1171 results about "Topology information" patented technology

The invention discloses a PCFarm resource scheduling method and system based on dynamic load prediction, and relates to the technical field of resource scheduling, and the method comprises the steps: constructing a feature extractor based on multi-modal feature fusion, mapping original data into a high-dimensional feature vector, and obtaining a dependency relationship between tasks; modeling a cluster topology, predicting a load propagation effect between nodes, constructing a spatio-temporal joint prediction framework, and fusing a time sequence and spatial topology information; proposing a multi-objective optimization function; training a scheduling strategy generator, and generating an optimal scheduling scheme based on the current cluster state; designing the elastic expansion and contraction of the prediction drive; constructing a migration cost model, and quantifying the influence of task migration on the performance; the overall load balance degree of the cluster is calculated through the global controller, and a coarse-grained migration instruction is generated. By arranging the load prediction module and the resource scheduling module, resource allocation is dynamically adjusted according to the real-time state and task requirements of the cluster, and a better resource scheduling effect is achieved.

The present disclosure relates to the technical field of power systems. Disclosed are a distributed fault self-healing method and system for an active distribution network, and a device and a medium. In the present disclosure, using a two-terminal waveform difference as a faulty section locating principle is applicable to an active distribution network, which contains various types of DGs and different DG permeabilities. The present disclosure has a relatively low requirement for data computation and storage regarding an STU, and topology identification and update can be automatically completed during operation by simply configuring basic static network topology information for the STU, thereby adapting to a change in the operation mode of the network. By means of the present disclosure, a rational fault recovery rule is set, and important load nodes can be recovered as much as possible while meeting a network power constraint. In a recovery solution, the presence of distributed generation is taken into consideration, and different processing methods are designed according to whether the distributed generation can operate in an island manner for a long period of time.

The invention discloses an intelligent cable fault diagnosis and positioning method and system, and relates to the technical field of intelligent operation and maintenance of a power system. The method is used for accurately identifying and positioning high-resistance faults and external damage. According to the method, electric field, current, temperature and vibration signals are synchronously collected, a multi-source fusion enhanced signal flow is constructed, and multi-physical field features are extracted; based on a coupling mechanism of an electromagnetic-thermal field and a mechanical-electric field, generating a fault type label and a space coordinate; executing targeted impedance correction for different fault types, establishing a dynamic topology network, and inputting a time-space diagram neural network to output a preliminary positioning result; and multi-source verification is carried out by further fusing salinity dielectric, a harmonic thermal field, a vibration electric field and stress topological information, a high-confidence-coefficient fault positioning result is finally output, a closed-loop diagnosis mechanism is formed, and the fault recognition accuracy and the system adaptability under complex working conditions are improved.

The invention discloses a wide-narrow integrated ad hoc network method, and belongs to the technical field of network communication. The method comprises the following steps: a network node generates dynamic topology information based on a narrow-band communication module periodic broadcast state beacon; according to the link quality and the node load in the topological information, coordinating the resource allocation of broadband and narrowband channels by adopting a load-aware time slot allocation algorithm; data are classified according to service types, broadband data are modulated and transmitted by adopting orthogonal frequency division multiplexing, and narrowband control instructions are modulated and transmitted by adopting differential phase shift keying; when the link quality is deteriorated, the link is switched, and the transmitting power and the frequency band parameters are dynamically adjusted; and adaptive switching and data distribution of wide and narrow band communication modes are realized. According to the method, the problems of difficulty in topology maintenance and low bandwidth and narrowband resource allocation efficiency in a high-dynamic network scene are solved, the network adaptability, the anti-interference capability and the resource utilization rate are remarkably improved through a dynamic cooperative bandwidth and narrowband communication mode, and meanwhile, the protocol switching overhead and energy consumption are reduced.

The invention provides a civil air defense engineering concealed conduit leakage positioning method based on acoustic characteristics, and belongs to the technical field of civil air defense engineering. An acoustic signal acquisition network is constructed by deploying a high-sensitivity hydrophone array, and signals are preprocessed by using a time division multiple access technology and a pulse compression technology; and carrying out time-frequency analysis by applying wavelet transform to extract acoustic features. A time reversal mirror technology is introduced to identify direct propagation and multipath reflection signals, a fluid acoustic coupling propagation equation is constructed to analyze a leakage sound source mechanism, and pipe network topological information and a sound wave speed correction function are combined to compensate a measurement error. Wherein the deep learning pipe network acoustic propagation multi-modal model is fused with a pipeline structure encoder, an acoustic feature extractor and a position prediction decoder, and high-precision leakage positioning and degree evaluation in a complex environment are realized through a three-stage pre-training strategy; the technical problem that it is difficult to accurately locate the position of a leakage point in the complex pipe network environment of civil air defense engineering is solved.

The invention discloses a multi-modal large model interpretable diagnosis method and system for a wind turbine generator, and relates to the technical field of wind turbine generator fault diagnosis, comprising the step of combining multi-modal data (vibration, time sequence, image and text) and topological information to realize fault diagnosis through cross-modal contrast learning and topological modeling. The method comprises the steps of multi-modal feature extraction, standardization and alignment, and feature fusion through topology embedding optimization and a cross-modal attention mechanism. In the fault diagnosis process, dynamic correction and path reliability evaluation are introduced by using a regular Agent and a topology consistent Agent, weighted fusion is performed on each modal feature and a topology structure, and finally an accurate fault type and a component positioning result are output. Through combination of knowledge retrieval and a multi-Agent decision model, the adaptability and precision of fault diagnosis are improved, especially in a complex environment, the fault mode of the wind turbine generator can be effectively identified, and the system reliability is improved.

The invention discloses a game reinforcement learning-based reactive power dispatching method for an electric vehicle to participate in a power distribution network, and the method comprises the steps: collecting node voltage, line current, electric vehicle charging and discharging states and topological information, and constructing graph attention embedding; generating an initial reactive power regulation action in the multi-agent game reinforcement learning framework, and shaping and updating a strategy gradient through double rewards; the improved particle swarm is initialized according to the updating strategy gradient, and a self-adaptive inertia coefficient is set; iteratively searching according to the node voltage sensitivity, the line reactive margin and the expected convergence step number to obtain an optimized particle swarm; the particle speed is mapped to a main strategy network through online cooperative training, the inertia coefficient is synchronously adjusted, a cooperative optimization strategy is output, a reactive power dispatching instruction is generated, and the algorithm is updated in a closed-loop mode according to real-time feedback. According to the invention, rapid and cooperative reactive dynamic scheduling of an electric vehicle group is realized, and the voltage stability and the electric energy quality are remarkably improved.

The invention relates to the technical field of power grid self-healing control, and discloses a power grid self-healing control intelligent optimization method and system, and the method comprises the steps: obtaining the topological structure information of a power grid, dividing n power grid toughness evaluation regions according to the topological structure information of the power grid, calculating a hierarchical toughness evaluation index, and forming a comprehensive toughness evaluation index according to the hierarchical toughness evaluation index; a self-healing control intelligent optimization model is constructed based on the comprehensive toughness evaluation indexes, a self-healing control optimal solution is obtained by solving the self-healing control intelligent optimization model, the self-healing control optimal solution is dynamically calibrated, and a self-healing control strategy is generated; according to the method, the topological structure, the historical fault data and the real-time operation state of the power grid are comprehensively considered, a layered evaluation and dynamic calibration combined mode is adopted, the intelligent level and adaptability of self-healing control of the intelligent micro-grid are improved, the toughness and recovery capacity of the micro-grid for coping with complex faults are enhanced, and safe, reliable and efficient operation of the power grid is guaranteed.

The invention discloses a large language model low-voltage power distribution network topology perception method based on cloud edge fusion, and relates to the technical field of power systems, and the method comprises the steps: collecting the data of a low-voltage power distribution network through an edge device, and analyzing a topological structure through a cloud model; standardizing the power distribution network data based on a large language model, and verifying topological information; and establishing an n-shaped model to realize direct-current and alternating-current power distribution network differential modeling, and based on the standardized power distribution network data, automatically generating a standard power distribution network model and generating power distribution network topology output. According to the method, low-voltage power distribution network data is collected through edge equipment, a topological structure is analyzed through a cloud model, multi-data source input is supported, topological information is verified, standardized processing and data quality control are achieved, direct-current and alternating-current power distribution network differential modeling is achieved by establishing an n-shaped model, and the modeling efficiency is improved. The calculation efficiency and precision of the model are improved, the application efficiency of topological data is improved through automatic output, and the multi-scene requirement is met.

The invention discloses an online prediction method and device for node voltage of a time-varying topology direct-current power distribution network, and the method comprises the steps: building a simulation model based on the historical operation data of the direct-current power distribution network in combination with the real topology of a power grid, and generating a time sequence training data set containing voltage, power and line states through dynamic topology simulation; based on the training data set, constructing a dynamic space-time diagram neural network fused with topology dynamic change, and learning a voltage time sequence evolution rule in combination with a time sequence module; node power data and topology information are collected in real time by using a dynamic space-time diagram neural network, network topology input is dynamically updated, and a future short-time domain voltage predicted value is output; and for a power grid topology change scene, graph connection parameters associated with change nodes are updated, rapid fine adjustment of the model is realized, and stable prediction is carried out. According to the method, network topology information can be dynamically input under the condition that the topology of the direct-current power distribution network is changed, and rapid and stable prediction of the node voltage of the direct-current power distribution network under the time-varying topology is achieved.

The invention discloses a distributed model training-oriented routing reconstruction method in a hybrid photoelectric network, and belongs to the field of network communication optimization. The method comprises the following steps: constructing a collaborative routing model based on a current network flow state and topological information, and introducing an adaptive weight mechanism to dynamically adjust the priority of optical channels and electric packet forwarding, thereby realizing efficient collaborative utilization of multi-layer forwarding resources. According to the method, path reconstruction can be rapidly completed according to task characteristics and a network dynamic state, so that the system resource utilization rate is improved, and the maximum task completion time delay is reduced. Compared with a traditional fixed routing and semi-dynamic routing method, the method has the advantages that the average completion time delay and the maximum time delay are improved by more than 45% in the large-flow environment generated by large-scale model training, and the method has good expandability and real-time performance and is particularly suitable for burst flow intensive intelligent computing cluster scenes.

The invention discloses a distribution network recording fault accurate positioning method based on big data analysis, and relates to the technical field of power distribution network fault positioning, and the method comprises the steps: deploying a fault indicator; setting a trigger threshold value, and triggering high-precision transient recording to generate recording data when the current or the electric field suddenly changes; carrying out time alignment on the recording data and synchronously acquired current intensity, electric field intensity, cable temperature and environment data, extracting transient characteristics and environment characteristics after wavelet transform denoising, and combining the transient characteristics with the environment characteristics to generate a comprehensive characteristic matrix; binding the comprehensive characteristic matrix and the line topology information, and uploading the information to a monitoring center through encryption communication; formulating a dynamic threshold association rule, and judging a fault type; the monitoring center matches the current transient waveform with a historical database and calculates a fault signal propagation time difference; and calculating a fault distance by combining line topology information, and fusing the fault signal propagation time difference and the fault distance to obtain a fault coordinate.

The invention discloses a cross-domain workflow scheduling system for dynamic resource arrangement, and the system comprises a resource abstraction subsystem which is used for collecting all resource platform information and network topology information, and constructing a resource directory and all adaptation layers; the task granularity division subsystem is used for receiving a target task, determining task characteristics of the target task, the cost of each task and an initial dependency graph, determining a target granularity according to the cost of each task, and optimizing the initial dependency graph to obtain a target dependency graph; the workflow arrangement subsystem is used for determining an initial execution plan according to the target dependency graph and the resource directory, determining a corresponding adaptation layer after each target execution platform receives target data, and submitting a corresponding subtask to each target execution platform; and the data preheating management subsystem is used for determining the transmission priority of the cross-platform sub-tasks and sending the target data to each target execution platform in combination with the network topology information, so that the effect of uniformly managing different platform resources is realized, and the stability of the cross-domain tasks is improved.

The invention discloses a TSN scheduling optimization method and device based on flow sensing autonomous learning, equipment and a medium. The method comprises the following steps: deploying a lightweight flow detection module in a switch or a router, and after a controller receives a data request, automatically identifying a newly arrived unknown service flow by using the lightweight flow detection module, and judging whether the newly arrived unknown service flow is a periodic TT flow or an unpredictable burst flow; the controller performs classification management on the identified service flow types, collects topological information and flow requirements of the whole network and issues the topological information and the flow requirements to the terminal nodes through a network management interface; the controller constructs an intelligent queue scheduling task based on the collected network topology information and traffic demand and converts the task into a Markov decision process MDP, network resources, queue states and priorities are used as a state space, a scheduling strategy is used as an action space, a reward function is designed in combination with throughput and delay indexes, and an intelligent queue scheduling task is obtained. Driving a dynamic environment through real-time data and training a DRL model; an enhanced queue scheduling mechanism Pro-CQF is adopted, different priority labels are configured according to classified flow types, and then mixed flow scheduling is carried out; the controller periodically collects time delay, packet loss rate and end-to-end transmission delay indexes and feeds back the indexes to the DRL model, and a scheduling strategy is updated online. According to the method, the traffic sensing and scheduling efficiency is greatly improved in a network environment in which multiple service flows coexist and end-side equipment functions are different, and the reliability and the expandability of the TSN in industrial Internet of Things, edge computing and other high-real-time application scenes are remarkably enhanced.

The invention relates to a satellite network traffic spatio-temporal characteristic prediction method based on multiple convolutional neural networks, comprising the following steps: step 1, preprocessing input data including satellite network traffic data and satellite network topology information; 2, extracting spatial features, and processing a spatial dependency relationship of satellite network traffic through a GCN module; 3, time features are extracted, global features and local features of the satellite network traffic in the time dimension are extracted through a double-branch Transform encoder structure, the global branch processes long-term dependence through causal convolution and a global attention mechanism, and the local branch processes short-time mutation through dynamic convolution and a local attention mechanism; and 4, fusing the spatio-temporal features and output prediction, fusing the spatial features and the time features, and outputting a prediction result of the satellite network traffic through a linear layer. The accuracy and robustness of satellite network traffic prediction can be improved.

The invention provides a distribution agent ad hoc network method and system based on dynamic node collaboration, and the method comprises the steps: firstly obtaining a network state data set of all distribution nodes in a target network region, including real-time communication load, link stability and resource occupation characteristics, generating dynamic collaboration topology information based on the network state data set; comprising a node cooperation relation topological graph and a communication path weight table, a distribution agent strategy set is generated according to dynamic cooperation topological information, path optimization processing is carried out on a target distribution task according to an initial path distribution strategy, a dynamic load adjustment rule and a path switching trigger condition, and an optimized distribution path set is obtained; and deploying to the corresponding distribution node to execute dynamic cooperative distribution, and updating the network state data set according to the distribution real-time feedback data, so that the distribution efficiency and reliability can be improved.

The invention provides a transient recording fault intelligent judgment method combining waveform matching and cloud edge cooperation, and relates to the technical field of fault diagnosis. Comprising the following steps: collecting global topological information and fault judgment reference parameters of a master station synchronous power system, and outputting initial transient recording data; carrying out preprocessing and feature extraction on the initial transient recording data, realizing edge side layered fault detection by adopting an improved waveform matching algorithm, and outputting multi-layer fault information; constructing a global fault judgment model, updating the global fault judgment model through the hierarchical model, and optimizing the fault judgment model; multi-level fault information is input into the optimized fault judgment model for accurate fault judgment, and graded response measures are taken for real-time judgment results with different confidence degrees; according to the invention, through deep fusion and system-level optimization of multiple technical streams, a transient recording fault intelligent judgment system integrating real-time performance and accuracy is created.

The invention relates to a multi-branch power distribution network double-end traveling wave fault positioning method and system, and the method comprises the steps: deploying a plurality of measurement points based on the head and tail ends of a radial power distribution network and branch terminals, collecting topological information, and calculating the linear mode wave velocity; using the three-phase voltage data after the fault to extract traveling wave line mode and zero mode components through modulus decoupling; a line mode component with the strongest global mutation is selected as a reference, and a wave head time sequence table is generated through wavelet decomposition and mutation detection; and correcting the zero-mode wave velocity by combining the topology, the time sequence table and the line-mode wave velocity, and determining a fault interval and outputting a fault position after compensation and correction by using the line-mode and zero-mode wave head time difference. According to the invention, through a wave head time sequence analysis and time difference compensation mechanism based on cooperation of multiple measuring points, positioning errors caused by clock asynchronization and parameter disturbance are effectively overcome, accurate distance measurement which is not influenced by clock errors is realized, and the distance measurement precision is significantly improved; and meanwhile, the point distribution requirement is simplified, accurate judgment and positioning of a fault interval can be realized only by a branch single measuring point, and high economy and deployment flexibility are both achieved.

The invention provides a power distribution network fault recovery method and device, a storage medium and computer equipment. The method comprises the steps of obtaining operation state data of a power distribution network; determining a fault detection result according to the operation state data in combination with a neural network algorithm and digital twinborn simulation; based on the fault detection result, after a fault area is determined, an isolation measure corresponding to the fault area is applied; and based on the topological information of the power distribution network, calculating a load transfer path, and according to the load transfer path, carrying out load transfer by adopting a flexible interconnection system and a household energy management system so as to recover power supply of a fault area. Therefore, cooperative scheduling of the power distribution network level flexible interconnection system and the user side household energy system is realized, the self-healing capability and the operation efficiency of the power distribution network are effectively improved, and the technical effects of rapid fault isolation and power supply recovery are achieved.

The invention discloses a memory resource processing method, a memory resource processing device, a memory resource processing system and electronic equipment, and relates to the technical field of computers. The method comprises the following steps: firstly, determining a memory resource reservation strategy according to extension equipment configuration information, and finally, performing memory resource reservation according to the strategy corresponding to each hardware equipment, so that the technical problem of unreasonable memory resource allocation of the hardware equipment in the prior art can be solved; therefore, the situation that the extension equipment connected with the downlink port of the Switch switching equipment in the hardware equipment cannot be normally used due to insufficient memory resources is avoided, and the technical effect of improving the performance and the stability of the computing equipment is achieved.

The invention discloses a space-ground integrated dynamic traffic flow analysis and road topology reasoning method and device. A cloud server generates and issues real-time traffic information through multi-level processing such as map matching, traffic flow analysis, prediction and early warning and road topology analysis and updating; specifically, firstly, the high-precision positioning and communication capability of a low-orbit satellite is utilized, data fusion of an inertial measurement unit and a wheel speed sensor is combined, and a multi-source fusion algorithm is constructed to improve the positioning precision; secondly, traffic flow dynamic analysis and prediction are carried out based on a Transform interaction analysis model, and real-time updating of a road topological structure is realized at a cloud end by using a clustering algorithm and a graph theory model; finally, the traffic flow and road topology information is sent to related vehicles and a traffic management system, and real-time monitoring, early warning and optimization of the traffic state are achieved. According to the invention, the accuracy and real-time performance of traffic information acquisition, processing and issuing can be effectively improved, and the intelligent level of traffic management and decision making is enhanced.

The embodiment of the invention discloses a chip verification environment generation method and system, electronic equipment and a storage medium, relates to the technical field of integrated circuit verification, can adapt to verification environments with different configurations, and improves verification efficiency. The method comprises the following steps: constructing or calling a verification environment template; analyzing the initial configuration information and the routing topology information, and generating a target configuration file according to an interface protocol and an interconnection rule of a register transfer level and a model; the routing topology information indicates an interconnection link topology structure between the devices, and the target configuration file indicates a component type corresponding to each device required by simulation forming of each interconnection link and a matched verification component type; loading and instantiating the register transfer level component and the model component based on the verification environment template according to the component type; and according to the verification component type, interconnection is carried out between the register transfer level component and the model component through the verification component, and a chip executable verification environment is generated. The method and the device are suitable for a scene for constructing a chip verification environment.

The invention discloses a distribution network line facility defect detection method based on line magnetic field variable characteristics, and relates to the technical field of magnetic variable measurement. According to the method, high-precision spatio-temporal data are obtained through multi-point synchronous non-contact magnetic field sensing, codes are efficiently preprocessed through self-adaptive voxel division and bytes, and the detection accuracy of the distribution network line facility defects is improved; and the line topology information is deeply fused to Token for representation. A pre-trained large language model is combined with low-rank adaptation (LoRA), sliding window attention and an external memory bank for efficient reasoning, and understanding of long sequences and topological association is enhanced. Characterization quality is improved through mixed mask self-supervised learning, and multi-task (classification, positioning and scoring) output is optimized by adopting uncertainty weighted loss. And privacy protection federal training is supported. And finally, an interpretable diagnosis report is generated, and an alarm and work order process is automatically performed. The accuracy, efficiency, intelligent level and automation degree of distribution network fault detection are remarkably improved, and safe and reliable operation of a power grid is guaranteed.

The invention discloses an electric power Internet of Things equipment anomaly detection method and system based on federal map learning. The method comprises the following steps: constructing graph structure data; in a client layer, using a graph neural network as a local node classifier of a downstream task, and using an adaptive graph interpolation generator to repair sub-graphs and lost cross-sub-graph connections; adopting a plurality of servers as central nodes, and uploading obtained model parameters to the central server of a local area after a client is locally trained to converge; in the server layer, potential global features of covered clients are obtained according to an automatic encoder, model parameters are allowed to be transmitted between adjacent servers, and a global model is obtained by adopting an aggregation algorithm and issued to each client; and after each client updates the model, the next round of training is continued until the optimal detection model is obtained. Potential connections between sub-graphs are mined by using an adaptive graph interpolation generator, and the blank of lack of cross-client topological information is filled.

The invention discloses an electric power collaborative operation method, system and device based on new energy and a medium, and belongs to the technical field of electric power collaborative operation, and the method comprises the steps: obtaining the real-time output data of a new energy access point and the demand data of a load node, and constructing a collaborative operation unit according to the topological information of a power grid; evaluating the load regulation capability of each cooperative operation unit in combination with the new energy access condition, and generating a first regulation and control instruction; when the new energy load regulation capacity in the cooperative operation unit is insufficient, a second regulation and control instruction is generated; the scheduling process is monitored in real time, the cooperative operation strategy is dynamically corrected according to the difference between the new energy output and the load demand, and the overall operation efficiency of the power system is improved. According to the method, the cooperative operation unit is constructed, new energy output prediction and load regulation capability evaluation are fused, local quick response and cross-regional intelligent cooperation of new energy fluctuation are realized, and the operation flexibility and new energy consumption capability of a power system are effectively improved.

The invention discloses a deburring method combining a 2D image and a 3D point cloud, and relates to the technical field of image processing, and the method comprises the steps: carrying out the point cloud registration under the verification of the surface orientation consistency of a standard point cloud and an actual point cloud; constructing a pixel mapping matrix between the workpiece image and the RGB image; semantic-level edge coarse extraction is carried out on the workpiece image, and extraction of a sub-pixel-level actual edge contour is carried out on the basis of coarse extraction in combination with a traditional edge detection algorithm; aligning the actual edge contour to the actual point cloud through the pixel mapping matrix; and the target polishing surfaces are aligned to the actual point cloud, to-be-polished edge contours are screened based on the distances between the contours of all the target polishing surfaces in the actual point cloud and all the actual edge contours, and a polishing track is obtained in the workpiece image. According to the method, the sub-pixel edge positioning capability of the 2D image and the spatial topology information of the 3D point cloud are fused, and the inherent defect of a single mode is overcome.

The invention relates to the fields of computer graphic processing, industrial software technology and the like, discloses a Cocos engine-oriented pipeline loop perception linear decomposition method, and aims to solve the problem that a complex topological graph which cannot be effectively processed and comprises a high-density nested loop in the prior art. The structure is completely and accurately decomposed into a linear drawing instruction sequence which can be accepted by a Cocos engine, and lossless conversion of original topological information is ensured; comprising the steps that after undirected topological graph data are input, a ring detection algorithm is called to recognize and record all independent ring structures in an undirected topological graph and nodes and edges contained in the independent ring structures; judging whether at least one ring is detected in the previous step, and if the at least one ring exists, assigning a determined traversal direction for each detected ring according to a preset ring direction rule; and if the ring does not exist, skipping, and directly performing depth-first search traversal on the undirected topological graph data.

The invention discloses a traffic scheduling method and an electronic device, and relates to the technical field of traffic scheduling, and the method comprises the steps: determining the priority weight of a micro-service, and predicting a target traffic according to the historical traffic information of a network device; constructing a graph model according to the topological information of the network equipment and the dependency relationship of the micro-service, and performing embedded learning on nodes in the graph model to generate a state vector representing a network state; the priority weight, the state vector and the target traffic of the micro-service serve as input of a reinforcement learning model, and a traffic scheduling strategy of the network equipment is obtained; performing iterative search according to iterative particles formed by encoding the strategy network parameters of the reinforcement learning model and the feature learning network parameters of the graph model to determine reinforcement learning model parameters; and issuing the traffic scheduling strategy to the network equipment and executing the traffic scheduling strategy so as to solve the technical problem that a traffic scheduling method in related technologies is difficult to adapt to a dynamic and complex network environment and service requirements under a micro-service architecture, and the reliability of traffic scheduling is improved.

The invention relates to a distributed photovoltaic transformer area metering data abnormity tracing method, belongs to the field of intelligent power distribution networks, and solves the problem of low metering abnormity user positioning efficiency caused by topological information loss of a low-voltage photovoltaic transformer area. Comprising the following steps: acquiring electrical data of a distribution transformer and each user when a distributed photovoltaic transformer area has no photovoltaic output at night, and constructing a transformer area trunk line approximate topology; constructing an inter-node final adjacency matrix between nodes on the basis of the acquired voltage data of each user in the afternoon photovoltaic high output time period of the transformer area; clustering based on the final adjacency matrix among the nodes to divide the users into a plurality of user clusters; constructing a cluster power curve for each user cluster, and calculating the fluctuation consistency between the voltage of each user in the cluster and the cluster power curve of the cluster to which the user belongs based on the approximate topology of the main line of the transformer area so as to identify a suspected abnormal cluster; and performing node-level identification on the suspected abnormal cluster to obtain an abnormal metering user. Automatic traceability and accurate positioning of abnormal measurement data are realized.

The invention relates to the technical field of deep learning model training optimization, in particular to a fine-grained resource scheduling method and system in DNN model parallel training, and the method comprises the steps: converting a DNN model into calculation graph representation, and constructing a calculation graph in combination with the topological information of calculation equipment; modeling calculation stage division in a DNN model training process and resource mapping from each calculation stage to calculation equipment as a stage division and resource mapping combined optimization problem; and iteratively solving the combinatorial optimization problem by using a heuristic algorithm, and optimizing the computing equipment resource mapping of each computing stage by dynamically adjusting the weight of constraint conditions and minimizing the overhead and load difference of the computing equipment. According to the method, the calculation cost, the communication overhead, the overall load balancing and the solving time are comprehensively considered in the target function, efficient calculation resource allocation and task scheduling optimization are achieved, the training and reasoning efficiency of a large-scale deep learning model is improved, and the method has a good application prospect in the field of deep neural network distributed parallel training.