712 results about "Greedy algorithm" patented technology

The invention provides an intelligent boxing method and system. The intelligent boxing method comprises the steps of obtaining boxing task information and generating a boxing scheme. The boxing task information comprises the size, the weight, the stacking limitation, the number and the priority of each cargo, and the size, the bearing capacity and the occupation state of each loading unit. And under the condition that the size, bearing and stacking constraint conditions are met, a plurality of candidate boxing combinations are generated, the loading utilization rate, the residual space distribution and the shape matching degree are evaluated based on a preset optimization target, and an optimal scheme is selected. The system comprises an input terminal, a task information receiving module, a data storage module, a task and scheme storage module, an operation processing module and a final boxing scheme generation module. According to the method, multi-dimensional constraint matching and weight balance are considered, and space waste and transportation risks are avoided; the maximum greedy algorithm and the adaptive large neighborhood search optimization are combined to realize the combination of automation and experience. The method can be expanded to batch task scheduling and grouping sequencing, so that the calculation complexity is reduced, and the overall boxing and transportation efficiency is improved.

The invention discloses a task scheduling method and device, equipment, a storage medium and a product, and relates to the technical field of artificial intelligence. According to the method, a large model is divided into a plurality of sub-tasks, the computing power demand is quantified in real time, and the matching score of the schedulable task and each neural network processor node is calculated in combination with the computing power evaluation information of each neural network processor node; the schedulable task with the highest priority in the scheduling queue is allocated to the neural network processor node with the highest matching score, and all subtasks of different large models are reasonably allocated through loop execution; besides, priorities are distributed to the to-be-scheduled sub-tasks in the scheduling queue based on the scheduling strategy, scheduling is carried out through a greedy algorithm, it can be guaranteed that each scheduling behavior is a real-time optimal scheduling scheme, and therefore the problem how to schedule large model tasks more efficiently is solved, and the scheduling efficiency is improved. And the computing power resource of the neural network processor is efficiently utilized to reduce the technical problem of computing power waste.

The invention discloses a broadcast type radio positioning system base station site selection analysis and evaluation method and system, and relates to the technical field of wireless positioning, and the method comprises the steps: building a radio ground base station deployment sample database and an elevation model database; based on a combined optimization strategy of a greedy algorithm and a genetic algorithm, selecting an optimal base station set covering a pre-working area from the database; constructing a three-dimensional evaluation system, and comprehensively evaluating the optimal base station set from three dimensions of deployment cost, system robustness and positioning precision; optimizing the base station set based on a digital twinning technology, wherein the optimization comprises multi-band signal emission modeling and machine learning recommendation of a historical case library; and dynamic presentation and parameter adjustment of a base station deployment effect are realized through a three-dimensional visual interface and an interactive sand table system. According to the method, a large-range working area can be well covered, the situation that a signal coverage blind area exists is avoided, then the working efficiency is improved, and flexible and efficient site selection evaluation on the radio ground base station in a local sudden battlefield can be achieved.

The invention discloses an information retrieval system and method based on semantic normalization, and relates to the technical field of artificial intelligence information, and the method comprises the steps: collecting a semantic query record input by a user, carrying out the preliminary semantic analysis, and generating structured data; on the basis of the structured data, entity disambiguation is carried out by utilizing a knowledge graph, abstract classes are generated through a neural network, calibration and dynamic weight adjustment are carried out, and high-confidence entity abstract classes and confidence scores are generated; entity abstract classes and confidence scores are combined with user contexts, an action-value function is calculated through a value network, and an optimal action is selected by utilizing a-greedy algorithm; executing semantic normalization mapping according to the optimal action, and obtaining an intermediate expression by using a meta-symbol dynamic generator; and performing index retrieval and multi-dimensional sorting based on the intermediate expression to generate a sorted retrieval result list. According to the method, the semantic fragmentation problem of multi-modal query is solved, and deep semantic alignment and dynamic weight calibration of heterogeneous data are realized.

The invention discloses a laser power optimization method based on a deep reinforcement learning strategy, belongs to the technical field of simulation and artificial intelligence, and is suitable for a laser powder bed melting (LPBF) process. The method comprises the steps that firstly, an LPBF simulation model is built, a molten pool three-dimensional temperature field distribution model is built through a moving heat source, and the forming process under a laser scanning path is simulated in combination with a part STL file; a state space and a reward function are constructed by extracting state features of a molten pool in real time, and a laser power adjustment strategy is optimized by using a deep reinforcement learning DQN network model. An optimal laser adjustment action is generated by adopting an epsilon-greedy algorithm, and intelligent agent interaction data is stored and updated based on an experience pool, so that network model parameters are optimized, and the stability and uniformity of the depth of a molten pool are ensured. According to the method, internal defects are effectively reduced, the quality and intelligent control level of the LPBF process are improved, and a new thought is provided for efficient and accurate process parameter optimization.

The invention provides an automatic identification method and system for a linear structural plane in a complex rock mass outcrop, and belongs to the field of rock mass engineering, and the method comprises the steps: obtaining a DeepLabV3 + model of a rock mass outcrop image, carrying out the semantic segmentation of an input image, outputting a binary mask image of the trace pixel of the structural plane, and carrying out the topology reconstruction; establishing a mapping relation between the image coordinate system and the three-dimensional point cloud; calculating a space included angle between the direction vector of the trace and the average occurrence of the structural plane groups, classifying the three-dimensional point cloud data according to the structural plane groups according to an angle threshold value, and performing three-dimensional line segment extraction by adopting an improved iterative Hough transform algorithm; and realizing global multi-segment connection and trace length calculation of discrete segments based on a greedy algorithm. The method breaks through the limitation of a topographic region when the rock mass structural plane data is obtained through manual measurement, and improves the precision and efficiency of structural plane extraction in the complex rock mass.

The invention discloses an equipment prediction maintenance framework based on probability residual life, and belongs to the technical field of fault prediction and health management (PHM). According to the framework, through combining a Bayesian neural network and a reinforcement learning technology, probability prediction and dynamic maintenance decision optimization of the residual life of equipment are realized. The method specifically comprises the following steps: acquiring sensor data in equipment operation, and preprocessing to generate a training data set; constructing a Bayesian neural network (BNN), utilizing variation reasoning to approximate posteriori distribution, and outputting probability distribution of residual life through Monte Carlo sampling; based on a probability prediction result, constructing a reinforcement learning environment model, and defining a state space containing residual life distribution, a spare part state and a maintenance action and a reward function; a Double DQN algorithm is adopted to optimize a maintenance strategy, and intelligent decision-making of the optimal maintenance time and the optimal ordering time of equipment is dynamically realized through an epsilon-greedy algorithm, so that the maintenance cost and the fault risk are minimized. According to the method, probability residual life distribution and reinforcement learning decision are innovatively combined, the problem that a traditional point estimation model ignores uncertainty is solved, an intelligent maintenance framework is constructed through a reinforcement learning method, and dynamic optimization of maintenance decision is achieved. The NASA aero-engine data set verification shows that compared with a traditional method, the optimization effects of prediction errors, uncertainty quantification, the maintenance cost rate and the like are remarkable, and the reliability and economical efficiency of equipment are effectively balanced.

The invention discloses a resource collaborative awareness-based Storm flow calculation dynamic scheduling method, and provides the following technical scheme aiming at the problem that the performance of a Storm default scheduling algorithm is reduced in node abnormity, resource fluctuation and rescheduling scenes: firstly, dynamically sensing node busy, downtime and data flow fluctuation events through a real-time monitoring module; triggering a rescheduling process; secondly, constructing a resource collaborative optimization model based on historical task instance resource requirements, node performance indexes and communication overhead, and generating a task instance allocation scheme by taking minimization of inter-node communication cost as a target and combining CPU / memory dynamic threshold constraints; further, a greedy algorithm is adopted to sort high-relevance task instances, the high-relevance task instances are preferentially distributed to nodes with the optimal historical performance, and it is ensured that the node resource utilization rate does not exceed a dynamic threshold value; meanwhile, node load balancing parameters are corrected in real time through time window smoothing processing, and the remaining resource state is updated; and finally, outputting an optimized topology division result, so that the system delay after rescheduling is remarkably reduced, and the throughput is improved. According to the method, rapid recovery and stable operation of the heterogeneous cluster are realized through resource collaborative modeling, historical data driven dynamic scheduling and load balancing optimization.

The invention belongs to the technical field of electric power engineering digitization and intelligent drawing auditing, and particularly relates to a dynamic rule engine and multi-target auditing task decomposition method oriented to electric power drawings. Obtaining a power rule file, analyzing and verifying to generate qualified rules, and screening the qualified rules to obtain an effective rule pool; the method comprises the following steps: reserving candidate areas by preprocessing an electric power engineering drawing, generating structured cells through line segment classification and combination and table detection, extracting and cleaning a text, and affiliating the text to the corresponding cells to form a structured result; screening candidate rules from the effective rule pool, extracting auditing objects from the structured result, generating auditing points, calculating the priority of the auditing points, and packaging the auditing points into an assignment unit set capable of being executed in parallel; on the basis of the dispatching unit set, optimizing task distribution by adopting a dynamic model and a greedy algorithm in an execution environment with the maximum concurrency, and doubly optimizing task execution by quantifying benefits and verifying rules; and performing rule judgment and result measurement on a task execution result.

The invention discloses an imaging satellite task scheduling automatic modeling method based on a large language model, which automatically generates different types of optimization models and corresponding greedy algorithms through LLMs, not only can efficiently solve the scheduling problem of different scales of tasks, but also can flexibly adjust when processing complex constraints, and improves the task scheduling efficiency. The method shows strong automatic correction capability, and can realize superior resolving capability and adaptivity when processing large-scale tasks.

The invention provides a wireless resource scheduling method and device used in a scene in which a mobile phone is directly connected with a satellite, and relates to the technical field of satellite communication, and the method comprises the steps: constructing an electromagnetic map which represents the corresponding relation of a geographic position, a physical resource block and ground interference power in a satellite beam coverage area; link budgeting is carried out on the multiple user terminals in combination with the real-time orbit geometric parameters of the satellite and the electromagnetic map, and the signal to interference plus noise ratio of each user terminal on each physical resource block is predicted; based on the predicted signal to interference plus noise ratio, adopting a proportional fair marginal gain greedy algorithm to allocate an optimal continuous spectrum resource block for each user terminal; on the allocated spectrum resources, optimizing and reallocating the total transmitting power of the satellites by adopting a water injection algorithm; and a final spectrum-power joint scheduling scheme is generated, and subsequent link self-adaption and data transmission are guided. According to the invention, the spectrum efficiency, throughput and connection stability of the system are obviously improved.

The invention discloses an OTA upgrade task dynamic configuration control method and system based on machine learning, and the method comprises the steps: collecting vehicle upgrade data, and carrying out the preprocessing of the vehicle upgrade data; designing a prediction model, carrying out training and effect evaluation on the model, and outputting the upgrade probability of each vehicle in a future time period and a recommended upgrade time window according to the preprocessed vehicle upgrade data; and carrying out OTA upgrade task allocation according to the output of the prediction model to complete vehicle upgrade. According to the method, the vehicle upgrading data are input into the prediction model for prediction, the predicted upgrading time window is obtained, OTA upgrading task allocation is performed, the vehicle upgrading demand can be predicted, the upgrading time window can be recommended, overlarge system load and current limitation caused by centralized task issuing are avoided, and the reliability, stability and success rate of system upgrading are improved; and a greedy algorithm is added in OTA upgrade task allocation, so that the allocation scheduling of the upgrade tasks is optimized, and the network and system loads are balanced to the maximum extent.

The invention discloses a multi-specialty and multi-nest scheduling method and related device based on electric power joint inspection, and the method comprises the steps: formulating the inspection strategy of each specialty of transmission, transformation and distribution to design an inspection plan, and integrating the inspection plan into an unmanned aerial vehicle joint inspection plan library to obtain an inspection task set; clustering a plurality of types of routing inspection tasks with similar features in the routing inspection task set by adopting a clustering algorithm to obtain a routing inspection flight task consisting of a plurality of devices; based on the established waypoint of the inspection flight task and the endurance time of the unmanned aerial vehicle, generating an inspection route of the unmanned aerial vehicle through a TSP algorithm; and based on the inspection route of the unmanned aerial vehicle, performing one-to-one matching on each inspection flight task and each unmanned aerial vehicle nest by adopting a greedy algorithm, and finally forming automatic scheduling of the unmanned aerial vehicle nest. The problems that in the prior art, routing inspection tasks are distributed unreasonably, cross-professional multi-nest routing inspection is not fully considered, and systematic and comprehensive research on the whole process of routing inspection operation is lacked are solved.

The invention discloses a semantic retrieval generation method based on a clustering and classification model, and particularly relates to the technical field of information retrieval and natural language process.The method comprises the steps that after a text is preprocessed, semantic vectors are generated through Sension-BERT, and multi-level technical theme branches are formed through UMAP dimension reduction and HDBSCAN clustering; extracting subject terms and expanding synonyms to construct an initial retrieval formula, and iteratively optimizing through a greedy algorithm to obtain an optimal retrieval formula; defining a retrieval range in combination with the fine-tuned BERT classification model and the IPC classification number prediction model; and dividing a retrieval range according to theme hierarchy, integrating results, sorting and returning, and recording retrieval history support multiplexing at the same time. According to the semantic retrieval generation method based on the clustering and classification model, the accuracy, coverage and flexibility of semantic retrieval are improved by constructing a multi-level theme system, optimizing a retrieval formula, precisely limiting a range and intelligently integrating a result, and target information meeting requirements is efficiently returned.

The invention provides a truck-unmanned aerial vehicle cooperative dynamic scheduling method and system in emergency logistics, and relates to the technical field of unmanned aerial vehicle intelligent scheduling. Constructing a truck-unmanned aerial vehicle cooperative distribution network, wherein each truck is equipped with a cooperative unit comprising a large unmanned aerial vehicle and a small unmanned aerial vehicle; the network comprises warehouses, hub points and demand points; establishing a mixed integer programming model by taking minimization of the total operation cost as a target and taking electric quantity management, service distribution, a time window and collaborative feasibility as constraints; and solving the model by adopting a randomized greedy algorithm, and synchronously deciding a truck path and an unmanned aerial vehicle task in single iteration to obtain a truck path scheme, an unmanned aerial vehicle service distribution scheme and a dynamic electric quantity scheduling scheme. Therefore, the cost optimization and dynamic coordination of emergency logistics scheduling are realized, and the efficiency and feasibility of emergency distribution are effectively improved.

The invention discloses a multi-station robot photographing task allocation method for hub surface defect detection, and the method comprises the steps: forming a hub triangular patch model, and carrying out the Gaussian mapping of the hub triangular patch model; finding out the task area boundary of each station robot based on a simulated annealing algorithm, wherein the included angle between the normal vector of the triangular patch in the same task area and the Z axis is in a set range; generating a camera viewpoint of each station robot for completing task area shooting, and adjusting the viewpoint with collision into a collision-free viewpoint; planning a local optimal path passing through all collision-free camera viewpoints based on a greedy algorithm; and carrying out collision detection on the local optimal path, and re-planning the road section with collision. According to the multi-station wheel hub defect detection method, the task area corresponding to each station is determined according to the wheel hub curved surface features by adopting the simulated annealing algorithm, finally, automatic balanced distribution of the multi-station robot operation tasks is achieved, collision detection is conducted on the generated shooting path, and the reliability of multi-station wheel hub defect detection is improved.

The invention discloses a flexible dynamic distribution method and system for a power module based on a charging pile, and relates to the field of power distribution, and the method comprises the steps: collecting dynamic demand parameters of a charging terminal in real time, and calculating a target power demand in combination with a battery safety constraint; obtaining the total number of available power modules and rated power, and generating an effective module set; calculating the minimum module demand number according to the total demand and the module average power; if the modules are sufficient, an efficiency priority strategy is adopted, and distribution is optimized through a greedy algorithm; if not, high-power terminals are preferentially distributed, and then residual power is distributed in a balanced mode; through real-time data, distribution is dynamically adjusted when a threshold value is exceeded; starting dormancy for the module with the load rate lower than 30%, migrating the terminal to other modules, and awakening when the load rises to 40%, so as to improve the efficiency and the resource utilization rate. The method has the advantages that through real-time monitoring and dynamic distribution of the power modules, the charging efficiency and safety are optimized, and the resource utilization rate and the system reliability of the charging pile are remarkably improved.

The invention discloses a dynamic planning method and system for an intelligent patrol point location of power transformation equipment, and belongs to the technical field of intelligent patrol of power systems, the dynamic planning method for the intelligent patrol point location of the power transformation equipment comprises the following steps: mapping dynamic features and static attributes to the same feature space and carrying out cross-modal association; constructing a defect severity model, and introducing defect severity in risk quantification calculation to generate a point location priority list; a transformer substation three-dimensional point cloud model is built, path nodes are initialized according to a point location priority list, and a greedy algorithm is utilized. A closed-loop decision-making system of multi-modal data fusion is constructed; according to the method, automatic planning driven by a risk quantification model based on defect history, generation of a three-dimensional space non-blind area coverage path, dynamic adjustment triggered by two factors of equipment change and inspection effect, deep collaborative analysis of machine account-defect-real-time data and continuous inspection of a complex scene are guaranteed by a semantic compensation mechanism.

The invention provides a demand response bus intelligent scheduling method and system driven by real-time data. A demand response bus intelligent scheduling method driven by real-time data comprises the following steps: S1, collecting and fusing multi-source data in real time, and constructing a dynamic data set; s2, carrying out short-time demand prediction and clustering analysis based on LSTM and DBSCAN; s3, dynamic service area division and greedy algorithm line generation; s4, elastic vehicle scheduling based on combination of mixed integer programming and a simulated annealing algorithm; s5, dynamically adjusting the real-time ticket price driven by the supply and demand game model; s6, passenger personalized service matching is achieved through a collaborative filtering algorithm; s7, performing multi-target simulation verification on the digital twin platform; and S8, reinforcing a learning-driven dynamic feedback optimization mechanism. According to the real-time data-driven demand response bus intelligent scheduling method and system provided by the invention, the operation efficiency, the passenger satisfaction degree and the enterprise benefit of the demand response type customized bus can be remarkably improved, and the method and the system are suitable for construction of an urban intelligent bus system and have remarkable advantages.

The invention relates to the technical field of the Internet of Things, and discloses a security assessment method and system based on the Internet of Vehicles, and the method comprises the steps: obtaining multi-dimensional original data, context data and a traffic data flow; attack event features are generated through data classification, cleaning, standardization and PCA feature extraction; constructing spatial and temporal distribution characteristics in combination with context data, determining a damaged object and a damage radius through a K-means clustering and morphological expansion algorithm, and outputting a damage area; analyzing abnormal behavior distribution based on the area and the traffic data flow, and generating a distribution diagram; analyzing the threat level according to the distribution map, the damaged object and the hazard radius, distributing resources by combining a greedy algorithm, and planning a time sequence to construct a dynamic defense strategy; optimizing a protection scheme according to the real-time feedback data; finally, behavior changes are continuously monitored, and a risk thermodynamic diagram and a safety gap list are generated. According to the method, the Internet of Vehicles safety assessment capability is improved.

The invention relates to a remote sensing satellite group adaptive task planning method and system based on a large language model, and the method comprises the steps: building an optimization model based on decision variables, constraint conditions and optimization targets of remote sensing satellite group task planning; generating an initial task planning strategy by adopting a greedy algorithm based on the optimization model; performing iterative optimization on the initial task planning strategy through a large language model, generating a new round of task planning strategy in each iteration, performing scoring by taking an optimization target as an evaluation function, stopping optimization until a preset number of iteration optimization times is reached, and outputting the task planning strategy with the highest score; and based on the task planning strategy with the highest score, performing remote sensing satellite group adaptive task planning. According to the method, multi-satellite distributed task planning can be effectively carried out on the dynamic task, the task income is improved, and the satellite imaging time consumption is reduced.

The invention relates to the technical field of electric energy distribution, in particular to an intelligent power distribution system and a power distribution method.According to the intelligent power distribution system and method, power flow paths and load density distribution are defined by collecting load distribution values of nodes of a power distribution network and calculating power flow, line transmission loss and load value density of load transmission paths among the nodes; according to the method, a data foundation is laid for subsequent dynamic adjustment, partition node power requirements and transmission path characteristics are integrated by adopting an alternating direction multiplier method, a dynamic partition model is established, a global problem is decomposed into a local optimization sub-problem in a partition optimization process, partition boundary adjustment and power balanced distribution are realized through iterative calculation, and a partition node is optimized. A greedy algorithm is adopted to screen load priorities and demanded quantities, the electric energy guarantee capability of high-priority loads is improved, through adjustment of a voltage offset range and dynamic adjustment of reactive power compensation equipment, the influence of voltage fluctuation on power balance between partitions is effectively reduced, the voltage stability is improved, and the energy consumption and the power loss are reduced.

The invention discloses an intelligent power grid load prediction and dynamic response coordinated scheduling method, and relates to the technical field of power system automation, and the method comprises the steps: accessing intelligent ammeters, distributed power controllers and other devices of Modbus, IEC61850 and DL / T645 protocols through a multi-protocol adaptive gateway, and achieving data standardization; time stamps are calibrated by means of Beidou time service and an IEEE1588PTP protocol, and it is ensured that multi-source data synchronization errors are controllable; deploying an edge computing node cluster, distributing high-priority tasks to low-load nodes through an edge coordinator in combination with a load fluctuation level and a greedy algorithm, and ensuring real-time processing efficiency; the edge nodes generate short-term load prediction, and the cloud platform outputs medium and long-term prediction based on a historical data training model; and finally, the coordinated scheduling decision module fuses the two types of prediction results and the real-time parameters of the power grid, and generates a dynamic instruction to control the output of the adjustable load and the distributed power supply.

The invention discloses an HTAP database query optimization method and device and a storage medium. The query performance is remarkably improved by combining the advantages of row storage and column storage. Comprising the steps of analyzing SQL query to generate a logic query plan; evaluating the execution cost of each logic operator in different storage modes by using a mixed cost model, wherein hyper-parameters of the mixed cost model are optimized through Bayesian search; selectively adopting a dynamic programming algorithm or a greedy algorithm to generate a hybrid physical query plan with the minimum execution cost according to the table number; and executing the plan to obtain a query result. In addition, an MCTS algorithm is further utilized, the optimal column subset is selected in the preset memory budget limit based on the query historical record and loaded into the memory column storage, and the subsequent query performance is further optimized. The problem that row storage and column storage in an existing HTAP system cannot work cooperatively is solved, operator-level mixed execution is achieved, and meanwhile memory resource utilization is optimized.

The invention provides a greedy algorithm-based unreinforced hull plate inverse finite element measuring point optimal arrangement method, and belongs to the field of ship structure health monitoring. The problem that it is difficult to reduce the number of arranged measuring points under the condition that the required reconstruction precision is kept is solved. The method comprises the steps that discretization processing is conducted on an unreinforced hull plate, and the unreinforced hull plate is divided into four-node units meeting preset conditions; generating a primarily selected measuring point scheme, namely arranging measuring points along the diagonal direction of a ship body plate, and performing array copying and segmented movement to obtain a plurality of initial measuring point arrangement schemes; and inputting the primarily selected measuring point scheme into a greedy algorithm, performing cyclic displacement reconstruction and node error calculation in combination with an inverse finite element method, iteratively removing the measuring point which has the minimum influence on the reconstruction error until all unit errors meet a preset precision threshold value, and outputting an optimized arrangement scheme which has the minimum number of measuring points and meets the precision requirement. The method is mainly used for hull structure safety assessment in the whole life cycle.

The invention relates to a multi-fan blade angle non-uniform distribution structure for reducing beat vibration noise, and aims to solve the technical problem that the beat vibration noise is difficult to reduce through scientific and accurate blade number setting and blade angle non-uniform distribution design of existing similar products. Blades at a fan of the structure are arranged at non-uniform angles by taking a shaft core in a hub as a center, and the non-uniform angles are circular angles formed by the blades which are arranged at non-uniform angles by taking the shaft core in the hub as the center; the method is characterized in that the blades modulate the real-time rotating speed of each fan through a control center so as to avoid frequency discretization or set the number of the blades according to discrete noise frequency distance, and finally the non-uniform distribution angles between the blades are obtained through blade angle non-uniform distribution modulation; the control center is a control module which comprises a control chip and a signal conversion input and output port and is used for data calculation and processing. When the mode B is selected in the real-time regulation and control stage of the blade angle non-uniform distribution modulation, a sorting and interval greedy algorithm is adopted.

The invention relates to the technical field of order allocation, in particular to a housekeeping platform order allocation system based on big data analysis, which comprises the following steps of: comprehensively analyzing housekeeping service request information, identifying service types, emergency degrees of time requirements and resource requirements, and integrating to form a preliminary service order list based on clear rules; the resource allocation is ensured to have pertinence from a demand end, and the matching probability of the order and the service personnel is optimized and calculated by updating the position and state data of the service personnel in real time and adopting the ant colony algorithm, so that the order matching is quickly adjusted and accurately allocated in a dynamic scene, the low-efficiency matching possibility is reduced, the resource utilization rate is improved, and the user experience is improved. The optimal service path is designed from the angles of cost and efficiency through the greedy algorithm, the service response time is ensured to be minimized, meanwhile, the operation cost is reduced, the economical efficiency and the performability of path planning are enhanced, and the distribution efficiency, the user satisfaction degree and the overall operation efficiency of the platform are improved.

The invention discloses an optimized node selection method based on a BGP-iSec protocol partial deployment scene, which comprises the following steps of: in a first stage, calculating three static topological characteristics of nodes according to a key node selection algorithm, weighting the three static topological characteristics and AS node levels in proportion, calculating a mixed weight of the nodes, and selecting the three static topological characteristics according to the mixed weight; preliminarily determining candidate nodes; and in the second stage, a greedy algorithm is adopted, the node with the maximum gain for the current path coverage rate is gradually selected from the candidate nodes according to the priority and added into the to-be-deployed set until the path coverage rate is stable or the number of the nodes reaches the specified deployment cost. According to the method, the key AS node selection strategy is formulated, so that the defense effect of the path manipulation / prefix hijacking attack is optimized under the limited deployment cost.

The invention relates to the technical field of computer networks, and discloses a link state sensing neural network model training flow path selection method, which comprises the following steps: collecting path information between each pair of hosts by sending a detection data packet carrying an in-band network telemetry field; calculating a priority index based on the number of sent bytes of the training task, wherein the priority index is determined by the proportion of the number of sent bytes of the training task to the total number of bytes needing to be sent and the number of sent bytes; a greedy algorithm is adopted, the training tasks serve as current training tasks according to priority indexes from high to low, paths with the lowest congestion degree are preferentially allocated to the current training tasks, and if the paths with the lowest congestion degree are occupied, suboptimal paths are allocated to the current training tasks; the congestion degree of the path is evaluated through the link utilization rate of each link in the path; according to the method, the link state and the task characteristics are monitored in real time, the path selection strategy is dynamically adjusted, and the communication efficiency of distributed neural network model training is remarkably improved.

The invention relates to the technical field of business management, in particular to an artificial intelligence business intelligent docking service platform, in the invention, task dependency relationships are rearranged by adopting topological sorting, and waiting time and resource waste caused by improper dependency relationships among tasks are reduced through scientific execution sequence adjustment, so that the service efficiency is improved. The method comprises the following steps of: extracting a business semantic structure chart, mining tasks which can be executed in parallel, recombining an execution path, reducing redundant nodes and unnecessary delay of a process, improving the execution efficiency of the whole process, extracting key operation rules through natural language processing, establishing the business semantic structure chart by adopting a support vector machine, accurately reflecting the hierarchy and association relationship of the operation nodes, and improving the execution efficiency of the whole process. Visualization and operability of complex business logic are achieved, flow optimization is carried out through the greedy algorithm, node priorities are defined, adaptive execution paths are matched, high efficiency and execution accuracy of flow design are ensured, and the response speed, the resource utilization rate and the execution accuracy of the business flow are improved.