42 results about "Weighted least squares estimation" patented technology

The invention provides a two-stage linear weighted least-square power system state estimation method which is characterized by comprising the following steps of: forming a network model, and calculating a node admittance matrix and a branch-node relationship matrix; transforming a measurement vector and a state vector; forming an accurate linear measurement equation; performing the first stage of linear weighted least-square estimation to obtain the estimated value of the transformed state vector; performing inverse transformation, and performing the second stage of linear weighted least-square estimation to obtain the estimated values of the voltage amplitudes and phase angles of all nodes; and identifying bad data. According to the two-stage linear weighted least-square power system state estimation method provided by the invention, more scientific state estimation result can be obtained, the calculation efficiency is higher, and the engineering application prospect is good.

The invention relates to a TDOA-AOA positioning method considering a position error of a base station. The method comprises the following steps: 1) arranging a wireless sensor network in a two-dimensional space, and measuring, by each base station, an arrival time dereference and angles of a measurement target node and a reference node under a noise condition; 2) selecting the first base station as the reference base station, and obtaining a signal transmission path distance difference relative to the reference base station according to the arrival time dereference of signal transmission paths; 3) establishing a least square target estimation algorithm according to the measured values in the previous two steps and the known position of the base station; 4) due to the existence of the basestation error and the measurement error, analyzing the errors, and importing a weight matrix to achieve weighted least square estimation; and 5) reconstructing the weighted least square estimation, and using the semi-definite relaxation technique to transform the weighted least square into a semi-definite planning problem with linear equality constraints and linear inequality constraints via laxity rank constraints, and then solving a globally optimal solution by using an inner point method.

The invention discloses a compressed sensing based sparse channel estimation method in a two-way relay network. The method includes: according to actual signals received in a second time slot by a current user terminal and a Gaussian random training sequence sent by each user terminal in the two-way relay network, subjecting an equivalent channel vector between the current user terminal and a relay node to channel iterative loop estimation to obtain a channel estimation value; and using a weighted least square estimation method during channel iterative loop estimation to continuously weaken influences of abnormal samples corresponding to positions with large residual values after the last iteration and enhance attention to favorable samples corresponding to positions with small residual values, so that precision and robustness in sparse two-way relay channel estimation can be effectively improved. Further, since the compressed sensing based sparse channel estimation method in the two-way relay network makes full use of potential sparse characteristics of two-way relay channels, the Gaussian random training sequences required by estimation can be greatly reduced, estimation efficiency is improved, and energy and bandwidth utilization rate is greatly increased.

Method for establishing a commercial real estate price change index supporting tradable derivatives. The method utilizes a database of price changes actually experienced by individual commercial properties including allowing for gradual accumulation of price data. The database is filtered with selected data filters and time-weighted dummy variables are specified. A repeat-sales regression is performed on the filtered database to create the index. The repeat-sales regression includes a weighted least squares estimation in which weights are determined in a three-stage regression process. A ridge regression noise filter in which a first order autocorrelation coefficient in estimated index price-change returns controls the ridge estimation, and the first autocorrelation coefficient is near 0. The index is optimized for derivative trading purposes by excluding backward adjustments and a scope and frequency for the index is selected. The index may be used for tradable derivatives.

The invention discloses a parameter estimation method for an electric power circuit in T-type connection, which belongs to the technical field of parameter identification and estimation for circuits of power systems. A computer and data acquisition devices at three ends of the circuit in T-type connection are adopted, SCADA (supervisory control and data acquisition) measured data for the circuit in T-type connection under a plurality of load levels at multiple time intervals are imputed by the aid of a program, then a weighted least squares estimation model is solved to obtain estimated values of parameters of the circuit in T-type connection under one load level, namely a sample of estimation values, then a mean value of a plurality of samples of estimation values is calculated, coefficient variations of the samples are used as convergence tolerance of parameter estimation precision, and finally estimation values of the parameters of the circuit in T-type connection are obtained. By the aid of the parameter estimation method, the parameters of the circuit in T-type connection can be effectively estimated, and the parameter estimation method has the advantages of simplicity, quickcalculation speed, high estimation precision, good stability of data, high engineering practicality, convenience in popularization and application, and the like. In addition, the parameter estimationmethod can be widely applied to parameter estimation of optional electric power circuits which are in T-type connection and are provided with data acquisition devices.

The invention discloses a method for detecting a satellite clock anomaly. After a satellite signal is decoded, message information is obtained and is sent to a CPU. The CPU judges whether an information stream read from the message information is accurate or not. If the information stream is accurate, a time processing module is controlled to tame PPS through utilization of a1PPS. Standard difference between the 1PPs and world standard time, and an accumulated error of the PPS are estimated through establishment of a weighted least square estimation model. The estimated standard difference iscompared with a standard difference set value of a GPS / BDS receiver. When the estimated standard difference is smaller than the set value, the time processing module carries out error correction on the PPS and carries out continuity and tendency judgment on a time interval error of the corrected PPS. If no error exists, a satellite clock is normal. Corrected PPS is output. According to the method,through multistep detection, whether the satellite clock is abnormal or not can be effectively identified, a high precision clock signal is output, and influence of satellite clock anomaly on outputtime of a time synchronization device is prevented.

Embodiments provide systems and methods for determining the geolocation of an emitter on earth based on weighted least-squares estimation based on two TDOA and two FDOA measurements, none of which need to be acquired at the same time. The four TDOA and FDOA measurements and the errors in each of the measurements are determined. Weights for the errors in the TDOA and FDOA measurements are determined, and the weights are applied in a weighted errors function. The weights account for the errors in the measurements and the errors in the satellite positions and velocities, and are dependent on the localization geometry. The weighted errors function is minimized to determine the location estimate of the unknown emitter.

The invention discloses an AUV intelligent motion observation method based on data driving. Current observation ocean data acquired by an underwater autonomous submersible vehicle is represented by alinear combination of a group of primary functions; secondly, the marine data to be observed is processed by adopting a recursive weighted least square estimation method to obtain an observed marine data coefficient, an estimated value of the observed marine data at the latest moment is obtained by utilizing the observed marine data coefficient, and a gradient of the estimated value of the observed marine data is obtained; and finally, the underwater autonomous submersible vehicle reaches the position of the marine observation center in the gradient direction. According to the invention, underthe drive of new observation data, the AUV can quickly reach the central position of the marine characteristics with the optimal observation path, and help is provided for efficient application of the underwater autonomous mobile observation platform in marine science.

The invention discloses a structural risk minimization based weighted least squares power system state estimation method. According to the characteristic that the measurement number of power system state estimation is limited, a structural risk minimization based weighted least squares estimation model is provided through a statistical learning theory, so that the norm of a residue can be minimized, and meanwhile the confidence interval of a state variable can be minimized. Besides, a detailed solving process of the method is provided. The method accords with the structural risk minimization thought in the statistical learning theory, can obtain an estimation result more approximate to the true value of the state variable under limited measurement conditions, and has good engineering application prospects.

The invention discloses a pulse ultra wide band signal detection method based on compressed sensing. The pulse ultra wide band signal detection method based on compressed sensing includes the steps: utilizing random sampling in the compressed sensing theory to sample on received signals at a signal detection end, greatly reducing sampling rate which is needed by detection, and meanwhile arranging a stop condition which conducts part reconfiguration on the emitted pulse ultra wide band signals according to a specific parameter of a pulse ultra wide band emitter. The pulse ultra wide band signal detection method based on compressed sensing effectively reduces calculation cost of a detection method, utilizes a weighting least square estimation method to continuously reduce influence of an abnormal sample which corresponds to a position with a large residual value after the previous iteration, effectively improves rate of convergence of part reconfiguration process, guarantees orthogonality of the residual value and a chosen optimal vector, improves robustness of the detection method, obviously improves detection success rate under the condition of low signal-to-noise ratio, finally conducts detection judgment of signals by directly processing a part reconfiguration value through compressed sampling and simplifies achieving difficulty of the detection process.

The invention discloses a GB-RAR high-speed rail bridge deformation information estimation model considering influence of colored noise. The method includes the steps of obtaining phase values in twoground-based radar images corresponding to the first moment and the second moment, carrying out interference on the ground-based radar images, obtaining interferometric phases, removing an atmosphericphase component and the random noise phase, then obtaining radar sight direction deformation quantity, acquiring amplitudes of white noise and colored noise, and combining sedimentation rate, obtaining a sedimentation time sequence and a random model of a target object, carrying out noise reduction processing on the settlement time sequence by utilizing a wavelet analysis method, analyzing the sedimentation time sequence by adopting a maximum likelihood method, after logarithm taking processing is carried out on the obtained joint probability density expression, adjusting the spectral index of colored noise, after the obtained set likelihood value is processed through a weighted least square estimation method, obtaining a GB-RAR high-speed rail bridge deformation information estimation model, thus the accuracy of railway bridge monitoring and evaluation is improved.

The invention relates to an RSSI (Received Signal Strength Indicator) ranging based WLS (WebLogic Server) node self-positioning method in a wireless sensor network, which comprises the following steps that: firstly, in a positioning process, a channel fading factor n is calculated by weighting; secondly, a distance between an unknown node and a beacon node is measured by RSSI ranging; and finally, a coordinate of the unknown node is estimated by a weighted least squares algorithm (WLS). Based on the same hardware complexity of the positioning algorithm and the least squares (LS) estimation algorithm, the positioning accuracy is greatly enhanced, therefore, the method provided by the invention is simple and easy to implement, and has a stronger practical value.

The invention, which relates to the technical field of passive positioning, discloses a method for estimating a measuring angle error in bearings-only passive positioning. The method includes the following steps: step 1, performing first-order Taylor series expansion on a measurement equation to obtain a bearings-only positioning equation; and step 2, performing loop iteration on the positioning equation to obtain weighted least squares estimation of a target position and thus obtaining an azimuth estimation error. The positioning equation contains a measuring angle error term, so that the estimation accuracy of the measurement angle error can be effectively improved and the effect is obvious especially under the condition of having a large measurement angle error. The measurement angle error obtained by the method can be applied to a weighted or deviation compensation positioning algorithm, so that the positioning accuracy of the algorithm can be effectively improved.

The invention provides an airborne radar real-time Doppler beam sharpening super-resolution imaging method, which belongs to the field of radar detection and imaging. The present invention adopts the fast iterative weighted least squares estimation to realize Doppler beam sharpening (DBS) super-resolution imaging for the echo signal of the current adjacent coherent processing interval (CPI) by means of scanning-while-acquiring-while-processing , and then use the spatial correlation of adjacent CPI azimuth echoes to recursively update the next CPI weighted least squares estimation result, thereby improving the convergence speed and greatly reducing the computational complexity. Compared with the traditional DBS imaging technology, the present invention can not only effectively improve the imaging resolution, but also can reduce the number of iterations, improve the convergence speed, and satisfy the real-time imaging performance while ensuring the imaging resolution.

The invention discloses a moving target positioning method under the condition of clock deviation and clock drift, which constructs a weighted least squares WLS problem including four unknown quantities of the source node's position and speed, clock deviation and clock drift, and then adopts positive semi-definite The relaxation method transforms the non-convex weighted least squares WLS problem into a convex positive semi-definite programming problem, and then uses the interior point method to solve the convex semi-positive definite programming problem to obtain the global optimal solution; In the case of simultaneous existence and unknown, from the perspective of weighted least squares estimation model estimation theory and convex optimization theory, while realizing the position and velocity estimation of the source node, the joint estimation of clock bias and clock drift is realized, and the positioning accuracy High, low computational complexity, suitable for complex deployment environments.

The invention discloses a parameter estimation method for an electric power circuit in T-type connection, which belongs to the technical field of parameter identification and estimation for circuits of power systems. A computer and data acquisition devices at three ends of the circuit in T-type connection are adopted, SCADA (supervisory control and data acquisition) measured data for the circuit in T-type connection under a plurality of load levels at multiple time intervals are imputed by the aid of a program, then a weighted least squares estimation model is solved to obtain estimated values of parameters of the circuit in T-type connection under one load level, namely a sample of estimation values, then a mean value of a plurality of samples of estimation values is calculated, coefficient variations of the samples are used as convergence tolerance of parameter estimation precision, and finally estimation values of the parameters of the circuit in T-type connection are obtained. By the aid of the parameter estimation method, the parameters of the circuit in T-type connection can be effectively estimated, and the parameter estimation method has the advantages of simplicity, quick calculation speed, high estimation precision, good stability of data, high engineering practicality, convenience in popularization and application, and the like. In addition, the parameter estimation method can be widely applied to parameter estimation of optional electric power circuits which are in T-type connection and are provided with data acquisition devices.

The invention provides a two-stage linear weighted least-square power system state estimation method which is characterized by comprising the following steps of: forming a network model, and calculating a node admittance matrix and a branch-node relationship matrix; transforming a measurement vector and a state vector; forming an accurate linear measurement equation; performing the first stage of linear weighted least-square estimation to obtain the estimated value of the transformed state vector; performing inverse transformation, and performing the second stage of linear weighted least-square estimation to obtain the estimated values of the voltage amplitudes and phase angles of all nodes; and identifying bad data. According to the two-stage linear weighted least-square power system state estimation method provided by the invention, more scientific state estimation result can be obtained, the calculation efficiency is higher, and the engineering application prospect is good.