18 results about "Algorithmic complexity" patented technology

The invention discloses a blockchain zero-knowledge proof collaborative credit risk control data privacy protection query method, and belongs to the technical field of financial science and technology and information security. According to the method, a four-layer architecture of a feature importance analysis layer, a zero-knowledge proof verification layer, a block chain intelligent contract layer and an audit log layer is constructed, so that the contradiction between privacy protection and data availability in a traditional credit risk control system is solved. According to the system, an SHAP value is adopted to calculate feature importance, a zk-SNARKs verification circuit is deployed for high-sensitivity features, zk-STARKs lightweight verification is adopted for low-sensitivity features, and differential privacy protection is achieved. And dynamically adjusting the complexity of a verification algorithm according to the credit risk score of the user, and realizing dynamic authority control based on a verification result through an intelligent contract. The verification metadata is recorded to the block chain to form a tamper-resistant auditing log, and it is ensured that the query process is transparent and traceable. According to the method, the real-time risk control requirement is met while the privacy of the user is protected, and a compliant and efficient data query solution is provided for financial institutions.

The invention relates to the technical field of power electronics and battery testing, discloses a charging and discharging control method and system for a multi-channel program-controlled battery simulator, and aims to solve the problems of response lag, high prediction algorithm complexity, timing alignment failure and poor dynamic adaptability in the prior art. The method comprises the following steps: synchronously acquiring current and voltage data with microsecond-level timestamps through hardware; predicting future current by using a lightweight gating cycle unit and transient attention mechanism hybrid model; and generating a feed-forward compensation quantity based on an inverse power level model, combining a PID feedback quantity, and dynamically fusing the feed-forward compensation quantity and the PID feedback quantity according to a transient activation factor to generate a final PWM instruction. The system comprises a synchronous acquisition module, a dynamic prediction module, a feed-forward compensation module, a closed-loop feedback module and a dynamic fusion module, and supports multi-channel independent control. According to the invention, the ultra-low delay response to the hundred-ampere / millisecond step load is realized, and the current tracking precision and the system stability are remarkably improved.

The application discloses a satellite slant beam inter-beam interference avoidance method and system based on cost flow and dynamic graph coloring, and accurately models the interference range of slant beams in a multi-satellite satellite hop-beam system, considers the influence of the slant beams when modeling the anti-beam inter-beam interference hop-beam pattern design problem, divides the dynamic graph coloring problem into two sub-problems of satellite-wave position matching and hop-beam time slot allocation which are coupled with each other, pre-solves the satellite-wave position matching problem by using a minimum cost maximum flow algorithm, obtains a satellite-wave position matching initial solution which maximizes the sum of satellite elevation angles of all wave positions in the system under the constraint of the maximum number of simultaneously opened beams of the satellite, and on the basis, combines a tabu search algorithm and a graph coloring algorithm to jointly optimize the two sub-problems, so as to further improve the inter-beam interference avoidance performance of the same hop-beam time slot opened beam. The application can obtain a multi-satellite hop-beam scheme with minimum inter-beam interference, and the algorithm has low complexity and is convenient for actual system implementation.

The disclosure relates to a method and system for extracting disparity information out of a pair of stereo camera images for processing images of camera sensors for Automated Driving or Advanced Driver Assistance Systems in a vehicle. The method includes: a) receiving a pair of unrectified stereo camera images, b) performing a 2D-matching of the received pair of images using a trained convolutional neural network and providing the 2D-displacement on a pixel-by-pixel level as disparity information, and c) outputting the 2D-displacement on pixel level. The method is robust which does not require a high precision rectification. Applying rectification on the disparity values after the matching process leads to reduced algorithmic complexity with simultaneous preservation of all information.

The present application provides a kind of port structure deformation monitoring method and system, belong to port safety technical field.Utilize GNSS sensor to the position information of port terminal is monitored to obtain the spatial coordinate information of port terminal;According to wharf horizontal coordinate information and wharf vertical coordinate information in spatial coordinate information, based on the first prediction method and the second prediction method, the lateral displacement and settlement displacement prediction of port terminal is carried out;Finally, based on lateral displacement prediction curve and settlement displacement prediction curve, port structure deformation early warning is carried out.The monitoring method of the present application only needs to utilize the spatial coordinate information of port terminal, and the required input data amount is small;Different prediction methods are used to predict the lateral displacement and settlement displacement of port terminal, which can take into account the relationship between algorithm complexity and prediction accuracy;When setting the prediction coefficient of the prediction method, the fluctuation of the data observation value is considered, and the stability and prediction sensitivity of the prediction method are considered.

This invention relates to a ground identification method and apparatus, mobile tools, and related products. The method includes: performing ring-direction alignment processing on a single frame of multi-ring point cloud transmitted by a multi-line lidar to generate a multi-ring aligned point cloud; for each ring aligned point cloud, forming data elements from each data point in the ring aligned point cloud and its forward, backward, and left points and storing them in a queue, and sequentially calculating the connectivity relationships of the data points in each data element; generating a connectivity graph of the ring aligned point cloud based on the connectivity relationships of the data points in each data element; determining suspected ground points from the multi-ring aligned point cloud based on the connectivity graph corresponding to the multi-ring aligned point cloud; fitting a ground equation based on the suspected ground points; and determining ground points from the multi-ring aligned point cloud based on the ground equation. This invention reduces the algorithmic complexity of ground identification in point clouds and improves computational efficiency by designing the connectivity graph of each data point in the multi-ring point cloud.

The present application relates to a kind of semi-supervised learning image defogging method based on physical imaging model, belong to computer vision and image processing technical field.The method includes: S1: design the light weight dense feature fusion network with attention mechanism, using feature extraction network sharing mechanism, using adaptive weight distribution module to estimate physical model parameters from degraded image;S2: atmospheric scattering model describes the imaging process of haze degraded image, proposes an improved physical model with adaptive information compensation, calculates and restores clear image;S3: design semi-supervised learning strategy, enhance the defogging performance of network in actual haze scene, weaken the domain difference between synthetic domain and real domain.The present application can solve the problem of single image defogging recovery with light algorithm complexity, and the superiority of the present application is proved by experiment verification in real data set.

The application discloses a kind of intelligent vehicle trajectory tracking robust control method and system, method includes the following steps: the vehicle model of target vehicle is simplified as bicycle model, establishes two degrees of freedom vehicle model;Based on two degrees of freedom vehicle model, establish vehicle trajectory tracking model;Using orthogonal decomposition method and state feedback method, establish the feedback dissipative Hamilton system of vehicle trajectory tracking model;According to the feedback dissipative Hamilton system, combined with Hamilton robust controller, design generalized Hamilton robust controller.The Hamilton robust controller designed in the application is compared with sliding mode controller and LQR controller, greatly reduces the algorithm complexity, improves the calculation efficiency, reduces the error of control vehicle tracking reference trajectory.

The application provides a path planning cooperative control method for ship-borne heterogeneous strong coupling unmanned equipment, relates to the technical field of intelligent control and path planning, and comprises the following steps: considering the longitudinal, transverse and yaw motion of an unmanned tractor and an unmanned carrier aircraft on the deck plane of an aircraft carrier, a three-degree-of-freedom coupling motion mathematical model of the heterogeneous unmanned equipment is established; a virtual force field is constructed according to an improved artificial potential field method, a field strength adjusting operator and a heading disturbance injection term are introduced, and discrete path points are generated; the discrete path points are converted into continuous expected trajectories by using a Gaussian smoothing function; precise trajectory tracking of the unmanned equipment is realized through an adaptive neural network double-layer control architecture, the upper layer controls the unmanned carrier aircraft, the lower layer controls the unmanned tractor, and cooperative control is realized by using the coupling relationship. The application introduces a first-order filter (dynamic surface control technology), effectively suppresses the differential explosion phenomenon, reduces the algorithm complexity, and improves the engineering applicability of the controller.

The present invention discloses a method and an apparatus for 2D regularized planar projection of a point cloud. The method includes: obtaining original point cloud data; initializing a planar structure of 2D projection of the point cloud; calculating horizontal azimuth information of the point cloud based on the original point cloud data; and determining a mapping relationship between the original point cloud data and the planar structure of 2D projection based on the horizontal azimuth information, to obtain a planar structure of 2D regularized projection of the point cloud. In the present invention, a large-scale point cloud may be projected to a 2D regularized planar structure without 2D local search. Therefore, complexity of an algorithm can be reduced, time spent on 2D regularized planar projection of the point cloud can be reduced, and algorithm performance can be improved.

A repository replica management method. The method comprises: receiving a backup command, which backup command is used for backing up a primary repository of a user, wherein the primary repository comprises at least one piece of object data and reference data of the at least one piece of object data, and the reference data is used for pointing to the at least one piece of object data; and in response to the backup command, compressing the at least one piece of object data in the primary repository to generate a compressed file, and storing replicas of the compressed file in a replica storage medium corresponding to the primary repository. In the method, replicas are created only for object data that occupies a large storage space and has a low update frequency in a repository, and reference data that occupies a small storage space and is frequently updated remains shared, such that an object-level repository replica of the repository is implemented. A synchronization and consistency algorithm for the object-level repository replica is simple, and thus both the high algorithmic complexity associated with strong consistency algorithms used to resolve repository divergence due to concurrent writes and the risk of repository corruption are avoided.

The invention provides a rotor position estimation method of a high-order sliding-mode observer based on a variable reaching law, and the method comprises the steps: obtaining an alpha-axis stator voltage, an alpha-axis stator current, a beta-axis stator voltage and a beta-axis stator current of a permanent magnet synchronous motor in an alpha-beta coordinate system; the obtained data are input into the preset high-order sliding-mode observer with the variable reaching law, the rotor position is estimated through fusion of the variable reaching law and the high-order sliding-mode observer, and rotor position estimation can be achieved under the condition that the rotor position does not need to be measured through a sensor; and the hardware cost required by the rotor position determination scheme is effectively saved. The high-order sliding-mode observer based on the variable reaching law under the alpha-beta coordinate system achieves estimation of the rotating speed and the rotor position, the number of times of coordinate change can be reduced, the algorithm complexity is reduced, a low-pass filter is avoided, and the dynamic performance and the observation precision are improved.

The invention discloses a data intelligent application system based on large model driving, which comprises a multi-source data processing module, a large scene application module, a small and medium scene application module, an optimization feedback regulation module and an intelligent management information output module, relates to the technical field of data analysis, and solves the problems of accurate classification of small and medium scenes and large scenes and high efficiency. According to the method, a lightweight tool and a simplified model are adopted for small and medium-sized scenes, the data analysis cost is reduced, and the large-scale scenes are subjected to real-time streaming through the cooperation of a distributed computing framework and a real-time streaming engine, so that the real-time streaming capability of the large-scale scenes is improved, and the real-time streaming capability of the large-scale scenes is improved. The real-time requirements of complex scenes such as smart cities and industrial internet are met, the processing efficiency is improved, feedback is collected from multiple dimensions of a data end, a service end and a model end, data hysteresis and tool parameters are corrected for medium and small scenes, flow engine delay and algorithm complexity are optimized for large scenes, and the long-term precision retention rate of the model is improved.

The present application relates to the technical field of optical fiber communication, and particularly relates to a GRU neural network nonlinear equalization method based on a regression coupling value, which first reshapes the data structure of a received signal sequence and matrixes it; then in the training stage, a nonlinear damage error factor based on a nonlinear Schrodinger equation is used to guide network updating, and in the application stage, a signal is input into a GRU neural network, and a signal nonlinear damage compensation result is obtained through the propagation algorithm of the network. The method of the present application uses a GRU neural network algorithm based on a regression coupling value to realize the capture and compensation of signal nonlinear damage, solves the problems of high algorithm complexity, multiple iteration numbers and limited nonlinear compensation capacity in digital back propagation and learning type equalization algorithms, further improves the effectiveness and practicability of the nonlinear equalization algorithm, and has an important application prospect in the field of digital signal processing related to optical communication.

The application provides a 2.5D global positioning method, system, equipment and medium for exhibition visitors, relates to the technical fields of meta universe and artificial intelligence, and comprises the following steps: S1, a zero Z-axis plane of a global coordinate system is arranged in a positioning domain, global grid points are divided for the whole scene, and the global grid points are projected into each camera to obtain image grid points of a video image; S2, a nearest neighbor point is acquired according to the image grid points; and S3, the mean value of the neighbor points acquired by a plurality of cameras is taken as the global position of the target. The application can adjust the positioning accuracy and avoid the algorithm complexity caused by three-dimensional reconstruction.

This invention proposes an AI-based intelligent retrieval and recommendation system for nursing literature, comprising a literature retrieval module, a nursing staff profiling module, a knowledge recommendation module, and a real-time update module. The literature retrieval module is configured to analyze nursing staff needs and retrieve relevant nursing literature. The nursing staff profiling module includes a profiling model, which calculates a weighted vector representing the distribution of nursing staff interests based on their search history, reading preferences, and other behavioral data. The knowledge recommendation module includes a recommendation model that uses a collaborative filtering algorithm with regularization to combine the nursing staff profiling vector with the retrieved nursing literature to generate personalized recommendation results. The real-time update module updates the nursing literature database using web crawling technology and updates the profiling model and recommendation model in conjunction with it. This system improves the accuracy of nursing literature retrieval, personalization, and timeliness of recommendations. It features low algorithm complexity, high recommendation accuracy, and efficient real-time adjustment and recommendation optimization capabilities.

The invention provides a path planning cooperative control method for ship-borne heterogeneous strong-coupling unmanned equipment, and relates to the technical field of intelligent control and path planning, and the method comprises the following steps: considering the longitudinal, transverse and yawing motions of an unmanned tractor and an unmanned ship-borne aircraft on the plane of an aircraft carrier deck; establishing a three-degree-of-freedom coupling motion mathematical model of the heterogeneous unmanned equipment; constructing a virtual force field according to an improved artificial potential field method, introducing a field intensity adjustment operator and a course disturbance injection item, and generating discrete path points; converting the discrete path points into a continuous expected trajectory by using a Gaussian smoothing function; accurate trajectory tracking of unmanned equipment is realized through a self-adaptive neural network double-layer control architecture, an upper layer controls an unmanned shipboard aircraft, a lower layer controls an unmanned tractor, and cooperative control is realized by using a coupling relationship. According to the invention, a first-order filter (dynamic surface control technology) is introduced, the differential explosion phenomenon is effectively inhibited, the algorithm complexity is reduced, and the engineering applicability of the controller is improved.

The application discloses a low-orbit satellite network multi-service distribution method based on SRv6, solves the problem that the flexibility of service distribution is low when the prior art needs to weigh and decide multiple service requirements, and the problem that the algorithm complexity is increased due to the mutual conflict or intersection between constraint conditions.The implementation steps of the application are as follows: 1, constructing a multi-level low-orbit satellite network topology structure; 2, calculating the SRv6 SID and Segment List forwarding path of the low-orbit satellite node; 3, calculating the shortest node path; 4, designing a service tunnel; 5, distributing the SRv6 TE Policy; and 6, generating a low-orbit satellite network routing table.The low-orbit satellite network generated by the application realizes a more flexible multi-service distribution mode and can fully meet the reliability requirement of the low-orbit satellite network routing.