40 results about "Generalized likelihood ratio" patented technology

The invention relates to the technical field of robot navigation, and discloses a multi-sensor fusion navigation method and system of a robot. The method comprises the following steps: collecting point cloud data of a laser radar sensor and motion data of an IMU six-axis sensor; executing generalized likelihood ratio test and sliding window interval search on the point cloud data and the motion data to obtain a zero-speed abnormal interval, and creating a detection threshold value for the current ground environment according to the zero-speed abnormal interval; performing spatial superposition on the position coordinates of the zero-speed abnormal interval and a laser radar scanning blind area to obtain an obstacle distribution map; cleaning control parameters are adjusted based on the detection threshold value, a re-planning cleaning path is generated based on the obstacle distribution map and the cleaning control parameters, the method can actively identify the potential trapped area and re-plan the cleaning path, and the autonomous navigation ability and cleaning efficiency of the intelligent cleaning robot are improved.

The invention relates to an array IMU (Inertial Measurement Unit) fault detection and data fusion method and system based on reinforcement learning, which are characterized in that a generalized likelihood ratio detection threshold is dynamically adjusted according to a real-time environment through an Actor-Critic network, and the problems of false alarm and missing alarm caused by a fixed threshold are solved; meanwhile, the fusion weight is dynamically calculated based on the detection result, the data contribution of the fault sensor is automatically inhibited, and the fusion robustness is improved. Besides, by constructing a reward function based on a pseudo reference value, online adaptive learning under the condition of no real label is realized, the convergence time is shortened, and the requirements of real-time monitoring and high-precision fusion in a complex dynamic scene are met.

The invention discloses a sea target coherent detection method based on knowledge assistance, which considers introduction of prior information of a radar sea clutter covariance matrix, adopts a two-step generalized likelihood ratio test (GLRT) criterion, constructs an adaptive coherent detection framework under a CG-LN clutter background, and improves robustness and effectiveness of sea detection radar target detection.

This invention belongs to the field of anti-spoofing and interference technology for satellite navigation and inertial navigation integrated systems. It provides a GNSS / INS tightly coupled adaptive anti-spoofing method based on generalized likelihood ratio and sequential decision-making, applicable to multi-satellite spoofing interference and micro-ramp spoofing scenarios in autonomous driving. The method includes: first, constructing a GNSS / INS tightly coupled model; second, constructing a GNSS spoofing and interference model; third, constructing a generalized likelihood ratio statistic and designing an iterative isolation strategy, improving detection accuracy through sequential hypothesis testing; and fourth, embedding the detection results into a filter and constructing an adaptive factor to achieve anomaly weighting and prediction adjustment. This method integrates spoofing detection and spoofing suppression, significantly improving the anti-spoofing capability of the navigation system.

The invention discloses a weak cyclostationary signal detection method based on an unmanned aerial vehicle-mounted passive radar network, and the method comprises the steps: firstly building an active radar transmitting signal and unmanned aerial vehicle passive radar receiving signal model, and then building a binary hypothesis model based on cyclostationary generalized likelihood ratio detection through block cyclic matrix approximation; and constructing generalized likelihood ratio test detection statistics and obtaining local detection results of the unmanned aerial vehicles under a colored noise background by using the cyclostationary characteristic of target echoes, finally fusing the local detection results of the unmanned aerial vehicles based on a Bayesian minimum risk criterion at a fusion center, and outputting global detection judgment. And a detection process of cyclostationary generalized likelihood ratio test-Bayesian minimum risk fusion is formed. According to the method, the detection statistics are constructed by using the cyclostationary characteristics of the target echoes, the global detection accuracy and robustness can be improved under the conditions of low signal-to-noise ratio and complex colored noise, and the method is suitable for a multi-unmanned aerial vehicle cooperative passive radar detection scene.

This invention discloses a passive detection method, apparatus, and electronic device in a non-Gaussian noise environment. The method includes constructing an M-dimensional probability density function of the Rayleigh distribution of the received signal; obtaining a first probability density function of the received signal under assumption H1 and a second probability density function under assumption H0; assumption H1 represents the assumption that the received signal includes a radiating source signal; assumption H0 represents the assumption that the received signal does not include a radiating source signal; based on the first and second probability density functions, a generalized likelihood ratio (GLR) test formula is obtained according to the GLR criterion; the derivative of the M-dimensional probability density function under assumption H1 is taken to obtain a maximum likelihood estimate of the product of the channel coefficient and the Doppler frequency shift; the maximum likelihood estimate is substituted into the GLR test formula to obtain a test calculation value; and the test calculation value is compared with a threshold value to achieve passive detection in the absence of a reference signal through a GLR test algorithm based on the test assumption.

The application belongs to the technical field of equipment management, and discloses a nuclear detector fault diagnosis method, device, computer equipment and storage medium. The nuclear detector fault diagnosis method comprises the following steps: acquiring an output signal of the nuclear detector; inputting the output signal into a pre-trained deep belief network for fault identification to obtain a fault parameter corresponding to the output signal; and adopting a generalized likelihood ratio test classification model to perform fault classification on the fault parameter to obtain a fault category corresponding to the output signal, so that the fault of the nuclear detector can be quickly and accurately detected.

This invention discloses a method for detecting multi-channel constant-mode signals without prior electromagnetic interference, belonging to the field of radar signal processing technology. The method includes modeling multi-shot data received by a multi-channel antenna array to obtain high-dimensional multi-channel data; introducing weight vectors to project the high-dimensional multi-channel data onto a single channel for dimensionality reduction, resulting in single-channel weighted data; constructing a generalized likelihood ratio test detector without prior spatial electromagnetic interference information based on the probability density functions under two assumptions: the presence and absence of multi-channel constant-mode signals; and building an optimization model based on the generalized likelihood ratio test detector without prior spatial electromagnetic interference information. A hybrid optimization framework combining alternating optimization and alternating direction multiplier methods is used to solve the optimization model, obtaining the detection parameters and completing the multi-channel constant-mode detection.

This invention discloses a target localization method for multiple-transmitter, multiple-receiver passive multistatic radar under interference conditions. The method includes: acquiring reference channel signals and echo signals and performing signal characterization preprocessing to obtain reference channel signal matrices and echo signal matrices; estimating channel parameters based on the reference channel signal matrices and echo signal matrices using the generalized likelihood ratio principle to construct a DPI-GRLT detector; solving the DPI-GRLT detector and calculating the detection statistics at each grid point to obtain the radar target information estimation result. This invention can maintain high localization accuracy and achieve rapid estimation and resolution of multiple targets even under weak signal conditions. As a target localization method for multiple-transmitter, multiple-receiver passive multistatic radar under interference conditions, this invention can be widely applied in the field of radar target localization technology.

This invention provides a VIO zero-velocity detection and correction method that integrates visual and inertial information. It combines information from both a camera and an IMU (Inertial Measurement Unit) to detect zero-velocity states. For the camera, in a zero-velocity state, the pixel displacement of matching points in adjacent frames determines whether the current state is zero. For the IMU, it combines triaxial acceleration and angular velocity information, applies a generalized likelihood ratio test, and determines whether the current state is zero based on whether the calculated value is less than the likelihood ratio test threshold. When both the camera and IMU information sources indicate that the vehicle is currently in a zero-velocity state, zero-velocity correction is applied: the vehicle is considered to be in a zero-velocity state, a zero-velocity observation is constructed, and applied to an extended Kalman filter for pose update. This invention can more accurately and robustly detect zero-velocity states and effectively corrects the pose estimation drift problem of VIO in a zero-velocity state.

Embodiments of the application disclose a kind of distributed communication signal detection method, device and system based on generalized likelihood ratio, wherein the method is applied to fusion center, comprising the following steps: S1: receiving the observation data transmitted by multiple receiving nodes; S2: under the assumption that radiation source does not exist and radiation source exists, respectively construct the joint conditional probability density function of all observation data of all receiving nodes within the observation time; S3: adopt NP criterion to construct the expression of generalized likelihood ratio test statistic; S4: according to maximum likelihood criterion, estimate the covariance matrix in the joint conditional probability density function under the assumption of two kinds; S5: the estimated value of the covariance matrix under the assumption of two kinds is substituted into the expression of generalized likelihood ratio test statistic, and compared with the decision threshold set according to false alarm probability, to complete distributed communication signal detection. The application can improve the detection probability of communication signal under low signal-to-noise ratio, and has good noise resistance when there is noise fluctuation in different receiving nodes.

The invention discloses an angle measurement method based on maximum likelihood estimation, and the method comprises the steps: constructing a correlation expression of an array echo signal model under the condition of employing a uniform linear array for receiving; establishing two hypotheses of GLRT (generalized likelihood ratio test), including an original hypothesis representing only noise and a standby hypothesis representing signal plus noise; respectively constructing probability density functions; constructing an expression of GLRT statistics in a mode of maximizing an angle value after maximum likelihood estimation is carried out on a probability density function under two assumptions on an unknown parameter; unknown parameters are estimated; the estimated value is substituted into the constructed expression of the GLRT statistical magnitude, and the final expression of the GLRT statistical magnitude is obtained; and obtaining a plurality of preset discrete angle values, and determining the discrete angle value which enables the final representation of the GLRT statistical magnitude to obtain a peak value as an estimated angle, thereby solving the problems of poor adaptability and insufficient precision of a traditional method in a low signal-to-noise ratio and complex electromagnetic environment.

The invention discloses a secure communication method based on STAR-RIS assistance in a communication-sensing integrated network. The method comprises the following steps: analyzing channel characteristics among a base station, an active reconfigurable intelligent surface, a user, a target, a plurality of guards and a fusion center; solving an optimal detection threshold value and a minimum detection error probability based on generalized likelihood ratio test and maximum likelihood estimation under the condition that a guarder knows the communication transmission power of the base station; solving the minimum security communication detection error probability of the fusion center under the hard decision fusion scheme and the soft decision fusion scheme based on the independent detection result of each guard; the transmitting beam forming vector of the base station, the active beam forming vector of the active reconfigurable smart surface and the number of continuous transmission time slots are jointly optimized to maximize the total secure transmission rate at the user and meet the perception performance and security performance constraints at the same time; according to the invention, the contradiction among the sensing performance, the safety performance and the safety transmission efficiency is balanced.

The invention provides a cooperative random encryption industrial internet physical layer secure transmission and detection method, and belongs to the technical field of wireless communication and information security. The multivariate data is received by I local sensors after being transmitted through a wireless channel, and the ith local sensor carries out hard decision by using a maximum likelihood criterion and encodes and converts the multivariate data into a binary signal; randomly selecting a part of data from the sent data to perform flipping encryption, and forwarding the encrypted flipping data by a plurality of relay nodes and then sending the flipping data to a fusion center; and after the data arrives at the fusion center, the fusion center firstly performs flipping decryption on the flipped data before transmission and then decodes a receiving value obtained by each transmission link, and a decoded signal obtains a system judgment metric value based on generalized likelihood ratio inspection and compares the system judgment metric value with a judgment threshold to obtain a final detection result. According to the method, the anti-eavesdropping capability and the transmission reliability are remarkably improved in the wireless sensor network with limited energy, and the method has relatively high engineering applicability and robustness.

The present application relates to the technical field of rail transit contact rail state monitoring and fault early warning, and particularly relates to a contact rail temperature rise infrared online monitoring method and system, which comprises the following steps: firstly, mapping pixels to mileage and collecting multi-band long-wave infrared data, synchronously obtaining traction working conditions and applying phase-locked thermal excitation to generate infrared observation data in the mileage domain; then, correcting system response, calibrating radiation, and implementing phase-sensitive demodulation, and inversely obtaining an absolute temperature field according to a multi-band radiation transmission model; further inversely obtaining an equivalent line heat source according to a heat conduction equation, constructing an expected temperature rise baseline, and generating an alarm object containing a position interval and intensity by combining a generalized likelihood ratio with spatial continuity; finally, returning phase-locked and imaging parameters according to sensitivity and Fisher information. The present application realizes quantitative, locatable, low false alarm temperature rise early warning and adaptive observation.

The present application belongs to but is not limited to the underwater detection technical field, especially relates to a kind of underwater electromagnetic detection system optimum receiver derivation method and system, comprising: S1, target echo signal feature analysis;S2, optimum receiver derivation;S3, receiver operating characteristic analysis;S4, underwater electromagnetic detection system optimum generalized likelihood ratio receiver obtains.The present application first obtains the optimum receiver model of underwater electromagnetic detection system, aiming at the existing reality problem of underwater electromagnetic detection system receiver, the present application uses generalized likelihood ratio criterion to comprehensively deduce the model of optimum receiver, and the performance of receiver is analyzed, which provides certain reference significance for underwater electromagnetic detection system receiver design.

An occupancy detector includes a radar and an occupancy analyzer. The radar is provided to output a radar background signal and a radar detection signal which are time-domain signals. The radar background signal is detected by the radar from an area when a subject is not occupying the area, and the radar detection signal is detected by the radar from the area when the presence of the subject is unknown. The occupancy analyzer is electrically connected to the radar to receive the radar background signal and the radar detection signal, and it is provided to analysis whether a subject is occupying the area based on the radar background signal and the radar detection signal through generalized likelihood ratio test (GLRT).

A method and system for sea surface target recognition based on range-time images, belonging to the field of measurement image recognition and processing, firstly maps the actually observed sea surface RT image into sea surface RT image data with amplitude following a standard Rayleigh distribution. The covariance matrix of the mapped sea clutter data is constructed. A binary hypothesis testing method is used on the mapped sea clutter data to construct a generalized likelihood ratio test statistic. The optimal estimate is substituted into the log-likelihood ratio expression to obtain the standard form of the mapped generalized likelihood ratio detector. The detector is then used to identify whether a target exists in the range cell to be detected. This invention can convert arbitrary non-Gaussian sea surface RT image data into RT image data with amplitude following a standard Rayleigh distribution, thereby achieving unified processing in complex sea clutter environments and effectively improving the accuracy of target recognition under complex non-Gaussian sea clutter backgrounds.

The invention provides a non-zero mean clutter neutron space signal adaptive detection method and system. The method comprises the following steps: firstly, jointly estimating a mean vector and a covariance matrix of clutters by using to-be-detected data and auxiliary data; then, on the basis of the subspace signal model, deriving a detection statistic by utilizing a generalized likelihood ratio criterion; and finally, comparing the detection statistics with a preset threshold to complete target detection. According to the method, the non-zero mean value of the clutter is explicitly estimated and compensated, so that the adaptive detector keeps a false alarm rate characteristic for a clutter covariance matrix and also has a false alarm characteristic for the non-zero mean value of the clutter in a non-zero mean value clutter environment, and the detection probability is remarkably improved; the technical problem that a traditional subspace detector fails in the environment is solved.

The application discloses a kind of artificial noise auxiliary de-cellular perception integrated beamforming methods, comprising: multiple distributed access points AP, user, eavesdropper and a perception target are laid out, and access point AP is centrally controlled by central processing unit.First, AP sends signal and receives target reflection signal, and detection probability is calculated based on generalized likelihood ratio test;With this probability as optimization goal, combined with communication demand and eavesdropping restriction, optimization problem is constructed.By analyzing the structure of the problem, it is decomposed into two sub-problems of beamforming and artificial noise joint optimization, and power allocation.The former is transformed into a convex optimization problem using semidefinite relaxation method;The latter uses bisection method to iteratively optimize power allocation, and updates the power allocation parameter optimization interval at each iteration, thereby maximizing system performance under a given total power.The application can effectively improve the detection probability and improve the perception safety under different power constraints with lower complexity.

This invention relates to a network anomaly traffic detection method based on wavelet transform and generalized likelihood ratio test, belonging to the field of network security technology. This invention uses Daubechie wavelet transform on network traffic data to output a set of multiple and independent wavelet coefficients, enabling them to conform to a bivariate Cauchy-Gaussian mixture statistical model. The generalized likelihood ratio test is then used for verification, thus solving the problem that hypothesis testing cannot be applied to identify anomalies in time-series network traffic data.

The application discloses a data-driven replay attack detection method based on an equivalent space industrial control system, which comprises the following steps: acquiring input and output data of multiple time points from an industrial control system; constructing a historical data matrix and calculating a zero matrix, solving an optimal weight matrix, calculating an equivalent matrix, generating a system residual, designing a chi-square test and a generalized likelihood ratio test, and designing corresponding thresholds; introducing part of model information to enhance the detection capability of the original system; deploying the obtained detection system, acquiring the input and output of the system online, generating a residual, and using the chi-square test and the generalized likelihood ratio test to judge the system operation; when the test value is greater than the residual threshold, it is judged that the system is subjected to a replay attack, and an alarm is issued. The application can detect the replay attack based on the historical data of the equipment operation of the industrial control network through the input and output of the system without depending on the model parameters of the system, meanwhile, the application is easy to deploy and does not sacrifice the control performance of the system.

The application provides a moving platform distributed radar moving target rapid detection method, device and equipment. The moving platform distributed radar moving target rapid detection method comprises the following steps: a moving platform distributed radar system model is constructed, and a low-bit quantized signal model is obtained based on the moving platform distributed radar system model; a generalized likelihood ratio test (GLRT) detector of the moving platform distributed radar system is generated through the low-bit quantized signal model; joint estimation processing is performed on the GLRT detector by using a differential evolution (DE) method and a batch gradient descent (BGDA) method, so that state data is obtained; the state data comprises a radar cross section (RCS), a target position, a target speed, a radar position and a radar speed. In the application, the joint estimation processing is performed by using the DE method and the BGDA method, so that the accuracy and the convergence speed of the moving platform distributed radar system for target parameter estimation in a dynamically changing environment are improved.

The invention discloses an artificial noise-assisted honeycomb-removing and sensing integrated beam forming method, which comprises the following steps that: a plurality of distributed access points (APs), users, eavesdroppers and a sensing target are arranged, and the access points (APs) are controlled by a central processing unit in a centralized manner; firstly, an AP sends a signal and receives a target reflection signal, and the detection probability is calculated based on generalized likelihood ratio test; the probability is taken as an optimization target, and an optimization problem is constructed in combination with a communication demand and eavesdropping limitation. By analyzing the structure of the problem, the problem is decomposed into two sub-problems of beam forming and artificial noise joint optimization and power distribution. The former adopts a semi-definite relaxation method to convert into a convex optimization problem; and the latter iteratively optimizes power distribution by using a dichotomy, and a power distribution parameter optimization interval is updated in each iteration, so that the system performance is maximized under the given total power. According to the method, the detection probability can be effectively improved with low complexity under different power constraints, and the sensing safety is improved.

The application relates to a generalized likelihood ratio test method and device based on Bayesian inference. The method comprises the following steps: acquiring reference channel and monitoring channel signals of a transmitting source and a receiving node, and constructing a binary hypothesis test signal model of target presence / absence; modeling unknown channel response coefficients as random variables with a known prior distribution, and modeling a transmitting signal symbol sequence as a deterministic variable, and constructing a joint probability distribution function of two types of hypotheses; obtaining an edge probability function through integral edge processing, maximizing the edge probability function to obtain a deterministic variable estimation value, constructing a Bayesian generalized likelihood ratio test statistic, and comparing the Bayesian generalized likelihood ratio test statistic with a preset threshold to determine whether the target exists. By using the method, channel characteristics and signal structure prior knowledge can be fully fused, detection robustness and sensitivity in a low signal-to-noise ratio and high dynamic environment can be improved, calculation complexity can be simplified, spatial diversity gain can be efficiently utilized, and quasi-optimal detection under the constraint of a constant false alarm probability can be realized.

The application discloses a kind of based on STAR-RIS auxiliary in the security communication method of integrated network, including the following steps: the channel characteristics between base station, active reconfigurable intelligent surface, user, target, multiple guards and fusion center are analyzed;In the case where the communication transmission power of guard is unknown base station, based on generalized likelihood ratio test and maximum likelihood estimation, the optimal detection threshold and minimum detection error probability are solved;Based on the independent detection result of each guard, the minimum security communication detection error probability of fusion center under hard decision fusion scheme and soft decision fusion scheme is solved;The transmission beamforming vector of base station, the active beamforming vector of active reconfigurable intelligent surface and the number of continuous transmission time slots are jointly optimized to maximize the total security transmission rate at the user, while satisfying the sensing performance and security performance constraints;The application balances the contradiction between sensing performance, security and security transmission efficiency.

The present invention discloses an active and passive coordinated sensing fusion method for a radar communication dual-function system. The method includes: step 1 embedding a dedicated sensing symbol in a signal transmitted by a base station; step 2 solving a power allocation problem using an optimization problem; step 3 having a receiving access point receive a passive sensing signal and process direct path interference into white noise using a whitening filter; step 4 having the receiving access point determine whether a target is present using a generalized likelihood ratio test detector, and then the receiving access point and base station transmit the binary determination result and target detection probability to a fusion center; and step 5 having the fusion center vote and aggregate the received information to determine whether a target is present. The present invention can improve the accuracy and efficiency of target detection, positioning and acquisition of environmental information, and reconstruction, and has the potential to significantly improve the performance of wireless communication sensing networks.

The invention relates to the technical field of track traffic contact rail state monitoring and fault early warning, in particular to a contact rail temperature rise infrared online monitoring method and system.The method comprises the steps that firstly, pixel-to-mileage mapping is established, multi-band long-wave infrared data are collected, a traction working condition is synchronously obtained, phase-locked thermal excitation is applied, and the traction working condition is determined; generating mileage domain infrared observation data; then system response correction and radiation calibration are carried out, phase-sensitive demodulation is carried out, and an absolute temperature field is obtained through joint inversion according to a multi-band radiation transmission model; further inverting an equivalent line heat source according to a heat conduction equation, constructing an expected temperature rise base line, and generating an alarm object containing a position interval and intensity by combining a generalized likelihood ratio with spatial continuity; and finally, phase locking and imaging parameters are transmitted back according to the sensitivity and Fisher information. According to the invention, quantitative, localizable and low-false-alarm temperature rise early warning and adaptive observation are realized.

This invention discloses a method, system, and device for hyperspectral anomaly detection based on a multi-scale grid window, belonging to the field of hyperspectral image processing technology. The method includes: acquiring a hyperspectral image; constructing a spectral gradient tensor to calculate a multi-scale spectral abrupt change intensity map and determining the optimal window size; extracting heterogeneous spectral core regions; filtering pure background regions through spectral clustering and correcting the generalized likelihood ratio statistic to obtain anomaly scores; constructing an attention weight map to generate a multi-scale enhanced anomaly score map and calculating saliency weights for fusion; optimizing spatial consistency through a Markov random field and combining historical data with dynamic thresholds to output the detection results. This invention solves the problems of local contrast calculation being affected by multiple anomalies, background estimation ignoring spectral purity and spatial correlation, and poor detection robustness due to lack of historical time-series data utilization. It achieves accurate hyperspectral anomaly identification from coarse localization to fine detection, and from single-frame analysis to historical verification.

The invention provides a target detection method and system suitable for a reverberation edge environment, and relates to the technical field of active sonar detection. According to the target detection method, the detection model can be generated according to the prior information of the covariance matrix in the reverberation edge environment, the prior information of the covariance matrix is fully utilized, and the estimation precision of the covariance matrix is improved. The estimation value of the detection statistic parameter of the detection model is solved based on the two-step generalized likelihood ratio test criterion, the adaptive detector is obtained, the adaptive detector is used for carrying out adaptive detection on the target in the reverberation edge environment, and the target detection performance in the reverberation edge environment is remarkably improved.