of pixels of the bitmap formatted data of the source image is changed according to the position coordinates of the pixel on the basis of a first set image zoom scale to form bitmap formatted data of an intermediate image; the amount of the pixels of the bitmap formatted data of the intermediate image is changed according to the average of color values of adjacent pixels on the basis of a second set image zoom scale to form bitmap data of a target image. The invention realizes zooming for images in various sizes, meets different requirements on display equipment and easily realizes hardware circuit.

The invention discloses a method for determining a lubricating oil filling amount of a transmission. The method comprises the steps of: establishing a transmission analysis model, grouping components, setting the rotating speeds of moving components, setting the lubricating oil filling amount, obtaining n+1 groups of analysis data I by simulation analysis and calculation, changing the lubricating oil filling amount for n times, obtaining n groups of analysis data I by simulation analysis and calculation, determining a first lubricating oil filling amount, setting the rotating speeds and the lubricating oil filling amounts of the moving components, obtaining m+1 groups of analysis data II by simulation analysis and calculation, determining a second lubricating oil filling amount, and finally determining the lubricating oil filling amount of the transmission. By adopting the method disclosed by the invention, before the transmission is made into a finished product, the lubricating oil filling amount of the transmission is determined quickly and efficiently, thereby reducing the test cost and resources.