30results about How to "Stable estimate" patented technology

The invention relates to an airborne SAR radiation external calibration processing method based on weak arrangement of corner reflectors. The method comprises the following steps: obtaining original SAR image data of a calibration field; on the basis of a point target radar equation, reconstructing an radiation external calibration equation; performing stable estimation on calibrator image energy and geometric optical solution on a calibration RCS; carrying out fitting and correcting on a radar system transfer function to obtain preliminary radiation calibration SAR image data; carrying out the stable estimation on the calibrator image energy and the geometric optical solution on the calibration RCS; and estimating a system constant item so as to generate calibration data and obtain final radiation calibration SAR image data. According to the invention, on the basis of the point target radar equation, according to the characteristic of parameters of a radar system, the SAR radiation external calibration equation comprising a change system transfer function and a constant system constant item is reconstructed, the application process of radiation external calibration is simplified, and the problem is avoided that it is quite difficult for airborne SAR flight to obtain a large-scope homogeneous stable tropical rainforest area frequently used by similar spaceborne SAR radiation calibration.

The invention provides a physical code-channel activation and detection method based on channel estimation post-treatment and a device thereof. The method of the invention comprises the following steps: the power of all windows of the user is acquired; the window with the largest power amongst all windows and the position of the window with the largest power of the user are ascertained; activated windows amongst all windows of the user are ascertained according to the window with the largest power and the position of the window with the largest power; the actual total number of the activated physical code-channels of the user and the corresponding activated physical code-channels are ascertained according to the activated windows and the resource of the physical code-channels distributed to the user by the high level. The method and device of the invention effectively solve the problem of unreliability of estimation of key parameters, such as frequency deviation, etc. caused by lacking of data quantity, thus enhancing the performance of the system.

The invention provides a battery model parameter identification method based on a multi-innovation recursive Bayesian algorithm. The method comprises the following steps of: 1), measuring the terminalvoltage and load current data of a lithium ion battery in a certain period of time through an intermittent constant-current discharge method, and determining the function relation of an OCV-SOC of the lithium ion battery through a polynomial fitting method; 2) determining a dual-polarization equivalent circuit model of the lithium ion battery, and establishing a system equation representing a relation between a battery parameter identification vector and system output; and 3) constructing an identification process of the multi-innovation recursive Bayesian algorithm. According to the method of the invention, an ARX model for lithium ion battery parameter identification is established; the result of the previous moment is corrected by utilizing an innovation correction technology; an innovation length parameter is introduced based on the multi-innovation identification method, so that the influence of bad data on parameter estimation is overcome, and the parameter estimation precisionis improved; and the parameter identification result shows that the method is high in identification precision and has engineering value.

The invention provides a GNSS satellite clock error real-time estimation quality control method aimed at characteristics that real-time clock error estimation has high requirement on observation data quality and cannot perform post residual error edition. The method comprises two aspects: first step, before parameter estimation, preprocessing original observation data, including gross error rejecting and cycle slip detection; second step, after update of filter observation, based on a posttest observed value residual error, analyzing and processing a gross error and cycle slip which are not detected in a previous step, specifically including steps of establishing a square root information filtering algorithm model, processing noise; smoothing square root information; and based on square root information filtering, performing real-time quality control. The method can realize preprocessing of real-time clock error data and efficient quality control, and can efficiently find out problematic observed quantity, prevents filter divergence, so as to realize real-time clock error stable estimation and satisfy real-time high-precision positioning requirement.

Provided is an apparatus and a method for controlling a motor which stabilizes the estimation of a rotor position even when a control for lowering the applied voltage to the motor is performed. The apparatus for controlling the motor estimates the position of the rotor (31) based on the induced voltage of non-energizing phase of a brushless DC motor (29) to be driven by an inverter (32), outputs a PWM switching signal pulse-width-modulated based on this estimated result to the inverter (32), and controls the applied voltage to the brushless DC motor (29). When the switching signal having the duty ratio corresponding to the applied voltage is generated, the apparatus includes a controller (34) for regulating the carrier frequency of the PWM switching signal so that the pulse width of the switching signal becomes a predetermined pulse width or more.

The invention provides a high-resolution low-sidelobe deconvolution spectrum estimation method based on space-time processing, and belongs to the field of signal processing. The method is suitable forthe high-resolution power spectrum estimation of a received array signal at frequency and azimuth under the condition that the data sample is limited. According to the invention, a deconvolution algorithm is used for the multi-dimensional deconvolution operation of an array signal power spectrum with the limited length and a multi-dimensional window function power spectrum, so the method can completely eliminate the impact caused by a window function under the ideal conditions, and obtains a real power spectrum of a signal. Based on the conventional convolution technology, the method can greatly improve the frequency resolution and angle resolution of a power spectrum estimation result, can greatly inhibit the sidelobe caused by the limited length of data, provides an additional signal tonoise ratio, and provides a new idea for the multi-target detection and recognition and weak signal extraction in the array signal processing. In addition, the method is high in operability, is moderate in calculation burden, is stable in calculation result, and is very wide in application range.

The invention discloses a display image processing method and device. The method comprises the following steps: carrying out direction clustering on straight lines in a display image to obtain two straight line sets containing the largest number of straight lines; Calculating a blanking point corresponding to each straight line set in the two straight line sets; Performing data processing on blanking points corresponding to the two straight line sets respectively to obtain a perspective transformation matrix; And correcting the display image based on the perspective transformation matrix to obtain a corrected image of the display image on the orthographic projection surface. The technical problem that perspective transformation with high precision cannot be obtained in the image correctionprocess in the prior art can be solved.

The invention discloses a scatter matrix feature-based second order partial differential equation remote sensing image denoising method, and belongs to the field of image processing. Firstly, a control function of diffusion terms is established by introducing a 45N degree rotation module value invariance operator, so that diffusion is performed in a gradient direction and a vertical gradient direction in a smooth region and is performed in a curvature direction only in an edge region; and secondly, a better distinguishing capability for feature regions of an image is realized through a scattermatrix, and based on this, fidelity terms are created, so that the denoising effects of a sharp corner region and the smooth region are improved, and the over-smoothness effect of an MCM at a targetsharp point and the "blocky" effect generated by a PM model are effectively relieved.

The invention provides a hotspot grid identification method and device, electronic equipment and a readable storage medium. The method comprises dividing a specific area into a plurality of grids; forthe grids, acquiring AOD concentration data based on satellite inversion, and acquiring surface feature information of each grid; calculating difference data of the surface feature information and difference data of the AOD concentration data based on satellite inversion; performing learning through a machine learning method to obtain a correlation model of the difference data of the surface feature information and the difference data of the AOD concentration data based on satellite inversion; according to a correlation model, calculating a pollution emission influence weight for each of theone or more types of surface feature information; and screening hotspot grids by utilizing the pollution emission influence weight and the surface feature information. According to the method, the problems of data instability and insufficient availability caused by cloud shielding in the existing hotspot grid screening technology can be solved, and hotspot grids can be accurately, objectively andstably screened.

The embodiment of the invention discloses a water-fat separation method, device and equipment based on a magnetic resonance image and a storage medium. The method comprises the following steps of obtaining each pixel point to be processed in a magnetic resonance image, respectively calculating a phase factor candidate solution of the pixel point, and respectively dividing each phase factor solution in the phase factor candidate solution into a first solution set and a second solution set; based on the phase factor solutions in the first solution set and the second solution set, determining a water lipid conversion area, and calculating a target phase factor solution of each pixel point to be processed in the water lipid conversion area; determining a target phase factor solution of each ofthe rest to-be-processed pixel points based on the water-fat conversion region and the target phase factor solution in the water-fat conversion region; and extracting a water map and a fat map in themagnetic resonance image according to the target phase factor solution of each pixel point to be processed in the magnetic resonance image. According to the technical scheme provided by the embodiment of the invention, the field diagram can be rapidly and stably estimated, and the correct estimation of water and fat components is completed on the basis.

In various embodiments, methods and systems are disclosed for the real time detection of network conditions in conjunction with a remote presentation protocol. The link quality may represent the quality of the end-to-end connection between client and server with upper and lower bounds on the injection of additional traffic used for measurement. In some embodiments, the measurement technique may be selected based on the type of measurement that is desired. Accuracy may be maintained by selecting the type of measurement used based on current and previous network conditions. In one embodiment, a state model is used to determine the frequency of measurement and to determine when the measurements have produced a stable estimate of the link quality.

The invention discloses a space channel characteristic prediction method and system based on an m sequence and a storable medium, and belongs to the technical field of underwater acoustic channel communication, and the method comprises the steps: setting an array parameter and a transmitting signal parameter; sampling signals are intercepted according to the set array parameters and emission signal parameters; carrying out delay summation beam forming on the intercepted sampling signal; capturing data formed by the delay summation beam to obtain a capturing result; generating a delay angle distribution diagram according to a capture result, and predicting signal delay and an incident angle; and performing Doppler analysis on the path signal according to a signal delay and incident angle prediction result to complete spatial channel characteristic prediction. According to the method, the problems of traditional direction-of-arrival prediction, time delay and Doppler are solved, so that the prediction performance of spatial channel characteristics is improved.

The invention discloses an LDoS attack detection method based on an EMDR-WE algorithm, and belongs to the field of computer network security. The method comprises the following steps: in view of the characteristic of high complexity of a TCP flow sequence under the LDoS attack, quantifying the TCP flow sequence complexity of an attack window and a normal window by combining four characteristic entropies, namely approximate entropy, sample entropy, fuzzy entropy and permutation entropy; constructing a preprocessing model for empirical mode decomposition and reconstruction of a TCP flow sequence, filtering noise components of the TCP flow sequence through the model, obtaining a sliding window sequence of TCP flow, and extracting four feature entropies of each window; giving a higher weight to the feature entropy with the large information contribution degree by using an entropy weight method to obtain a comprehensive score of the complexity of each TCP flow window. comparing with a threshold value obtained by logistic regression training, and judging that the window with the comprehensive score higher than the threshold value has the LDoS attack. According to the LDoS attack detection method based on the empirical mode decomposition, reconstruction and entropy weight method, the LDoS attack can be accurately detected, and the performance is stable.

A control apparatus for estimating the flux, the current, and the speed of an AC motor using the current and the voltage, and controlling the vector of an AC motor using the estimated speed, a flux command, and a torque current command. The speed is estimated by adding a product of the deviation between the actual value and the estimated value of a magnetization current, the level of a torque current correspondence value, the sign correspondence value of a primary frequency command value, and a gain to an outer product of an estimated current deviation and an estimated flux. Thus, a stable speed estimating operation can be performed to successfully operate the motor in a low speed area in which a voltage frequency applied to the motor is extremely low.