64 results about "Factor graph" patented technology

The application discloses a marine buoy multi-source fusion positioning method and system based on a factor graph optimization, which acquires GNSS observation data, IMU measurement data and marine environment auxiliary data; a factor graph containing state nodes, GNSS position factors, IMU pre-integration factors and marine dynamics constraint factors is constructed, and wave and current theories are used to constrain buoy movement; in view of multipath effects, marine surface reflection geometry and marine root mean square wave height are combined to calculate a multipath weighting factor, and a GNSS covariance matrix is adaptively adjusted; according to IMU data, a sea state level is discriminated, and an edge window length and a trigger interval of incremental smoothing solving are adaptively linked and adjusted; through adjacent buoy ranging information, collaborative constraints are constructed, and based on Mahalanobis distance and chi-square distribution threshold value detection, abnormalities are detected and local reconstruction is performed. The application effectively suppresses marine surface multipath interference, slows down the accumulation of calculation errors during signal interruption, and realizes high-availability continuous positioning under limited computing power.

The application provides a low-density parity-check code decoding method and device based on a graph neural network, which comprises the following steps: step 1, constructing a factor graph comprising variable nodes, check nodes and edge relationships, and mapping log-likelihood ratio information of a received signal to initial embedding of the variable nodes; step 2, generating message features according to node embedding, node degree and iteration step length, and constructing attention weights for each edge to measure the importance of the message; step 3, inputting the message features and the attention weights into a gated recurrent unit to update the residual of the edge weight, and feeding back the updated edge weight to the graph structure; step 4, weighting and aggregating the messages from the adjacent nodes according to the edge weight, calculating the updated embedding of the variable nodes and the check nodes, and performing gated modulation combined with the check result; step 5, repeating steps 2 to 4 until a preset iteration number or a decoding convergence condition is reached, and mapping the final variable node embedding to a decoding result to realize LDPC code word recovery. By introducing the attention mechanism and the gated residual update into the message passing process, the application realizes adaptive modeling of the contribution degree of different edge messages, dynamically remembers the historical state, thereby improving the decoding performance and the convergence speed; meanwhile, the application can effectively reduce the bit error rate and is suitable for high-speed reliable data transmission scenarios in a 5G / 6G wireless communication system.

Techniques associated with generating and maintaining sparse geographic and map data. In some cases, the system may maintain a factor graph comprising a plurality of nodes. In some cases, the nodes may comprise pose data and sensor data associated with an autonomous vehicle at the geographic position represented by the node. The nodes may be linked based on shared trajectories and shared sensor data.

This invention discloses a vehicle cooperative positioning system for traversing continuous blind zones, comprising a module for acquiring multi-source observation data, a module for constructing a spatiotemporally consistent cooperative factor graph, a cooperative constraint optimization module, an adaptive cooperative strategy adjustment module, and a cooperative information re-fusion module. This invention belongs to the field of data processing technology, specifically a vehicle cooperative positioning system for traversing continuous blind zones. This scheme employs observability-driven cooperative constraint construction and spatiotemporally consistent cooperative factor graph modeling, maintaining high observability and stable positioning even in scenarios without satellite signals, achieving unified fusion of multi-source heterogeneous data. It utilizes communication quality-driven adaptive cooperative switching and predictive prior constraints for delay-compensated re-fusion, achieving autonomous positioning without drift during communication interruptions and unified global trajectory calibration after communication recovery, providing high-precision and continuous positioning support for traversing continuous blind zones.

The application relates to the technical field of surveying and mapping geographic information, and discloses a surveying and mapping intelligent analysis system based on multi-source data fusion, which comprises a multi-source acquisition and space-time reference alignment module, a factor graph consistency fusion and gate module and the like. The multi-source acquisition and space-time reference alignment module is configured to lock a main clock source through a hardware timing protocol, set a hard time alignment threshold, receive multi-source sensor data, judge a time synchronization deviation fed back by the sensor, and output original observation data flow when the time synchronization deviation is lower than the hard time alignment threshold. The factor graph consistency fusion and gate module is configured to receive the original observation data flow and build a global factor graph containing a full-source data constraint chain. The hardware timing protocol locking and hard time alignment threshold mechanism are adopted to establish a unified space-time reference at a physical level from the source, eliminate spatial misplacement caused by sensor sampling differences, realize high-precision consistency fusion of multi-source data, and solve the problems of difficulty in determining deviation sources and low fusion efficiency in traditional technologies.

This application discloses an AGV multi-sensor tightly coupled positioning method and system based on adaptive factor graph optimization, relating to the field of AGV navigation technology. By quantitatively evaluating the real-time positioning quality of LiDAR and GNSS, and dynamically adjusting the Gaussian noise covariance matrix and sensor fusion weights, it achieves intelligent switching of the dominant positioning sensor. It can accurately identify anomalies such as satellite signal failure and decreased LiDAR matching accuracy, quickly switching to a stable sensor to avoid positioning jumps, drift, and interruptions. This enhances the anti-interference capability of the positioning system, ensures the continuity of AGV autonomous navigation and operational safety, and facilitates the penetration of AGVs into high-end industrial scenarios, improving their adaptability and operational efficiency in flexible manufacturing, intelligent warehousing, and other scenarios. It effectively solves the technical problem of existing multi-sensor fusion positioning using fixed weights and being unable to dynamically adapt to environmental changes, significantly improving the positioning performance of AGVs under complex working conditions.

The present application relates to the technical field of navigation positioning, and in particular to a multi-source sensor adaptive weight fusion positioning method and related equipment, which comprises the following steps: collecting multi-source positioning sensor observation data, calculating corresponding health degree factors according to the data characteristics of each sensor, which are used to represent the data quality at the current time; inputting the health degree factors into a pre-trained reinforcement learning model, the model taking the maximization of the reward function constructed by the positioning error and the sensor health degree as the goal, and outputting the fusion weight of each sensor. According to the output weight, the multi-source observation data is fused and solved through factor graph optimization, and the final positioning result is output. This method can dynamically adapt to the differences in sensor working conditions and environmental disturbances, suppress the influence of abnormal data, realize the collaborative improvement of positioning accuracy and system reliability, and is suitable for multi-source fusion positioning scenes such as automatic driving, robots, intelligent navigation, etc.

This invention discloses a multi-UAV dual-grid relative navigation method and device based on factor graph optimization, belonging to the field of multi-robot cooperative localization technology. Existing relative navigation technologies are affected by state modeling accuracy and measurement nonlinearity, making accurate relative position estimation difficult in challenging GNSS environments. This method first proposes a new relative position calculation method, based on which a dual absolute navigation state model is proposed, improving the accuracy of relative navigation state modeling. Then, IMU pre-integration is improved by considering gravity variations, and pre-integration of the error state is proposed. Based on this, a multi-UAV dual-grid relative navigation framework based on factor graph optimization is proposed, significantly reducing the impact of measurement nonlinearity on relative position estimation. The navigation method proposed in this invention can significantly reduce the impact of state modeling accuracy and measurement nonlinearity on relative position estimation, improving the accuracy of relative position estimation, and has practical engineering application significance.

This invention addresses the need for autonomous navigation of low-cost, high-spin munitions by providing a factor graph-based navigation method for dynamically optimizing trajectories and inertia under elliptical geometric constraints. This method uses an ideal trajectory as a benchmark, constructs elliptical geometric constraints, and generates multiple candidate trajectories that satisfy these constraints, forming a candidate trajectory set. By evaluating the similarity between measured inertial data (gyroscope and accelerometer) and the inertial data derived from each candidate trajectory, a trajectory sequence that meets threshold requirements is dynamically selected. Using the optimized trajectory and the inertial pre-integration results as constraint factors, a factor graph-based navigation scheme integrating the optimized trajectory and inertial data is designed to achieve high-precision autonomous positioning of high-spin munitions under satellite denial scenarios. This invention solves the technical challenges of rapid error accumulation and significant decrease in navigation accuracy inherent in pure inertial navigation schemes under satellite denial conditions.

Embodiments of the present disclosure disclose a SPP positioning model construction method and device based on factor graph optimization. The specific implementation of the method comprises: constructing a variable node set of a factor graph structure, wherein the to-be-estimated state quantity of each variable node in the variable node set at an epoch includes three-dimensional coordinates of a receiver and four parameters related to clock bias of the receiver; constructing a pseudorange factor set according to the variable node set, wherein the pseudorange factor set includes at least four pseudorange factors; correlating the position parameters between two adjacent epochs in each epoch in an observation window to construct a Doppler velocity factor, obtaining a Doppler velocity factor set, wherein the Doppler velocity factor is a binary factor; constructing a graph optimization cost function according to the pseudorange factor set and the Doppler velocity factor set, and obtaining a SPP positioning model. The implementation can improve the GNSS positioning accuracy in a complex environment.

The application provides an adaptive UWB-SLAM loose coupling fusion method based on graph optimization, comprising: S1: a space-time synchronization module acquires a mobile robot pose estimation value output by a front-end SLAM system and original measurement data generated by communication between a UWB anchor and a UWB tag, unifies the time and coordinate system of the pose estimation value and the original measurement data to a world coordinate system, realizes time and coordinate system alignment, and obtains a coordinate system transformation matrix; S2: an online statistical quality evaluator module acquires UWB distance measurement values, signal strength indication values and precision estimation factors between a UWB tag and a plurality of UWB anchors deployed in an environment, obtains a comprehensive quality score, generates a dynamic information matrix based on the comprehensive quality score; S3: a factor graph optimization module weights UWB distance measurement values in the original measurement values by using the dynamic information matrix, and outputs a robot global path, a current real-time pose of a mobile robot and a UWB anchor coordinate.

The application relates to a GMM-assisted vehicle fusion positioning method and system, which comprises the following steps: calculating an IMU pre-integration term according to the angular velocity and acceleration measurement information of an IMU; calculating a dynamics pre-integration term according to the speed measurement information of a wheel speed sensor and the angular velocity measurement information of the IMU in combination with a two-degree-of-freedom vehicle model; constructing an IMU factor and a dynamics factor by using the IMU pre-integration term and the dynamics pre-integration term after obtaining a GNSS measurement signal; constructing a pseudorange factor according to the original observation information of a GNSS receiver in combination with system states; constructing a clock drift factor based on the clock error of the GNSS receiver; constructing a factor graph by combining the constructed factors, wherein the noise of the IMU factor, the dynamics factor and the clock drift factor is Gaussian modeling, and the noise of the pseudorange factor is GMM modeling; and optimizing the factor graph to estimate the positioning information of the vehicle. Compared with the prior art, the application can effectively suppress the influence of abnormal GNSS measurement on positioning, and realizes low-cost and robust high-precision positioning.

The application relates to the technical field of navigation positioning of pipe network inspection equipment, and particularly discloses a high-precision positioning system for pipe network inspection equipment based on multi-sensor fusion. In view of the problem that no GNSS signal exists in the pipe network and long-distance inspection leads to serious cumulative drift of positioning, the system integrates IMU, a wheeled odometer and a three-dimensional laser radar, realizes space-time synchronization at the hardware level through a preprocessing module, uses an ESKF fusion model and IMU pre-integration for high-frequency pose estimation at the front end, corrects global errors through a factor graph optimization algorithm in combination with key node features and a global topological prior map at the back end, and introduces measurement weight adaptive adjustment logic to cope with environmental degradation. The application effectively solves the problem that the positioning accuracy is difficult to maintain for a long time in the complex environment of the underground pipe network, and significantly improves the reliability and trajectory consistency of the inspection operation.

The application discloses a decoding method and device, electronic equipment and a non-transitory computer readable storage medium, and belongs to the technical field of communication. The method comprises the following steps: determining the activation probability of each user based on at least one sparse code division multiple access signal received in a current time period and the codebook of each user in a communication network; the activation probability represents the probability of a user sending a non-all-zero code word in the current time period; determining each activated user from each user in the communication network according to the activation probability of each user; constructing an initial factor graph based on the activated user and the subcarriers occupied by the non-zero elements in the corresponding codebook; and performing decoding processing on each sparse code division multiple access signal based on a message passing algorithm, the initial factor graph and the codebook of each user, to obtain the sending code word of each activated user, so that data detection can be performed only on the activated user; when the activation probability of a user is low, the user is directly considered as a non-activated user sending an all-zero code word, and the calculation complexity of data detection can be reduced.

The invention discloses an unmanned aerial vehicle cluster collaborative navigation method realized based on a factor graph, and relates to the technical field of unmanned aerial vehicle collaborative navigation. According to the method, a factor graph comprising variable nodes and factor nodes is constructed for each unmanned aerial vehicle, the variable nodes comprise the position, the speed, the attitude and the IMU zero offset of the unmanned aerial vehicle, and the factor nodes comprise an inertial measurement pre-integration factor, a visual measurement factor and a radio distance measurement and angle measurement factor. By initializing a factor graph, collecting sensor data in real time to update factors, and adopting an L-M algorithm based on Lie group and a sliding window to optimize the factor graph, the unmanned aerial vehicles communicate and share pose information to realize collaborative optimization. According to the method, the problems of high calculation complexity, low information fusion efficiency, poor expansibility and the like of cooperative navigation of the existing unmanned aerial vehicle cluster in a complex environment are effectively solved, cooperative navigation between the unmanned aerial vehicle clusters is realized through a factor graph mode, and the method has reliable navigation performance.

This disclosure provides a motion planning method, apparatus, computer device, and storage medium for robots, applied to each robot in a robot swarm; each robot maintains its own local factor graph; the local factor graphs maintained by the multiple robots constitute a global factor graph; the method includes: in response to a state update event corresponding to the current robot being triggered, determining a trigger domain, and determining a replanning subnet from the current global factor graph based on the trigger domain; performing message passing processing based on Gaussian belief propagation in the replanning subnet with the goal of reducing the energy change of the local factor graph corresponding to the current robot, to obtain an updated local factor graph corresponding to the current robot; determining the confidence distribution information of the variables corresponding to multiple variable nodes in the local factor graph of the current robot based on the updated local factor graph of the current robot; and determining the motion replanning result of the current robot based on the confidence distribution information.

This invention discloses an intelligent analysis method for bridge inspection using an unmanned aerial vehicle (UAV) remote sensing platform. To address the problem of high-precision mapping and re-inspectionable localization of bridge defects from pixel coordinates, and to suppress localization drift under occlusion and multipath conditions caused by the Global Navigation Satellite System (GNSS), this invention collects and time-aligns image sequences, inertial measurement data, and GNSS observations. Based on satellite geometry and signal quality, it performs multipath evaluation, outputting reliability scores and observation biases, and corrects these biases. A factor graph containing inertial, visual, and GNSS constraints is constructed, and a switch variable is introduced for the GNSS constraints. Prior values ​​are set based on the reliability score to adaptively adjust weights. Nonlinear optimization is performed to obtain the pose sequence and covariance. Furthermore, a transformer matching method is used to generate closed-loop constraints. A joint decision is made combining matching confidence, pose covariance, and reliability score, and weighted additions are added to the factor graph for update optimization. Defect detection is performed on the images to obtain defect pixel coordinates, and back-projection is used to intersect the bridge component model to obtain component identification and 3D position. A re-inspection localization range is generated based on covariance propagation. This achieves the technical effect of high-precision localization of bridge defects and output of a re-inspectionable range.

This invention discloses a multimodal tightly coupled synchronous localization and mapping method and system with resistance to dynamic interference. The method simultaneously acquires images, point clouds, and inertial data. The visual front end combines a feature extraction network and a target detection model, using an adaptive extended Kalman filter to track dynamic targets and generate a mask for removing key points in dynamic regions. Dynamic feature points in these regions are removed, and the remaining static key points are used to complete inter-frame matching and relative pose estimation, forming visual odometry constraints. The laser front end uses inertial pre-integration to distort the point cloud and calculates laser odometry based on geometric feature registration. Loop closure detection performs candidate frame retrieval based on geometric descriptors and combines geometric consistency checks to generate loop closure pose constraints. The back end constructs a global factor graph and integrates visual odometry constraints, laser odometry constraints, inertial constraints, and loop closure pose constraints for joint nonlinear optimization. This invention effectively suppresses dynamic environmental interference and significantly improves positioning accuracy and robustness.

The application discloses a marine buoy multi-source fusion positioning method and system based on a factor graph optimization, which acquires GNSS observation data, IMU measurement data and marine environment auxiliary data; a factor graph containing state nodes, GNSS position factors, IMU pre-integration factors and marine dynamics constraint factors is constructed, and wave and current theories are used to constrain buoy movement; in view of multipath effects, marine surface reflection geometry and marine root mean square wave height are combined to calculate a multipath weighting factor, and a GNSS covariance matrix is adaptively adjusted; according to IMU data, a sea state level is discriminated, and an edge window length and a trigger interval of incremental smoothing solving are adaptively linked and adjusted; through adjacent buoy ranging information, collaborative constraints are constructed, and based on Mahalanobis distance and chi-square distribution threshold value detection, abnormalities are detected and local reconstruction is performed. The application effectively suppresses marine surface multipath interference, slows down the accumulation of calculation errors during signal interruption, and realizes high-availability continuous positioning under limited computing power.

This invention relates to a multi-sensor fusion localization and mapping method and system based on single-anchor-point extrinsic constraints, belonging to the field of robotic multi-sensor fusion SLAM technology. It aims to solve problems such as poor scale and heading observability in degraded environments, strong susceptibility of UWB to non-line-of-sight interference, and easy drift during optimization when using single-anchor-point UWB / LiDAR / IMU fusion. The invention constructs a joint optimization vector and a sliding window factor graph, assesses ranging reliability through geometric visibility and channel quality, and adaptively incorporates UWB factors. It utilizes the eigenvalues ​​of the information matrix to evaluate observability, and employs compensation strategies such as ranging filtering and weight adjustment during degradation. The output pose is iteratively optimized and stitched together to generate a 3D map. The system includes modules for data synchronization, reliability assessment, factor graph optimization, degradation compensation, and mapping output. This invention achieves stable localization and mapping with only a single anchor point, suppresses drift and divergence, is suitable for complex scenarios where GNSS fails, and features low deployment cost, strong adaptability, and high accuracy.

The application provides a visual inertial SLAM method and system based on visual feature points and feature lines, and the method comprises the following steps: acquiring camera image data and inertial measurement unit (IMU) data, performing feature extraction on the image data to obtain feature points and feature lines, performing pre-integration processing on the IMU data to obtain IMU pre-integration constraints between adjacent key frames; constructing a tight coupling optimization framework based on a factor graph: taking feature point re-projection errors, feature line re-projection errors and IMU pre-integration residuals as edges of the factor graph, and taking key frame poses and map features as nodes of the factor graph; performing nonlinear optimization on the factor graph by using a sliding window method, and minimizing an objective function to obtain optimized key frame poses. The method of the application can still maintain stable positioning in a weak texture environment, and effectively avoids system failure caused by insufficient features through complementary point and line features.

Various aspects of the present disclosure relate to techniques for low-complexity message passing. An apparatus is configured to create a precoding matrix for enabling a plurality of combinations of streams of multiplexed layers at a plurality of receiving antennas, wherein the plurality combinations of streams are associated with a plurality of user equipment UEs; transmit the precoding matrix to the plurality of UEs associated with the multiplexed layers; detect a plurality of signals at the plurality of receiving antennas using a factor graph-based detection algorithm, wherein the plurality of signals are encoded using the precoding matrix; and decode the plurality of signals based at least in part on the precoding matrix.

This invention discloses a method and system for intelligent matching of preferred products based on user profiles, specifically relating to the field of e-commerce data processing technology. The method includes obtaining user account rights transaction records, filtering transaction lines based on payment completion markers, using the receiving field as the object primary key, forming an occupancy interval based on the start and end times of rights, and attaching the redemption time and purchase price to output a successful purchase table; merging the successful purchase tables by object primary key and forming an observation sequence based on transaction time, forming a preference continuation factor based on the primary key, a demand occupancy factor based on the intersection of intervals, and a price response factor based on the price difference following the purchase result, outputting a factor graph; extracting the object primary key, occupancy interval, purchase price, and redemption feedback from successful transaction records, constructing a factor graph, and separating the preference retention state and demand extinction state using a factor hidden Markov model algorithm, and then generating a purchase suppression set and a purchase strategy sequence using a particle learning algorithm and a strategy price determination process.

This invention discloses a method and system for predicting coal mine gas concentration based on a dynamic factor graph network, belonging to the field of coal mine safety monitoring technology. Addressing the problem that existing technologies cannot explicitly and dynamically model the nonlinear coupling relationships of multiple factors, leading to low prediction accuracy, poor robustness, and insufficient interpretability of gas concentration, this invention constructs a coal mine gas concentration prediction network: the front end explicitly captures the dynamic coupling relationships between multiple factors and outputs high-order feature sequences by embedding the original data of each factor in the dynamic factor graph network, gated dynamic message passing, and a GRU state update mechanism; the back end learns the temporal dependencies through GRU and finally outputs the predicted value. The overall system can adaptively adapt to the evolution of factor coupling relationships during mining, significantly improving the model's adaptability to complex systems, and enhancing the model's learning efficiency and prediction accuracy; the entire process is automated with a fast response speed, meeting the needs of real-time safety monitoring in coal mines and possessing significant engineering application value.

The application discloses a tunnel vehicle positioning method based on preset target and factor graph optimization, and belongs to the technical field of tunnel vehicle positioning. The application pre-arranges targets with global coordinates and geometric structures in a tunnel; a vehicle acquires point clouds in real time by using a laser radar, identifies visible targets by two-stage matching algorithms, and calculates the poses of the targets in a radar coordinate system; the absolute correction pose of the vehicle is calculated in combination with the global coordinates of the targets; and the pose is taken as an absolute observation factor and integrated into a factor graph SLAM optimization framework, and is optimized in combination with a laser interframe motion constraint factor, so that a high-precision vehicle pose trajectory is output. The application effectively suppresses long-distance cumulative drift, and meets the centimeter-level positioning requirements of unmanned mine trucks.

This application belongs to the field of multi-target state estimation technology. This application provides a method for joint optimization of the states of multiple passive sensors and multiple targets based on a two-factor graph. The embodiments of this disclosure first construct a passive detection multi-target measurement solution layer to obtain coarse positioning points of multiple targets as state measurements; then, a joint optimization framework is constructed containing a multi-target management layer, a multi-target state optimization layer, and a multi-sensor state optimization layer. The non-Gaussian distribution characteristics of the fused measurements and the non-Gaussian distribution characteristics of the fused measurement reprojection error are modeled using a three-dimensional Gaussian mixture model and a two-dimensional Gaussian mixture model, respectively. Then, a two-factor graph is established based on the three-dimensional and two-dimensional Gaussian mixture models; finally, the sensor attitude state and target state are jointly optimized and estimated using sliding window nonlinear least squares, and the optimization results are fed back to the multi-target management layer to form a closed-loop update.

The application relates to the technical field of computer vision, in particular to a three-dimensional positioning method and system of array elements, wherein the method comprises the following steps: acquiring a multi-view high-bit-depth gray image frame sequence of an array element to be measured; completing dynamic range conversion from high-bit-depth to low-bit-depth through adaptive bit-depth mapping enhancement; performing primary detection on the low-bit-depth image to obtain a candidate coded visual marker and a corresponding ROI region; performing local mapping enhancement on each ROI region to obtain an enhanced ROI image; performing secondary precision detection on the enhanced ROI image to output a target coded visual marker containing a target corner point and a single-frame pose; determining a corner point observation covariance according to the ROI dynamic range; constructing an incremental factor graph taking the single-frame pose as a state quantity and taking a re-projection error as an observation constraint, establishing an adaptive noise-weighted multi-frame joint optimization target; and performing optimization solving based on the optimization target to obtain three-dimensional coordinates of the array element in a world coordinate system. Through the application, the precision, consistency and real-time performance of three-dimensional positioning of the array element are improved.

This application provides a message scheduling method, device, and program product for multi-view heterogeneous sensor data. It relates to the field of communication control technology. The method includes: receiving multi-view heterogeneous sensor data from an autonomous driving device and determining the source category of the sensor data; constructing a factor graph model of the sensor data based on the source category; representing each target state, each sensor observation, and the relationship between each target state and each sensor observation as nodes and edges in the factor graph based on the factor graph model, thereby constructing a Bayesian inference framework; and processing the received multi-view heterogeneous sensor data based on a sequential processing strategy. In this solution, a sequential / parallel switching scheduling strategy based on dependencies, combined with the factor graph model and the Bayesian inference framework, solves the non-convex optimization and convergence dilemmas of existing technologies.