8results about How to "Avoid over-occupancy" patented technology

The invention provides a data development debugging method and device, equipment, a storage medium and a program product. Relates to the technical field of data development. The method comprises the following steps: in response to a received debugging request initiated by a user through a data development platform, allocating computing resources based on the debugging request and starting an interactive debugging environment instance; the debugging request comprises a Python operator code to be debugged; executing the Python operator code to be debugged in the interactive debugging environment instance, and displaying a code execution result and a variable value; and backfilling the debugged Python operator code in the interactive debugging environment instance into a static code editor of the data development platform. According to the method provided by the invention, the code execution state can be checked in real time, the code writing and debugging process is deeply integrated, environment switching and manual operation are avoided, and human errors are reduced; through a reasonable resource management and scheduling mechanism, it is ensured that computing resources can be fully utilized in the starting and running process of the interactive debugging environment instance, and resource waste is avoided.

This invention relates to a load-balancing-based constellation communication-navigation coupled task scheduling method, belonging to the field of aerospace mission planning technology. It primarily improves resource allocation efficiency under equipment-level constraints and addresses the challenge of ensuring communication task completion while meeting the geometric accuracy requirements of navigation services. The invention uses satellite ephemeris data and mission requirement parameters as input to construct a multi-objective mixed integer linear programming (MILP) model. The optimal binary allocation variables are then obtained through a MILP solver, and the service satellites and service time slots for each mission are extracted to generate a scheduling table. This invention avoids the problem of navigation mission failure caused by communication tasks occupying multi-satellite convergence windows in separate planning. Simulations show that the overall mission completion rate is improved by an average of 12.3 percentage points, and the load variation coefficient is reduced by 66%–88%.

The invention relates to the field of data processing, in particular to an adaptive computing power segmentation scheduling method for large model reasoning tasks, which comprises the following steps of: synchronously acquiring characteristics of common dimensions of a plurality of tasks to be reasoned and computing power cluster equipment; clustering the tasks, setting a plurality of batch size schemes according to the tasks, and dividing task batches; for each scheme, the tasks are matched with equipment according to the batch sequence, the matching degree of batches is calculated, and the weights are determined by the ratio of the feature weights to the same-dimension features of the equipment tasks; after matching is completed, counting the equipment computing power, the memory utilization rate and the balance degree, and obtaining a scheme evaluation index in combination with the matching degree; and comparing the scheme evaluation indexes under all the processing modes, and selecting the highest scheme and the matching result thereof as a final scheme. According to the method, the multi-dimensional features and multiple processing modes are comprehensively considered, the optimal scheme is selected for task and equipment matching, the task requirement can be better met, the processing time of the task on the equipment is shortened, and therefore service delay is reduced.

The application discloses a kind of mobile multi-view picture data acquisition methods based on graph neural network, comprising the following steps: S1, collect multi-view picture data and carry out repair preprocessing, construct preliminary image dataset;S2, the image is divided into several image blocks, and input in the improved ViT model and carry out preliminary processing, according to the processing weight of each image block dynamic allocation according to task priority;S3, the arrangement order of image block is optimized in combination with bird swarm algorithm, maximizes labeling accuracy and minimizes the consumption of computing resources;S4, the optimized image block arrangement order is input into the improved ViT model again, and the labeled image block is obtained;S5, the relationship between the labeled image block is modeled using graph neural network, and the image block relationship graph is generated;S6, according to the labeled image block and image block relationship graph, generate final image dataset.The application combines graph neural network, improved ViT model and bird swarm algorithm, etc., and optimizes multi-view image data acquisition and labeling.

The application provides a server, a display device and a live source import method. First, a mapping table based on geographical position and network operators is used for accurate query to filter available live sources in line with the local network environment of a user, avoiding misjudgment across regions. For unknown or unplayable live sources, the server generates a distributed detection task and distributes it to a second display device in an idle state in the same region to detect availability through a lightweight HTTP request, which not only avoids the IP blocking risk of centralized scanning, but also improves detection accuracy by using the local network environment. The available live sources detected are mapped to unique identification codes after semantic normalization processing and are dynamically updated to the mapping table to form a continuously optimized data closed loop. Through resource optimization strategies, the entire process minimizes interference with user devices, and finally provides the user with an accurate and de-duplicated list of available sources, improving determination efficiency and user experience.

The application provides an accounting data management method, system, device and medium based on edge computing, which extracts a structured feature vector of newly added accounting voucher data; extracts a similar voucher set similar to the newly added accounting voucher data from a high-frequency subject area and a sparse subject area in historical accounting voucher data according to a data partition granularity; determines the distance between the structured feature vector of the newly added accounting voucher data and the feature vector of each accounting voucher in the similar voucher set, and then generates a risk confidence value of the newly added accounting voucher data based on the occurrence frequency of similar accounting vouchers and the feature vector distance; and when the risk confidence value exceeds a set threshold, updates the compliance verification result and the structured feature vector corresponding to the newly added accounting voucher data to an accounting voucher database. The technical scheme provided by the application can overcome the interference caused by the natural long-tail distribution characteristics of accounting subjects on similar voucher retrieval in accounting data management.

The invention provides a smart park resource computing power scheduling optimization method and device, equipment and a medium, and belongs to the technical field of resource scheduling. Determining a target computing power node group from the plurality of computing power node groups based on the computing power demand and the task priority of the computing power request task and the predicted load data of each computing power node; screening an optimal execution node based on the predicted load data, the task adaptation degree and the node health degree of each computing power node in the target computing power node group, and distributing the computing power request task to the optimal execution node for execution; and in response to detection that the energy consumption of the optimal execution node exceeds a preset energy consumption threshold value of the node type corresponding to the optimal execution node, or the task execution delay of the computing power request task is greater than a preset delay duration, migrating the computing power request task based on the task priority. The resource utilization rate can be effectively improved, and idle or excessive occupation of resources is avoided.

This invention belongs to the field of rail transit technology, and particularly relates to a train turnaround control method. It includes: after a first train obtains turnaround resources according to its travel path, it controls the switch to the reverse position; after controlling the switch to the reverse position, it begins to turn onto the route, clearing the turnaround resources in the area its tail has passed through, gradually releasing the cleared turnaround resources until the first train has completely cleared the movable area in front of the switch; after the first train has completely cleared the movable area in front of the switch, it controls the switch to the fixed position. This invention, by further refining the switch area, fully utilizes and controls the switch resources available to the train, thereby improving train turnaround efficiency.