62 results about "Confidence factor" patented technology

The invention relates to the technical field of image recognition, in particular to a fire behavior recognition method and system based on video monitoring. The technical problem that the early warning reliability of a video monitoring system on fire smoke is insufficient is solved. The method comprises the following steps: acquiring a monitoring video stream of a monitoring area and time sequence data acquired by a sensor in the monitoring area; determining an evaluation value of the to-be-detected area based on the image feature and the morphological change of the to-be-detected area; determining a confidence factor based on the form change trend of the to-be-detected region and the change trend of the time series data; determining a smoke authenticity value based on the evaluation value and the confidence factor; and adjusting an alarm threshold value of the smoke alarm based on the smoke authenticity value. The method is used for fire smoke early warning and monitoring scenes.

The invention belongs to the technical field of power distribution automation of a power system, and particularly relates to a self-adaptive load control method and system for an intelligent cable branch box, and the method comprises the steps: collecting and preprocessing the current data of a branch loop and the temperature data of a cable joint in real time, obtaining a data confidence factor, and extracting a current change rate and a temperature rise rate; obtaining a load prediction value in a future set time period through a load trend prediction model; calculating the residual heat tolerance time for reaching the limit tolerance temperature; calculating an output load control adjustment coefficient based on the adaptive load control decision model and a particle swarm optimization algorithm; and judging a current state interval according to the load control adjustment coefficient, and controlling an execution end to execute a grading response strategy of through-flow maintenance, short-time overload monitoring and early warning and breaking current limiting in combination with the residual heat tolerance time. According to the method, the residual heat tolerance time is calculated in combination with load prediction and a transient thermal circuit model, and hierarchical response of load control is realized through a particle swarm optimized radial basis function neural network decision.

The invention discloses an underground safety level intelligent evaluation response method and system fusing fuzzy credibility weighting and Bayesian reasoning, and relates to the technical field of underground safety monitoring. According to the method, a multi-source heterogeneous sensor is deployed to collect data, a fuzzy membership matrix is obtained through fuzzy membership function normalization, a credibility factor matrix is constructed by combining time sequence stability, spatial neighborhood consistency and node long-term health weight, and the two are subjected to weighted fusion to obtain a fusion index vector. And inputting the Bayesian network fused with the credible nodes, reasoning to obtain security level posterior probability distribution, and triggering a linkage response. The system correspondingly comprises a plurality of modules for implementing the steps. According to the method, the dynamic evaluation and credibility quantification of the data quality of the underground multi-source sensor are realized, and an intelligent evaluation mechanism for deeply integrating the credibility of the data source into a reasoning structure is constructed, so that the risk evaluation robustness and accuracy are improved, the closed-loop management from evaluation to response is realized, and the underground safety is guaranteed.

The invention discloses a freight logistics track abnormity processing method and system, and belongs to the technical field of intelligent logistics monitoring, and the method comprises the steps: obtaining a track original data set of a target freight vehicle; constructing a behavior evolution model based on the historical trajectory data, wherein the behavior evolution model is used for describing conventional operation boundaries of the vehicle under different road types and time periods; performing multi-source fusion correction on the trajectory data to generate a trajectory point sequence; calculating credibility factors of the track points based on GPS stability, historical track goodness of fit and road matching degree; constructing an anomaly judgment graph, and generating an anomaly distribution function by adopting a recursive scoring mechanism; identifying a track abnormal fragment based on a dynamic adaptive threshold value and outputting an abnormal label; generating a scheduling suggestion according to the abnormal label and a vehicle scheduling rule, and pushing the scheduling suggestion to a vehicle-mounted system in real time; according to the invention, individuation, high precision and real-time performance of track abnormity identification are realized, and the method is suitable for an intelligent logistics management and dynamic scheduling system in a complex scene.

The invention relates to the technical field of intelligent power grid abnormity diagnosis, in particular to an intelligent electric energy meter operation data abnormity diagnosis method and system. The method comprises the steps of establishing a sliding window, collecting magnetic induction intensity and load current data, maintaining references of the magnetic induction intensity and the load current data according to historical data, calculating the change direction and intensity of a sampling point relative to the references, screening data with enhanced magnetic induction intensity and reduced load current, and carrying out weighted covariance calculation by utilizing an effective mask. Constructing a negative correlation response coupling index; constructing an abnormal definite diagnosis confidence reasoning model; calculating a causal confidence factor based on the coupling index; calculating an aging confidence factor based on the duration of an abnormal state; according to the scheme, industrial interference and electricity stealing attacks can be accurately distinguished, hidden attacks are captured, the false alarm rate is reduced, and diagnosis blind areas are eliminated.

The invention discloses a natural road feature-based vehicle-mounted radar external parameter calibration method and device, a storage medium and a program product. The method comprises the steps of obtaining synchronous radar point cloud data and visual image data; determining a visual lane line according to the visual image data, and determining a radar lane line according to the radar point cloud data; lane line transverse deviation is calculated according to the visual lane line and the radar lane line, and deviation confidence is calculated according to the credibility factor and the factor weight; determining a deviation estimated value based on the effective deviation and the deviation confidence by adopting a one-dimensional Kalman filter; the deviation estimation value is evaluated according to preset calibration threshold values, the preset calibration threshold values comprise the minimum calibration threshold value and the maximum calibration threshold value, when the deviation estimation value is larger than the minimum calibration threshold value and the maximum calibration threshold value, the radar coordinates are calibrated based on the deviation estimation value, and when the deviation estimation value is larger than the maximum calibration threshold value, an early warning mode is started. And outputting early warning prompt information in response to the early warning mode, and recording a fault code.

The invention relates to the technical field of robot navigation, in particular to a robust visual inertial navigation method for dynamic and static factor decoupling in a high dynamic environment, and the method comprises the following steps: obtaining prior pose information and visual feature data of an inertial measurement unit from a navigation system; the method comprises the following steps: generating a stable confidence factor and a stable inertia factor of a visual feature based on priori pose information and visual feature data of an inertial measurement unit, and generating a double-factor residual error at the same time; based on prior pose information and visual feature data of the inertial measurement unit, effective dynamic features are identified, and a dynamic effect residual error is generated; and in combination with the basic residual error, the two-factor residual error and the dynamic effect residual error of the navigation system, decoupling dynamic and static residual errors and carrying out joint optimization to obtain an optimized navigation track. According to the method, the robustness of the visual inertial navigation system in a complex environment can be improved through dynamic and static factor decoupling and dynamic compensation constraint, and scene specific training is not needed.

The invention relates to the technical field of main power distribution network monitoring, in particular to a main power distribution network integrated dispatching monitoring method and system, and the system achieves main power distribution network integrated dispatching monitoring through the method steps of main network and distribution network data acquisition, data preprocessing, data fusion, main power distribution network fusion topology model establishment and the like. When the operation data of the main power distribution network is fused, the traditional Bayesian fusion algorithm is improved by using a confidence factor, so that the heterogeneous data sources can be more effectively integrated, and more accurate information is provided for the operation and maintenance of the main power distribution network; the fused state estimation value reflects the most probable value of the state parameter of the main power distribution network under the condition that all observation data and priori knowledge are given, the information of all data sources is integrated, and the reliability of different data sources is constrained through confidence factors, so that the state estimation value is more accurate and reliable.

The application discloses a method for determining the control range of bolt connection pre-tightening force, which comprises the following steps: firstly, establishing a finite element model of bolt connection under the condition of applying pre-tightening force; calculating the maximum equivalent stress of the connection area under different pre-tightening force conditions; constructing a radial basis function proxy model based on the pre-tightening force and the equivalent stress; considering the uncertainty of the pre-tightening force, obtaining a plurality of groups of sample points of the pre-tightening force, and calculating the corresponding equivalent stress through the radial basis function proxy model; calculating the reliability confidence factor of the equivalent stress, and taking the bolt pre-tightening force when the reliability confidence factor is greater than 1 as the control applied pre-tightening force range. Starting from the essence of the uncertainty of the pre-tightening force, the uncertainty of the pre-tightening force is described through the mean value and the multiple relationship of the pre-tightening force, the safe working range of the structure is determined according to the reliability confidence factor of the equivalent stress, the pre-tightening force control range is obtained, the traditional pre-tightening force control is expanded from a single deterministic value to a range interval, and the reliability requirement of the structure in the range is ensured.

Disclosed is a system (100) and a method (300) for determining a confidence factor to evaluate network performance. The method comprises initiating, by a test unit of a user device (110), a request for performing a network speed test. The method further comprises receiving, by a receiving unit of the user device (110) from a network, values of a plurality of network performance metrics based on results of the network speed test. The method further comprises determining, by a determining unit of the user device (110), a confidence factor based on the values of the plurality of network performance metrics.

The embodiment of the invention provides a risk assessment method and device for a bridge strut, equipment and a storage medium. Under the condition that the bridge supporting columns comprise a plurality of single supporting columns, first point cloud data, visible light images and positioning data corresponding to each single supporting column are obtained; constructing a strut three-dimensional model of each single strut according to the data; then determining a defect severity factor, a deterioration development trend factor and a detection confidence factor of each single pillar according to the pillar three-dimensional model; determining a first risk level of each single pillar according to the factors; and finally, determining a second risk level of the bridge strut according to the first risk levels of all the single struts. According to the method, the problems that the assessment dimension is single and the assessment result is not accurate enough when risk assessment is carried out on the bridge strut only depending on two-dimensional image data are effectively solved, and more comprehensive assessment on the risk of the bridge strut can be realized through multi-dimensional data.

The invention belongs to the field of information security, particularly relates to an identity authentication method and system, electronic equipment and a storage medium, and aims to solve the problem that balance among security, flexibility and accuracy is difficult to obtain due to lack of comprehensive analysis capability for dynamic behaviors of a user in the prior art. The method comprises the following steps: acquiring user authentication data; uploading the user authentication data to a preset graph database, and representing a corresponding relationship between the account information and the authentication factor by using the graph database; performing matching verification on the account information according to a high confidence factor in the graph database, and determining and distributing a first joint ID; performing cross validation on the account information which is not allocated with the first joint ID according to the middle and low confidence factors in the graph database, and determining and allocating a second joint ID; and uploading the first joint ID and the second joint ID to a preset ID library. The method is higher in accuracy and effectiveness when dealing with fuzzy matching or incomplete information.

This invention belongs to the field of image processing technology, specifically relating to an automatic measurement method and system for material quantity in a silo based on machine vision. The method includes: acquiring depth and infrared intensity images of the silo; preprocessing these images to generate a valid binary mask; determining a geometric confidence factor based on the deviation between the depth gradient and the natural angle of repose of the material being measured, as well as local curvature; and determining an environmental confidence factor by combining infrared intensity and measurement distance; repairing invalid pixels using anisotropic diffusion; predicting depth values ​​based on first-order Taylor expansion and multi-dimensional confidence weighting; converting the repaired depth image into physical height and projected area; and obtaining the total volume of material in the silo through infinitesimal element accumulation. This invention solves the problem of data loss in complex environments and improves measurement accuracy.

The invention discloses a function handling method, device and equipment based on a life cycle and a medium, and relates to the technical field of artificial intelligence. The method comprises the steps of obtaining a confidence factor set corresponding to a to-be-selected evaluation function in a current function evaluation service; performing quantitative judgment on the confidence factor set based on the target low-efficiency affirmation model to obtain a quantitative judgment result, and determining a corresponding intermediate evaluation function based on a preset function category field and the quantitative judgment result; generating a qualitative affirmation starting instruction based on the intermediate evaluation function, and determining a qualitative affirmation result corresponding to the current function evaluation service based on the qualitative affirmation starting instruction and a preset qualitative affirmation rule; and performing function disposal on the intermediate evaluation function based on the qualitative affirmation result, and generating a function disposal result corresponding to the to-be-selected evaluation function. According to the technical scheme of the invention, the method can achieve the accurate recognition and disposal of the low-efficiency function, improves the recognition efficiency and accuracy of the low-efficiency function, and reduces the maintenance and operation cost of a system.

A method includes acquiring a set of images from a plurality of cameras in a monitored environment, detecting a first entity based on the set of images, and determining a first set of locations of the first entity based on locations of the first entity in the set of images. The method also includes acquiring sensor measurements from a plurality of sensors and determining a second set of locations of the first entity based on the sensor measurements. The method also includes determining whether the first set of locations should be associated with the second set of locations based on a confidence factor, and, in response to determining that the first and second set of locations should be associated, determining a sequence of locations of the first entity through the monitored environment.

A multi-source localization method and system for quadruped robots based on geometric spatial decoupling and full-dimensional dynamic confidence is disclosed. The method involves: acquiring data and inputting it into a data preprocessing module to obtain distortion-free effective point cloud data; inputting the effective point cloud data into a geometric information matrix construction module to construct a geometric information matrix; based on the geometric information matrix, a full-dimensional spatial degradation decoupling and perception module performs subspace decoupling, algebraic stiffness calculation, and temporal inertial smoothing to form a geometric confidence factor in the [0,1] interval; inputting the geometric confidence factor into a multi-source covariance elastic reconstruction module to dynamically reconstruct the observation noise covariance of the lidar and leg constraints according to the geometric confidence factor, obtaining the reconstructed lidar observation constraints and leg contact observation constraints; inputting the reconstructed observation constraints into an error state Kalman filter to output the high-precision pose of the quadruped robot. This invention enables precise localization of quadruped robots.

The invention relates to the technical field of fan control, in particular to a speed self-adaptive adjusting method and system for an indoor fan, and the method comprises the steps that the thermal response hysteresis degree of each temperature control area is obtained; according to the temperature difference between any temperature control area and the adjacent area, boundary heat flux representing the heat invasion intensity of the adjacent area to the temperature control area is calculated in real time; calculating an uncertainty index for representing the deviation degree of the Q-Learning control model strategy; constructing a confidence factor, performing weighted correction on the basic learning rate of the Q-Learning control model by using the confidence factor to obtain an effective learning rate at the current moment, and updating the Q value table by using the effective learning rate; and determining the rotating speed of the target fan according to the updated Q value table. By means of the technical scheme, it can be ensured that the temperature of the whole plant is accurate and constant, and meanwhile the risks of frequent starting and stopping of a draught fan and too large air speed adjustment are effectively avoided.

The invention discloses a water use tail end abnormal state monitoring and dynamic service authorization method and system. The method comprises the steps that high-frequency fluid signals of a pipe network are collected in real time and preprocessed; extracting a physical feature vector, and constructing a two-dimensional phase space according to the flow velocity change rate and the pressure change rate; identifying a hysteresis loop formed in the two-dimensional phase space by a single water consumption event, and calculating the topological variation degree of the hysteresis loop; based on historical interaction data of the user, extracting a behavior feature vector; global model updating parameters are generated and distributed to all the terminals, and end-cloud collaborative federated learning type model optimization is achieved; constructing a behavior subspace composed of historical legal behavior feature vectors of the user, and projecting a current physical feature vector to the behavior subspace to obtain a feature deviation degree; and based on the quasi-non-Newtonian rheological dynamics evolution model, dynamically updating a confidence factor according to the feature deviation degree, and self-executing a corresponding hierarchical service authorization control strategy according to the predefined risk level of the confidence factor.

This invention provides an intelligent handwritten mathematical formula recognition and correction method, system, and medium, relating to the field of text recognition technology. The method includes: acquiring raw image data of handwritten mathematical formulas; preprocessing to obtain standardized image data; determining a structural feature matrix; identifying formula error types and patterns based on the changing trends of formula grammatical topological relationships and grammatical logic deviation coefficients over multiple recognition cycles; configuring correction confidence factors; determining the formula error level; performing graded correction feedback; and outputting a standard formula. It also drives a voice broadcast unit to provide synchronous voice correction prompts. This invention solves the technical problems of instability in handwritten mathematical formula recognition, difficulty in guaranteeing recognition and correction accuracy, and resulting in erroneous or inconsistent recognition results. It achieves a complete closed loop from image acquisition to intelligent correction of handwritten mathematical formulas, thereby improving the accuracy, stability, and correction precision of handwritten mathematical formula recognition.

The invention provides an industrial coal blending optimization method and system based on multi-agent collaboration and a program product, and relates to the technical field of industrial intelligent control and data processing. The method comprises the following steps: acquiring production state data of a current time step, wherein the data contains coal inventory, physicochemical indexes and raw material resource weight coefficients; inputting the data into a mechanism calculation module, and calculating a reference ratio vector by using a linear programming algorithm with the purpose of minimizing a comprehensive resource consumption index; meanwhile, inputting the data into a strategy correction agent, and outputting a correction action vector aiming at a reference ratio and a confidence factor of a current working condition; and carrying out dynamic weighted fusion on the reference ratio and the correction action. A double-layer coupling framework of mechanism guarantee and AI efficiency improvement is adopted, a confidence factor is used as a safety valve, on the premise that production safety is guaranteed, the problems that a traditional coal blending model is poor in non-linear adaptive capacity, lagged in feedback and the like are solved, and global optimization of industrial raw material configuration is achieved.

The application belongs to the technical field of navigation, and proposes an adaptive robust filtering navigation method based on multivariate t distribution and Bayesian shrinkage. First, multi-source heterogeneous sensor observation data is obtained. Second, the multi-source heterogeneous sensor data is modeled based on multivariate t distribution, and the innovation of the multi-source heterogeneous sensor observation data is extracted. Then, a continuous confidence factor is constructed, the fault detection is converted into a Bayesian inference problem, and the confidence diagnosis of the observation innovation is performed based on the confidence factor. Then, the innovation discount factor is constructed based on the confidence factor, and the weighted fusion target function is constructed based on the observation fitting term and the virtual robust term, so as to obtain the equivalent innovation observation, the equivalent observation noise covariance matrix, the equivalent innovation covariance and the robust Kalman gain. Finally, the observation information weight is dynamically adjusted, the equivalent parameters are used to complete the robust filtering iteration, and the optimal navigation solution is obtained. The application can improve the navigation positioning precision and robustness in the multi-source fusion navigation process.

The application discloses an electricity-energy-carbon multi-source space-time data fusion method, system, device and medium, and belongs to the technical field of energy data management, and comprises the following steps: acquiring multi-source space-time data, identifying and extracting key event points, and generating an initial event response sequence; performing credibility screening on data segments corresponding to the key event points of each data source to obtain a high-credibility data segment set; analyzing the collaborative relationship among electricity, energy, carbon and economic multi-source activities, and extracting characteristic segments representing system collaborative behaviors; identifying system collaborative abnormal time periods; and fusing the collaborative behavior characteristic segments and the identification results of the collaborative abnormal time periods to output a multi-source space-time fusion data set. The application directly solves the defect of insufficient data credibility of the prior art by calculating multi-dimensional confidence factor of the data segments of each data source at the key event points, such as data acquisition accuracy, time synchronization and consistency with historical rules, screening and eliminating low-confidence data segments.

The invention discloses a vehicle stability control multi-domain safety fusion method and system based on double neural networks, and the method comprises the steps: obtaining a vehicle operation state and local perception data, and preprocessing external cooperation data to construct an external environment information set; calculating the curvature by using multiple independent sensing modes and calculating normalized consistency deviation characteristics; identifying physical dynamic parameters, and adaptively calculating a comprehensive stability risk value by combining a road surface excitation confidence factor; constructing a double-neural-network cascade topological structure and a constraint coupling path; inputting the multi-source heterogeneous joint feature space into a first neural network to solve a fusion safety coefficient; and introducing the fusion safety coefficient into a variable weight loss function of the second neural network to form a structural constraint, dynamically resolving a PID gain and generating a target control instruction. According to the invention, deep fusion of four domains of function security, expected function security, information security and active security is realized, and the problem of multi-source risk nonlinear coupling and adaptive control is solved.

The invention relates to the technical field of data processing, and discloses a multi-step wind power prediction method based on heterogeneous seq2seq and NWP refinement, and the method comprises the steps: collecting high-frequency inertia and low-frequency meteorological sequences, and carrying out the interpolation alignment; calculating a dynamic time warping path to extract an original slope and a global alignment cost; generating a confidence factor based on the path linearity to elastically correct the slope; constructing a heterogeneous neural network to respectively extract high and low frequency hidden state vectors; calculating a virtual mapping index by using the correction slope, and extracting a meteorological channel context according to the virtual mapping index; and generating a dynamic gating coefficient in combination with the alignment cost, weighting the meteorological features and fusing the meteorological features with the inertial features, and outputting a predicted value. According to the method, by correcting the NWP time sequence deviation and introducing position constraint attention, the problem of nonlinear drift of heterogeneous data is solved, and the precision and robustness of multi-step wind power prediction are remarkably improved.

The present application belongs to the technical field of power system distribution automation, and particularly relates to a self-adaptive load control method and system for an intelligent cable branch box. The method comprises real-time acquisition and preprocessing of current data of a branch circuit and temperature data at a cable joint, obtaining of a data confidence factor, and extraction of a current change rate and a temperature rise rate. A load trend prediction model is used to obtain a load prediction value in a future set time period. A residual thermal tolerance time to reach a limit tolerance temperature is calculated. Based on a self-adaptive load control decision model and a particle swarm optimization algorithm, a load control adjustment coefficient is calculated. According to the load control adjustment coefficient, a current state interval is determined, and in combination with the residual thermal tolerance time, a hierarchical response strategy of maintaining through-flow, short-time overload monitoring and early warning, and branch current limiting is executed by an execution end. The present application combines load prediction with a transient thermal circuit model to calculate the residual thermal tolerance time, and uses a particle swarm optimized radial basis function neural network decision to realize hierarchical response of load control.

This disclosure discloses a method, device, storage medium, and program product for calibrating extrinsic parameters of vehicle-mounted radar based on natural road features. The method includes: acquiring synchronous radar point cloud data and visual image data; determining visual lane lines based on the visual image data and radar lane lines based on the radar point cloud data; calculating the lateral deviation of the lane lines based on the visual lane lines and radar lane lines, and calculating the deviation confidence level based on a confidence factor and factor weights; determining a deviation estimate based on the effective deviation and the deviation confidence level using a one-dimensional Kalman filter; evaluating the deviation estimate based on preset calibration thresholds, which include a minimum calibration threshold and a maximum calibration threshold; calibrating the radar coordinates based on the deviation estimate when the deviation estimate is greater than both the minimum and maximum calibration thresholds; and activating a warning mode when the deviation estimate is greater than the maximum calibration threshold, outputting a warning message and recording a fault code in response to the warning mode.

The invention provides a data association method and device, electronic equipment and a storage medium, and the method comprises the steps: when a data association request is detected, based on a preset spatio-temporal data index and a preset feature data index, extracting to-be-associated historical data from the historical data of each data monitoring device of a plurality of data monitoring devices; for each data monitoring device, determining a confidence coefficient corresponding to the to-be-associated historical data of the data monitoring device based on the current emission site working condition and the confidence factor corresponding to the to-be-associated historical data of the data monitoring device; and determining a data association strategy based on the current launching site working condition and the confidence coefficient corresponding to the to-be-associated historical data of each data monitoring device, and associating the to-be-associated historical data of all the data monitoring devices according to the data association strategy. By adopting the technical scheme provided by the invention, efficient and credible fusion of multi-source heterogeneous data is realized.

The invention discloses an automobile key point detection method based on intelligent identification and structure correlation optimization, and belongs to the field of computer vision. The invention aims to provide an accurate detection method for complex key points of a vehicle, and the method carries out the intelligent recognition of the key points of the vehicle through a key point detection and image matching hybrid model: firstly, carrying out the detection and matching of the whole and part of the vehicle through a deep learning network, and carrying out the screening of a result through combining a confidence factor, the precision of the detection algorithm is effectively improved; secondly, a matching method based on deep learning is applied, feature extraction of key points can be efficiently carried out, detection and matching and top view and side view combination optimization are carried out in sequence, then the detection efficiency of the key points is improved, and the precision of the key points can be guaranteed under the condition that images are complex; meanwhile, automobile key point information optimization is performed by applying structure association, and a model for performing key point optimization by using a vehicle contour and an intersection structure is established in combination with a traditional image detection method. Finally, all key point information is corrected, the problem of complex key point detection deviation is solved, and the key point detection precision is improved. The method can accurately detect the automobile key points, and is mainly applied to the fields of vehicle structure design, vehicle detection, intelligent transportation and the like.

The invention discloses a multi-modal data fusion method, system and device applied to charging evaluation, which are used for improving the health evaluation precision of a charging module. The multi-modal data fusion method comprises the steps of collecting multi-modal original data of a charging module; processing the multi-modal original data to construct a deep representation tensor; performing forward semantic association mapping on the deep representation tensor based on a bidirectional cross attention mechanism to generate a matrix in a first direction; performing reverse semantic backflow inference in a bidirectional cross attention mechanism to generate a matrix in a second direction; performing modulation fusion on the matrixes in the first direction and the second direction according to the modal confidence factor to obtain an inter-modal correlation intensity matrix; and weighting the deep representation tensor based on the inter-modal correlation intensity matrix, performing aggregation by using a preset cross-modal sparse projection network to obtain a low-dimensional fusion semantic representation, and inputting the low-dimensional fusion semantic representation to a preset charging health assessment model for assessment.

The invention discloses a fusion positioning method and system based on a laser radar ICP confidence factor. The method comprises the steps of obtaining point cloud data output by an AGV vehicle-mounted laser radar and map information of a scene map where the AGV vehicle-mounted laser radar is located; based on the point cloud data and the map information, point cloud registration is completed through an ICP algorithm to determine an ICP pose, and an ICP confidence factor is calculated according to a matching error, convergence characteristics, a characteristic overlapping degree and point cloud quality; adjusting an observation covariance matrix based on the ICP confidence factor; based on the AGV pose, the ICP pose and the observation covariance matrix, constructing an observation noise model, and outputting fused pose information; the AGV poses comprise the AGV pose of the wheel type odometer and the AGV pose of the inertial measurement unit. According to the method, a dynamic mapping mechanism from the ICP confidence factor to the observation covariance is introduced into the EKF local fusion framework, mismatching and trajectory jump are inhibited under the working conditions of low overlapping / shielding / noise fluctuation and the like, and the convergence speed and precision are improved under the scene of full geometric constraint.