175 results about "Path network" patented technology

The invention provides a multi-path cooperation method, device and equipment for a cluster quadruped inspection robot, and the method comprises the steps: obtaining gait dynamics data and topographic data, and recognizing a topographic adaptive rhythm based on the spectrum decomposition of a foot end contact force sequence; converting the topographic data into a potential energy gradient field to construct a dynamic potential energy diagram, and generating a natural path network; a cooperative trigger point is set based on the stable support moment, a path ripple field is formed, and a multi-machine cooperative potential well is constructed; decomposing the path network into a dominant task chain and an auxiliary task chain, and weaving to form an initial collaborative path; rigid nodes and flexible sections are identified through topological transformation, and random disturbance is injected to reconstruct a self-adaptive collaborative path; gait phase information is extracted to construct a synchronous network, a potential energy flow vector is used for phase alignment to generate a group resonance frequency, and a resonance cooperation path is formed; and finally, an execution instruction is generated based on the resonance cooperation path and the terrain adaptability rhythm, and multi-path cooperation of the cluster quadruped inspection robot in the complex terrain environment is achieved.

The invention belongs to the technical field of hyperspectral image unsupervised, and discloses a hyperspectral image self-supervised clustering method based on a local-global double-branch network, which comprises the following steps: (1) constructing a 3D pixel block; (2) acquiring shallow depth features through a ResNet module; (3) injecting shallow layer features into a dual-path network module; (4) calculating the similarity between the local features and the clustering center to obtain local semantic probability distribution, and then obtaining target distribution; (5) the global features are sent to a feedforward neural network to obtain the probability of each pixel, and global semantic probability distribution is obtained; and (6) constructing a network loss function through a dual self-supervision mechanism to guide the updating of the whole network model. According to the method, the local detail features and the global semantic features of the hyperspectral image are mined by introducing the dual-path network architecture, so that the recognition precision of the hyperspectral ground features is effectively improved.

The invention discloses a crack segmentation method based on condition debugging and a hierarchical dual-path network, and relates to the technical field of computer vision, and the method comprises the steps: 1, collecting and marking a crack image, and generating a training data set; 2, constructing a condition debugging and hierarchical dual-path network model, and training by using the training data set to obtain a crack segmentation model; the conditional debugging and hierarchical dual-path network model comprises a ResNet50 branch network, a CMT branch network and a double-flow cascade segmentation header network, the CMT branch network comprises a conditional modulation gating convolution module, and the double-flow cascade segmentation header network comprises a hierarchical adaptive feature modulation module; and 3, acquiring a to-be-detected image, and inputting the to-be-detected image into the crack segmentation model to obtain a crack segmentation result. According to the method, global context and local feature extraction are effectively combined, and the most advanced accuracy and robustness are achieved in various crack scenes.

The invention discloses a digital twinning-based flow machine operation intelligent collaborative management system, and the system comprises an operation data collection module which is used for obtaining multi-source data in a target scene and carrying out the preprocessing of the multi-source data; the twinning modeling module is used for constructing a flow machine operation digital twinning model and generating a twinning operation initial situation; the space-time path analysis module is used for constructing a time hierarchical path network and generating a path conflict cone set and a feasible space-time path set; the operation load calculation and scheduling module is used for establishing task link representation, generating a storage yard congestion potential field and forming a multi-stream-machine cooperative scheduling candidate scheme; and the virtual-real correction module is used for issuing execution and collecting operation feedback, correcting the twin model and generating an optimized target collaborative scheduling scheme. According to the method, the space-time-behavior-link coupled digital twinborn model is constructed to realize flow machine operation predictive collaborative scheduling, and the method has the advantages of conflict recognition in advance, congestion trend suppression and remarkable improvement of scheduling efficiency.

The invention discloses a surface mine transportation real-time intelligent scheduling method and system based on multi-objective optimization. The method comprises the steps that S1, an initial data loading module initializes the system; s2, the algorithm module verifies the vehicle state after receiving the scheduling request; s3, executing a path generation strategy according to states, wherein a normal vehicle generates all feasible routes based on the global path network; when a user appoints a starting point and a terminal point, a point-to-point optimal route is intelligently generated; s4, performing dynamic optimization scheduling based on the multi-source data; optimal task allocation is solved through a multi-objective optimization model, and a vehicle and route mapping relation is output; and S5, a scheduling result is issued to the vehicle-mounted terminal through the service layer, and a terminal interface displays a recommended route and operation guidance. According to the invention, a dynamic event-driven mechanism and a multi-source data collaborative optimization intelligent scheduling framework are fused, so that efficient distribution of vehicle tasks, real-time generation of paths and quick response to abnormal conditions are realized.

The invention relates to the technical field of logistics collaborative optimization, in particular to an urban logistics distribution range collaborative optimization method and system, and the method comprises the following steps: obtaining a path network node and order coordinate time sequence, constructing a path mapping number, extracting a direction vector and an included angle difference value, and recognizing a path sudden change node and a density change node. Dividing boundaries, calculating density ratio differences, screening unbalanced sections, extracting direction and distance abrupt change nodes, reconstructing a path structure and outputting a boundary graph. According to the method, the corresponding relation between the order and the path node is enhanced by constructing the path mapping number sequence, the adaptive capacity of the boundary to the path trend is improved, the density ratio analysis is combined with the spatial distribution difference to identify the non-uniform region, the responsiveness of the boundary to the order aggregation situation is enhanced, and the boundary node is jointly judged through the included angle and the centroid distance. And target judgment and structural stability control of boundary adjustment are realized, connection coherence is kept in path structure reconstruction, and a boundary path shape with dynamic adaptation capability is output.

The invention provides a low-altitude networking of intelligence multipath data transmission system based on MPQUIC, and belongs to the technical field of novel network protocols, and the system comprises a network state analysis module which is used for obtaining the network state indexes of all communication paths of an unmanned aerial vehicle under different position coordinates; the flow scheduling module is used for predicting the track of the unmanned aerial vehicle, scoring the communication paths by combining the network state indexes of the communication paths, and obtaining the communication paths corresponding to the data flows with different priorities of the unmanned aerial vehicle according to the priorities of the data flows with the scores of the communication paths; and the path selection optimization module is used for dynamically adjusting weight parameters in communication path scores based on deep reinforcement learning by taking network throughput maximization and time delay minimization as targets to obtain optimal communication paths corresponding to data streams with different priorities. According to the invention, the problem that the prior art is difficult to have high reliability and low time delay due to the dynamic change of the network environment caused by the position change of the unmanned aerial vehicle in the low-altitude Internet of Things is solved.

Devices, systems, methods, and processes for enabling path monitoring in multi-path networks are described herein. Load balancing of packets is often performed based on outer encapsulating headers and inner packet headers. However, inner packet headers of probe and data packets can be different. Thus, the probe packet can traverse a different path as compared to the data packet, causing inconsistent hashing. To address this, a device is provided with a routing logic that encapsulates a packet, having a packet header, with an outer header. The outer header has a control bit, source and destination address fields, and a flow label field. When a value of the control bit is set to a predetermined value, load balancing is performed exclusively based on the outer header, exempting a utilization of the packet header. Therefore, load balancing for probe and data packets performed exclusively based on the outer header is consistent.

The invention discloses an unmanned aerial vehicle route planning method based on a graph neural network, and belongs to the field of unmanned aerial vehicle route planning. According to the method, nodes and edges are generated through random sampling in a target area by using a probability route map, the safety distance with a building is considered, a complex map structure is simplified, and a continuous space is converted into a discrete space; then path node selection is carried out on the path network diagram, so that the problem of low search efficiency in a complex space is solved, and meanwhile, the safety of the unmanned aerial vehicle in the flight process is ensured; according to the graph attention network of an encoder-decoder structure, an encoder uses an attention mechanism to carry out feature extraction on an obtained probability route map, different weights are given to nodes through a self-attention mechanism, and influences of key nodes related to paths and neighbors of the key nodes are emphasized; and the decoder uses attention weight to guide node selection according to node features extracted by the encoder, and finally an optimal path is formed.

The invention belongs to the field of medical image classification, and discloses a novel dual-path network architecture combining Mama and a convolutional neural network, and the novel dual-path network architecture comprises a Patch Embedding module, a plurality of Conv-OSS layers, a Patch Merging layer, and a final feature classifier. An OSS module is developed, local features are extracted through parallel convolution branch, overall representation is enriched, and global features are effectively captured; an OSSM module is designed, spatial features are modeled from multiple directions, and the extraction capability of global features is enhanced; an MSAConv network is designed, and the overall feature representation quality is improved by aggregating eight diagonal features reserved after 8DScan; an FFM module is developed, Mama output is further optimized, and fine features related to pneumonia in a medical image are emphasized. Experimental results of a pneumonia data set show that the method is superior to an existing computer-aided pneumonia diagnosis method in multiple evaluation indexes, and the huge potential of the method in actual clinical application is highlighted.

Example embodiments are directed to generating an optimized network of flight paths and an operations volume around each of these flight paths. A network system creates a source network of paths, whereby the source network comprises a set of possible paths between two locations. The network system assigns a cost for traversing each edge of each path of the source network and aggregates the cost for traversing each edge of each path to obtain a cost for each path of the source network. Based on the cost for each path, the network system identifies a path having the lowest cost, whereby the path having the lowest cost is the optimized route between the two locations. The network system then generates an operations volume for the optimized route. The operations volume represents airspace surrounding the optimized route. The operations volume is transmitted to a further system for use.

The invention discloses a multi-modal dual-path fusion network structure based on data level fusion, and relates to the technical field of dual-path fusion network structures, comprising the steps of data set construction and preprocessing, selection and processing of a plurality of data sets, and image enhancement and normalization; constructing a network structure, constructing a dual-path network, pre-training on different data sets by using a mainstream backbone network, optimizing training, splicing feature layers, and performing classified training; and performing data fusion and classification, performing horizontal fusion on the double-path features, determining the weight by adopting a weighting strategy, and inputting the fused features into a full connection layer for classification according to a softmax function. According to the method, the ophthalmic disease diagnosis accuracy and model robustness are improved, the generalization ability is improved, network training is optimized, and medical intelligent diagnosis development is promoted.

The invention relates to the technical field of network security monitoring, in particular to a multipath network communication security guarantee method based on an adaptive encryption algorithm, comprising: acquiring a network undirected graph corresponding to a target network, the network undirected graph comprising a first node set, a second node set and an undirected edge set, the first node is used for collecting and managing data, the second node is used for monitoring the safety state of the first node, and the undirected edge is used for connecting the first node and the second node. According to the method, the behavior of the second node in the network is subjected to feature extraction, the trust level of each node can be dynamically evaluated and adjusted in combination with the pre-trained trust classification model, and the self-adaptive feature can effectively adjust the communication security level according to the actual situation, so that the security of the network is improved, and the user experience is improved. And potential security threats or malicious nodes can be identified by monitoring and analyzing the behavior of the second node.

According to an embodiment of the present disclosure, disclosed herein is a system and method for validating software upgrades and optimizing network path mapping. The method for validating a software upgrade and determining network path mapping at a Network Entity (NE) using a Management Data Analytics Service (MDAS) producer. The method includes predicting a set of Key Performance Indicators (KPIs) using a Machine Learning (ML) model, performing a software upgrade, and determining KPIs associated with the NE post-upgrade. The method includes comparing the predicted and determined KPIs, and　the software upgrade is validated. Further, the method focuses on determining optimal network paths. The method receives network traffic information reflecting network performance metrics for various paths, predicts a plurality of KPIs using the ML model, and determines the optimal network paths based on the predicted KPIs.

The invention relates to the field of thermal flow guide path optimization, and discloses an electronic communication equipment heat dissipation method based on thermal flow guide path optimization, which comprises the following steps: arranging a plurality of thermal paths between a chip and a radiator, combining the thermal paths into a network structure in parallel and series modes, and constructing a selectable thermal flow path network structure; collecting real-time temperature and load data of each module, and constructing a multi-dimensional thermal state database in combination with historical operation records and environmental parameters; thermal resistance and module temperature of each thermal path are analyzed, priority is set according to module types and power loads, key modules correspond to specific thermal flow paths, non-key modules distribute thermal flow through remaining paths, and a thermal flow planning scheme is generated; the thermal resistance of each path is adjusted through the phase change material, and a target thermal resistance value is generated; in the chip operation process, the temperature, load and thermal resistance state of each thermal path are continuously collected, and a thermal resistance adjusting instruction is generated; the method has the advantage that the thermal management efficiency and the temperature control precision are improved.

The invention discloses an apple leaf disease segmentation method based on a lightweight dual-path network, an apple leaf disease segmentation device based on the lightweight dual-path network, an apple leaf disease segmentation device and a computer readable storage medium. The problems that an existing disease segmentation method is high in model complexity and insufficient in multi-scale recognition capability are effectively solved. The lightweight encoder adopts a depth separable convolution and channel recombination technology, so that the parameter quantity and the calculation complexity are greatly reduced while the feature extraction capability is maintained, and the model can be deployed on edge equipment such as an unmanned aerial vehicle and a field robot. The enhanced cavity space pyramid pooling module constructs abundant multi-scale receptive fields through multi-branch parallel cavity convolution with different expansion rates, and can capture feature information of initial tiny disease spots and later fused disease spots at the same time.

The invention relates to the field of routing algorithms of a mobile ad hoc network, discloses a multi-path routing optimization method and system for the mobile ad hoc network based on a graph attention network, and mainly solves the problem that in a traditional mobile ad hoc network, due to the fact that link characteristics and a routing scheme can generate nonlinear influence on network performance, the network performance is not influenced. And the problem of network resource waste caused by an improper routing scheme is solved. The content provided by the invention comprises a mobile ad hoc network multi-path network model, a graph attention network prediction model and a routing optimization system. The method comprises the following steps: firstly, generating path information through feature vectors of nodes and links obtained by a software defined network technology; secondly, constructing a multi-path network model of the mobile ad hoc network according to the hidden state relationship between the links and the paths; and finally, constructing a graph attention network prediction model in an application layer, thereby realizing the mobile ad hoc network multipath routing optimization method based on the graph attention network. In addition, a cross-layer routing optimization system is established according to the mobile ad hoc network multipath routing optimization method based on the graph attention network. According to the mobile ad hoc network multipath routing optimization method and system based on the graph attention network provided by the invention, node aggregation is optimized, the requirements of simultaneously learning network features and network complex structures in the mobile ad hoc network are met, the accuracy of predicting the network performance is improved, the routing strategy can be selected in an auxiliary manner, and the routing efficiency is improved. And the utilization rate of network resources is effectively improved.

The invention discloses a multi-machine collaborative mobile measurement robot path control method, which belongs to the technical field of path planning, and specifically comprises the following steps: collecting surface height information of a target area, and binding the surface height information with geographic coordinates to generate a simplified three-dimensional model; based on the sun moving trajectory data in the current season, calculating an accumulated time length distribution diagram of receiving direct sunlight by each coordinate point on the ground in the current time period; dividing a high sunlight intensity region and a continuous shadow region according to the cumulative duration distribution diagram, and carrying out topological association on the coordinates of the target point to be measured and the continuous shadow region; establishing a dynamic path distribution protocol in a multi-robot communication network, enabling each robot to select a non-overlapping measurement path in a path network formed by continuous shadow areas according to a real-time position and a task queue, and updating a path occupation state through periodic broadcast; according to the invention, the path planning of the mobile measurement robot in the urban high-temperature environment is optimized.

The invention provides a remote control method and device for intelligent multimedia interaction equipment, equipment and a medium, which are used for improving the real-time performance and the execution success rate of remote control and enabling complex multimedia interaction to be reliably operated on a heterogeneous network and heterogeneous equipment. The method comprises the following steps: carrying out hierarchical analysis processing on interaction data input by a user to generate perception command data, and carrying out execution planning and resource allocation on the perception command data according to equipment performance characteristics to generate an executable operation sequence, and performing transmission quality evaluation and path selection on the propagation path network according to the executable operation sequence to obtain a target transmission path, and performing protocol encapsulation on the executable operation sequence according to the target transmission path to obtain a remote control instruction data stream.

The present invention discloses a deep learning-based image semantic segmentation optimization method and system, which belongs to the field of image processing technology. The method and system are used to solve the technical problem of poor boundary perception ability in existing solutions. By constructing a dual-path network architecture, the collaborative learning of semantic features and boundary features is realized, and a closed-loop optimization is formed from input preprocessing to feature fusion, which directly addresses the core problem of poor boundary perception ability. The dynamic attention fusion module realizes the closed-loop optimization of boundary area enhancement, key channel screening, and dynamic noise suppression. The designed weighted cross entropy can solve category imbalance, the designed improved boundary distance loss can enhance boundary positioning accuracy, and the designed multi-scale feature consistency loss can ensure semantic alignment of deep and shallow features. The dynamic weight adjustment mechanism can adaptively balance the contribution of each loss according to the training progress, which can significantly improve the model's segmentation robustness to complex boundaries.

The invention discloses a traffic scheduling method and electronic equipment, and relates to the technical field of traffic scheduling, and the method comprises the steps: responding to a condition that a to-be-sent data stream accords with a transport layer protocol and the traffic size of the to-be-sent data stream is greater than a preset threshold value, and generating a plurality of candidate port values used for identifying the to-be-sent data stream at a data sending end; creating a plurality of detection data packets, wherein source port fields of the detection data packets are set as different candidate port values; sending the detection data packets to a data receiving end, and obtaining performance indexes of candidate network paths corresponding to the plurality of detection data packets based on the response of the data receiving end; determining a target source port value from the plurality of candidate port values based on the performance index; according to the embodiment of the invention, the source port field of the data packet corresponding to the to-be-sent data stream is modified into the target source port value, and the modified data packet corresponding to the to-be-sent data stream is sent to the data receiving end, so that the technical problem of load imbalance of a multi-path network in related technologies is solved, and the reliability of traffic scheduling is improved.

The invention discloses a video behavior recognition method based on multi-scale time feature fusion, and the method comprises the steps: taking a dual-path network as a basic structure, carrying out the normalization and uniform frame extraction of an input video, and respectively inputting a Slow path and a Fast path for feature extraction. The Slow path focuses on low-frequency background semantic modeling, the Fast path integrates a dynamic Token sparse module (VDTS) constructed based on one-dimensional convolution, and a key action area is adaptively selected through a Gumbel Top-K mechanism. Furthermore, according to the method, a gating bidirectional cross attention module (Bi-CAM) is designed between a fast path and a slow path, dynamic complementary fusion of bidirectional information streams is achieved by introducing a Query-Key-Value fusion structure and a channel attention gating mechanism, and the coupling efficiency and expression ability between multi-scale features are effectively improved. And finally, accurate recognition of behavior categories is completed through a classifier. The method gives consideration to the recognition precision and the calculation efficiency, is suitable for scenes such as behavior monitoring, man-machine interaction and intelligent security and protection, and has a good application prospect.

The invention discloses a construction unit safety responsibility tracing method based on an HSE map, and the method comprises the following steps: S1, collecting multi-source HSE data of a construction site, and constructing a structured data set; s2, constructing an HSE atlas containing multiple types of nodes and relationships based on the data set; s3, inputting the HSE atlas into the logic attention path network model, extracting a causal path and embedding nodes; s4, performing fuzzy responsibility scoring on the nodes in the extraction path to generate an attribution score; s5, adopting a cuckoo search algorithm to perform joint optimization on the model structure and the scoring parameters to improve the recognition effect; and S6, outputting a construction unit responsibility node and grade result based on the optimized model for responsibility tracing and decision making. According to the method, the recognition efficiency and the decision transparency are remarkably improved while the construction responsibility is accurately traced, and the method is suitable for various types of safety management scenes.

The invention relates to the technical field of intelligent manufacturing, and discloses an intelligent control system for a battery shell aluminum plate stamping production line. A data cleaning module of the system receives original sensing data of a stamping production line and executes fluctuation smoothing processing and abnormal point removing operation to generate clean data flow; the feature sequence generation module detects starting and ending event points of a stamping process based on the clean data flow, and performs splicing according to a time sequence to form a process chain and marks key process nodes; the path network creation module maps process nodes into network vertexes, and adds directed connection edges based on a node precedence relationship to construct a path network map; an intersection analysis module traverses the atlas to search for a shared vertex group, calculates an aggregation degree index and screens a core intersection; the attribute grouping module is used for summarizing association path attribute data sets according to the core intersections and executing similarity grouping operation to distribute group labels; and the control topology generation module maps the path groups to a control unit output control topology structure table according to the group labels.

The invention discloses an image semantic segmentation optimization method and system based on deep learning, and belongs to the technical field of image processing. The method is used for solving the technical problem of poor boundary sensing ability in the existing scheme. By constructing a dual-path network architecture, collaborative learning of semantic features and boundary features is realized, closed-loop optimization is formed from input preprocessing to feature fusion, and the core problem of poor boundary perception capability is directly solved; the dynamic attention fusion module realizes closed-loop optimization of boundary region enhancement-key channel screening-noise dynamic suppression; the designed weighted cross entropy can solve class imbalance, the designed improved boundary distance loss can enhance the boundary positioning accuracy, and the designed multi-scale feature consistency loss can ensure semantic alignment of deep and shallow layer features; through a dynamic weight adjustment mechanism, each loss contribution can be adaptively balanced according to the training progress, and the segmentation robustness of the model to a complex boundary can be significantly improved.

Techniques are disclosed for the detection of different states of a session comprising a bidirectional flow of network traffic between client devices so as to enable a network device to apply different network policies to different states of the session. In one example, a computing device identifies multiple states of a session and defines a plurality of network policies. Each network policy defines performance requirements for network traffic during each state of the session. A network device receives the plurality of network policies and determines a state of the session. The network device selects a path based on the performance requirements of the network policy associated with the determined state of the session. The network device forwards traffic associated with the session along the selected path while the session is in the determined state.

The invention relates to a multi-modal emotion recognition fusion and communication method oriented to real-time human-computer interaction, and aims to solve the problems of asynchronous modal time sequence, non-uniform feature dimension, single fusion mode, high communication delay and the like of an existing emotion recognition system and improve the recognition precision and real-time performance of the system. According to the method, voice, images and physiological signals are synchronously collected through a multi-modal input module, feature alignment is achieved through time sequence interpolation, dynamic time warping and space coordinate mapping, double-domain feature fusion is conducted in combination with a time path network and a space path network, emotional state judgment is completed through a lightweight neural network, and the emotional state recognition accuracy is improved. And outputting six types of basic emotions and confidence coefficients. Meanwhile, a low-delay communication protocol based on UDP clipping extension realizes rapid feedback of emotion data, and in combination with modal priority scheduling, data compression and bandwidth sensing mechanisms, high-efficiency and low-delay transmission is ensured, and the real-time interaction requirement in a weak network environment is met. The method has the advantages of high accuracy, low power consumption, low time delay and flexible deployment, is suitable for various real-time interaction application scenes such as voice assistants, virtual customer service, emotion accompanying and telemedicine, and has wide application value and market prospect.

The invention provides an access rule management method, device and equipment based on an access control matrix, and relates to the technical field of network security, and the method comprises the steps: obtaining a source address and a target address; generating an access control matrix based on the source address, the target address, the protocol and the target port, and determining target network equipment and passing network equipment based on the access control matrix; automatically generating corresponding access control rules for the target network equipment and the path network equipment based on the target network equipment and the access control matrix thereof, and automatically generating corresponding access control rules for the path network equipment and the access control matrix thereof; and correspondingly configuring the access control rule to the target network equipment and the path network equipment to realize access control of the target network equipment and the path network equipment. According to the invention, through the access control matrix, a visual and structured interface is provided, so that a manager can clearly check, edit and manage the network access control rule.

The invention discloses a phi-OTDR (Optical Time Domain Reflectometer) vibration signal classification method based on multi-scale residual convolution and a bidirectional long-short-term memory network, relates to the technical field of phase-sensitive optical time domain reflectometer vibration sensing, and solves the problems of insufficient feature extraction, single model fusion mechanism, poor system real-time performance and the like of an existing phi-OTDR vibration signal classification method. The method is realized through the steps of a multi-scale deep convolutional network, a bidirectional long-short term memory network, a cross attention fusion module, a learnable gating mechanism and the like. A cross attention fusion module is adopted, association between spatial features and time sequence features is dynamically established in a double-path network, feature interaction and information coupling are deepened, and the sensing ability of the model to key information is remarkably improved. Meanwhile, the fusion proportion of the multi-path features is adaptively adjusted according to the actual distribution of the input features, background noise interference is effectively suppressed, and the robustness and discrimination performance of the model are enhanced. Accurate identification and classification of external vibration signals are realized.

According to one embodiment, an information processing apparatus includes: a weight determiner configured to determine weights of a plurality of travel paths in a travel path network based on timings at which a first moving object travels on travel paths included in a first route among the plurality of travel paths; and a route creator configured to create a second route on which a second moving object travels in the travel path network based on the weights of the plurality of travel paths.