12 results about "Node deletion" patented technology

The invention discloses a heterogeneous Internet of Things topology key node identification method in combination with a mutation theory, and belongs to the field of computers and information, and the method comprises the steps: firstly constructing an initial topology of a heterogeneous Internet of Things, generating a topology set through random disturbance, and calculating performance indexes such as robustness, redundancy, communication efficiency and a global clustering coefficient; then normalizing the indexes by using a mutation series method, evaluating the correlation between the indexes based on a Pearson's correlation coefficient, and obtaining a network performance mutation value through layer-by-layer fusion; simulating a deliberate attack process, establishing a corresponding relationship between a node deletion sequence and performance evolution, and fitting to obtain a performance evolution function; and finally, establishing a cuspidal point mutation model, and identifying critical nodes causing network performance mutation by analyzing the equilibrium state, the singular set and the bifurcation set of the model, thereby breaking through the limitation of traditional static network indexes, being capable of accurately capturing the mutation behavior of the network performance, being high in objectivity, and providing a theoretical basis for topological optimization and fault-tolerant design of the Internet of Things.

The invention provides a distributed workflow task scheduling method and system, and relates to the technical field of computers, and the method comprises the steps: defining a scheduling task node and a dependency relationship in a dragging manner through a visual canvas, and storing the scheduling task node and the dependency relationship; the Worker nodes with the ZooKeeper center hop state being normal are screened, node weights are calculated based on dynamic load balancing, and tasks are distributed to the Worker node with the highest weight to be executed; when the Worker node has a fault, monitoring a ZooKeeper node deletion event, and identifying an IP (Internet Protocol) and a port of the fault Worker; marking an unterminated task state on the fault node as a state needing fault tolerance and redistributing; when the Master node fails, the surviving Master takes over the workflow instance of the failed Master based on a slot allocation mechanism; and updating the task reporting address in the Worker to the current Master, and re-distributing the uncompleted tasks. The usability, efficiency and stability of workflow scheduling are integrally improved, and diversified business requirements are met.

The present invention provides a method for convolution operator fusion, comprising: S0 first identifying the operator type to be fused; S1 first traversing the entire graph to find a subgraph structure that satisfies the sequential connection of N operators; S2, after finding the subgraph, extracting the corresponding weights h and w, strides, input and output weight bit widths, and activation attributes of the N operators, and determining whether the operator meets the conditions for fusion. If not, the subgraph is skipped; if it meets the conditions, S3 is performed; S3, if the conditions for convolution fusion are met, a new node node is created and given a new operator name, all attributes within the N convolutions are extracted, renamed, and placed in the new node, and the corresponding inputs and outputs are connected to the new node so that it has a corresponding connection relationship in the graph, and the nodes to be fused are deleted from the graph. This method can reduce the number of nodes and reduce time consumption.

The present disclosure relates to the technical field of ASON network, and provides a method, which comprises: in an ASON network SPC service, a routing node containing residual state information of LSP activation, cross configuration, deletion state and node ID; when creating or deleting an LSP operation, the node updates its own information and transmits the information to a network management for storage; after creating an LSP, if LSP of node A is not activated or cross configuration fails, there is service residual; after deleting an LSP, if node A deletion is not activated or cross configuration fails, there is also residual; the network management initiates a clearing request according to the ID of node A with residual to clear the service residual. The present application can realize efficient tracking and clearing of residual, ensure stable operation of the network, and improve operation and maintenance efficiency and reliability of the ASON network.

The invention discloses an unstructured data multi-way tree chain table storage method, which comprises the following steps of: 1, defining a data structure, and establishing a dynamic multi-way tree node structure; step 2, maintaining a target chain node, and maintaining an index relationship of the node through a directory chain table; step 3, node insertion adjustment: dynamically adjusting a pointing relation of a linked list pointer array during node insertion to form a multi-way tree topological structure; step 4, recombining pointers, and updating child node pointers of the target nodes to reconstruct a local hash table; 5, deleting and recovering the nodes, marking the nodes to be deleted as invalid nodes and recovering the storage space of the invalid nodes; step 6, carrying out data fault tolerance and recovery, and storing data by multiple copies to ensure data integrity; through the innovative unstructured data multi-way tree chain table storage method, efficient, flexible and reliable data management is achieved, a layered compression structure is adopted, memory mapping, pre-read caching and asynchronous persistence technologies are combined, and storage occupation is remarkably reduced.

The application discloses a distributed industrial wireless network deterministic transmission scheduling system, which comprises the following modules: a synchronization module for time slot synchronization between nodes; a data transmission module for data transmission between nodes; and a node management module for node deletion and addition management. The application uses distributed negotiation scheduling design, reduces overall energy consumption and delay of the network, and avoids network paralysis caused by failure of a central control node.

The invention discloses an incremental updating method for a Graphrag community, which belongs to the technical field of information technology specially suitable for supervision purpose, and comprises the following steps: monitoring a change event of a knowledge graph, and defining an affected local sub-graph based on an N-order neighborhood expansion algorithm; for node deletion, key context information of deleted nodes is converted into semantic heritage vectors through a semantic heritage vectorization method, and the semantic heritage vectors are associated to an original community; local subgraphs and semantic heritage vectors are loaded into an iterative collaborative optimization engine to be initialized, a candidate community division scheme is proposed based on current topology and historical semantic feedback in an optimization loop, a draft digest is generated, and the community structure is adjusted by quantitatively evaluating the consistency score of community cohesion and semantic tension between communities, so that the candidate community division scheme is obtained. Iteration is carried out until convergence; finally, an optimization result is atomically solidified to a graph database and an abstract index. According to the method, high overhead of total reconstruction and semantic drift of local updating are avoided.

The application discloses a method for identifying key nodes of a heterogeneous Internet of Things (IoT) topology based on mutation theory, and belongs to the field of computers and information. First, an initial topology of the heterogeneous IoT is constructed, a topology set is generated by random disturbance, and performance indexes such as robustness, redundancy, communication efficiency and global clustering coefficient are calculated. Then, the indexes are normalized by using a mutation series method, the correlation between the indexes is evaluated based on a Pearson correlation coefficient, and a network performance mutation value is obtained by layer-by-layer fusion. Next, a deliberate attack process is simulated, a corresponding relationship between a node deletion sequence and performance evolution is established, and a performance evolution function is fitted. Finally, a cusp mutation model is established, critical nodes causing network performance mutation are identified by analyzing the equilibrium state, singular set and bifurcation set of the model, and the application breaks through the limitations of traditional static network indexes, can accurately capture the mutation behavior of network performance, is objective, and provides a theoretical basis for topology optimization and fault-tolerant design of the IoT.

Embodiments of the present application provide a node adjustment method and device, a management platform, a computer storage medium and a computer program product. The method comprises: obtaining resource usage information of a first CDN node; the first CDN node is any CDN node in a plurality of CDN nodes; generating a node control instruction according to the resource usage information, and performing a corresponding node control operation according to the node control instruction; wherein the node control instruction comprises an expansion instruction, a contraction instruction, a new node instruction or a node deletion instruction; after performing the node control operation, feeding back, to a network controller, capability update information of a second CDN node corresponding to the node control operation, so that the network controller adjusts the service capability of the second CDN node according to the capability update information.

The application provides a distributed workflow task scheduling method and system, and relates to the technical field of computers. The scheduling task nodes and dependency relationships are defined and stored in a drag-and-drop manner through a visual canvas. The Worker nodes with a normal ZooKeeper center hop state are screened, the node weights are calculated based on dynamic load balancing, and the tasks are distributed to the Worker nodes with the highest weights for execution. When a Worker node fails, the ZooKeeper node deletion event is listened to, the IP and port of the faulty Worker are identified, the task states of the un-terminated nodes of the faulty node are marked as needing fault tolerance and are re-allocated. When a Master node fails, the surviving Master takes over the workflow instance of the faulty Master based on a slot allocation mechanism. The task reporting address in the Worker is updated to the current Master, and the uncompleted tasks are re-dispatched. The overall usability, efficiency and stability of the workflow scheduling are improved, and the method is suitable for diversified business requirements.

The application discloses a trajectory prediction method based on a node deletion and merging deep echo state network, optimizes the deep echo state network by using a node deletion and merging method, selects two most similar nodes according to node similarity evaluation, and merges similar nodes in the same reserve pool; similar nodes in different reserve pools at higher layers are deleted, the reserve pool structure can be adaptively optimized, redundant nodes in the network are deleted, the network optimized by using the method can be used for completing a target trajectory prediction task in an optoelectronic tracking system, and prediction precision and prediction speed can be improved, so that the performance of the optoelectronic tracking system is improved.

The present application provides a maintenance method and related devices for automatically processing information, which are used to maintain automatically processing information in an arbitration system. The method includes: when receiving a node addition operation for a first target node, adjusting the first target node from a manual processing node to an automatic processing node; when receiving a node deletion operation for a second target node, adjusting the second target node from an automatic processing node to a manual processing node; when receiving a calculation rule maintenance operation for a third target node, adjusting the reward amount calculation rule and / or penalty amount calculation rule corresponding to the third target node, where the third target node is a node in the process of receiving an order or a node in the process of continuing to declare an order. This facilitates the maintenance of automatically processed information, reduces development costs, and meets the needs of arbitration personnel.