57 results about "Single task" patented technology

The invention discloses a distributed layout information reading method and system, a storage medium and electronic device.The distributed layout information reading method comprises the steps that a task configuration file is read through a master control node, the task configuration file comprises layout information configuration, and the layout information configuration comprises a layout processing area and a single task processing area; the main control node is used for dividing the layout processing area according to the single-task processing area to obtain a plurality of sub-task areas, and each sub-task area corresponds to one sub-task; adopting the master control node to respectively distribute the plurality of subtasks to the corresponding slave nodes according to the distribution strategy; the slave node sends a layout reading request corresponding to the subtask to a target database based on a load balancing mechanism; returning layout information corresponding to the layout reading request to the slave node through the target database; and performing corresponding task processing on the layout information through the slave node, and returning a processing result to the master control node. The layout information reading efficiency can be improved.

The invention discloses a communication computing parallel optimization method, a multiprocessor system, a medium and a program product, and the method comprises the steps: configuring a first computing core and a second computing core on a single task flow for a general matrix multiplication subtask allocated to a single processor; wherein the first calculation core is used for executing a general matrix multiplication subtask, and the second calculation core is used for executing a set communication task; the set communication task comprises a full accumulation operator or a protocol dispersion operator; then, starting scheduling is conducted on the first calculation core and the second calculation core according to a preset dependency mechanism, so that the general matrix multiplication subtask and the set communication task are executed asynchronously in an overlapped mode; according to the method, the problem of poor reusability of the original Kernel caused by intrusive modification of the GEMM or re-implementation of the Kernel can be effectively avoided through parallel optimization of GEMM calculation and ensemble communication operation, and the performance overhead of the processor is reduced.

The application relates to a multi-round scientific research workflow evaluation system and method for a multi-modal intelligent agent, and belongs to the technical field of large language models and multi-modal large models. The application comprises a multi-round task generation module, a multi-modal information processing module, and the like. The multi-round task generation module constructs a multi-round task scene closer to a real scientific research process. The multi-modal information processing module enables the measured intelligent agent to complete paper active retrieval, multi-modal evidence extraction alignment, cross-paper evidence integration, and round reuse of evidence. The protection points of the application are not limited to a single task sample or a single scoring index, but are focused on a multi-round evaluation benchmark constructed around a real scientific research workflow, a process-level evaluation protocol matched with the benchmark, and an overall implementation scheme supporting landing execution and fine-grained diagnosis of the benchmark.

The invention relates to the technical field of automation control, and discloses a PLC single task to multi-thread task realization method and system, and the method comprises the steps: obtaining PLC task arrangement information which comprises at least one PLC task and a corresponding program organization unit POU; based on the PLC task arrangement information, creating a real-time operating system (RTOS) thread corresponding to each PLC task in the RTOS, and distributing a corresponding stack space for each RTOS thread; when the RTOS task scheduler executes task switching, calling a preset task switching hook function to carry out stack pressing storage on the additional context of the PLC program unit corresponding to the currently running RTOS thread, and carrying out elastic stack recovery on the additional context of the PLC program unit corresponding to the to-be-running RTOS thread, according to the method for transforming the PLC single task into the RTOS-based multi-thread task, the practical requirement that multi-thread task transformation needs to be carried out due to the fact that the real-time performance of a single task mode of a traditional PLC manufacturer is limited is met.

Systems and methods described herein provide a computer-implemented method for process simulation using a multi-agent system. A software agent is trained using historical process data associated with a process. Each software agent represents a single task in the process. Based on the candidate process model, the process is simulated by executing the multi-agent system to generate simulation results. The software agent operates autonomously within the multi-agent system based on the trained behavior. The process is also performed using a runtime configuration based on the candidate process model. Execution of the process generates runtime results for comparison with the simulation results. In some examples, the candidate process model is adjusted based on the comparison, and simulation and execution of the process are repeated using the adjusted multi-agent system and the adjusted runtime configuration, respectively. A deployment process model is generated based on the adjusted candidate process model.

The application provides a node energy consumption optimization method, device, equipment, medium and product of a supercomputing cluster. The method comprises: obtaining resource occupation data corresponding to a task flow; predicting unit task energy consumption when the task flow is deployed on each computing node in the supercomputing cluster based on hardware configuration information and the resource occupation data of each computing node; determining a target computing node from the plurality of computing nodes according to the unit task energy consumption corresponding to each computing node, and distributing the task flow to the target computing node for execution; and adjusting physical operating parameters of the target computing node according to task load changes during task execution, so as to reduce overall operating energy consumption and achieve node energy consumption optimization. The two-stage collaborative control of the scheduling stage and the execution stage can realize fine energy consumption control for tasks, and reduce single task execution energy consumption and overall operating energy consumption of the supercomputing cluster.

The embodiment of the invention provides display equipment. In the embodiment of the invention, the display equipment receives a selection instruction of a multi-window template under the condition that the currently played media asset is a screen projection media asset picture, and supports the screen projection media asset which is played previously to be continuously played based on the screen projection instance by multiplexing the screen projection instance in a single task mode in a task stack. And supporting screen projection media resources indicated by the screen projection instruction sent by the user later based on the screen projection instance. And the multi-window layout in the multi-window playing mode is adjusted based on the screen projection instance, so that mutual shielding among the multiple windows is avoided. On the basis of reuse of the screen projection instance, the problem of playing interruption of previous screen projection media resources caused by screen projection signal conflicts in a multi-window mode is effectively avoided.

The invention relates to a space crowdsourcing task allocation method based on trajectory prediction, and the method comprises the steps: constructing an MMLSTM model based on a meta path and a Mogrimer LSTM module in combination with end-to-end training, and enabling the MMLSTM model to be used for predicting a user trajectory; and according to the predicted user track, adopting a dynamic strategy of combining grouping insertion and single task insertion to distribute the published new task to the corresponding user on line. Compared with the prior art, on one hand, by introducing the space-time diagram and the multi-type meta path, the complex space-time interaction relationship between the users can be explicitly and structurally modeled, and by combining with the Mogrifier LSTM, the time sequence evolution can be dynamically captured, so that the prediction of the future trajectory of the user is more accurate; and on the other hand, a dynamic strategy of combining grouping insertion and single task insertion is adopted, geographically adjacent tasks are packaged and allocated, and an access sequence is optimized, so that the total detour overhead when a user executes a plurality of tasks can be remarkably reduced.

The embodiment of the application belongs to the technical field of laser processing, and relates to an intelligent processing method, comprising the following steps: prestoring at least two processing modes; acquiring a task cycle size and a single task maximum size; calculating a processing width according to the acquired task cycle size and single task maximum size, and selecting one processing mode from the prestored processing modes according to the processing width; moving a marking device to a starting position of the marking device and moving a cutting device to a starting position of the cutting device according to the selected processing mode; controlling the marking device to mark according to the selected processing mode, and controlling the cutting device to cut according to the selected processing mode. The application further provides an intelligent processing device, a computer device and a storage medium. The method can adapt to batch processing of different tasks, has strong adaptability, does not need human participation and equipment replacement, and saves the cost of producing products of different sizes.

The invention relates to a reservoir water resource management technology, discloses a reservoir water temperature machine learning prediction method considering reservoir area vertical-discharged water temperature linkage, and solves the problems that numerical simulation in an existing reservoir water temperature prediction method is difficult to adapt to a dynamic boundary; and a machine learning single-task model does not consider the cascade relationship between the vertical water temperature and the discharged water temperature of the reservoir area, so that the generalization ability is insufficient, and the prediction efficiency is low. According to the scheme, firstly, monitoring data related to the water temperature of the reservoir are collected and preprocessed, and a data set is obtained; then, constructing an LSTM model architecture of'shared feature extraction + dual-task output '; then training the dual-task LSTM model by using the data set; in practical application, historical time sequence data of the preprocessed water temperature correlation indexes are input, and vertical water temperature data and discharged water temperature data in a future time period are obtained based on the trained double-task LSTM model.

A system and method for a using a genetic algorithm with an extra bit in the genome to obtain a best-fit of resource combinations for a particular task to be scheduled. The system can generate a genome from a plurality of jobs and then add an extra bit that can be used to find the best resource combination in the event that multiple resource combinations are equally suitable as an initial fit.

The invention discloses a CTC simulation practical training emergency disposal process task issuing method and system, and relates to the technical field of railway dispatching simulation practical training, and the method comprises the following steps: S1, a teacher imports a scheduling table of a dispatching station into a CTC simulation practical training system for visual interface display; s2, based on visual interface display, the teacher formulates a reasonable emergency disposal process task for the student and sets task basic information; s3, the teacher stores, modifies, exports and deletes the emergency disposal process task; s4, issuing an emergency disposal process task corresponding to the trainee, and enabling and completing the task by the trainee; and S5, carrying out training task training condition statistics and query. According to the invention, tasks can be distributed in a personalized manner, the training time and training tasks of trainees are reasonably arranged, and the flexibility of training task configuration and the objectivity of evaluation are improved. The problems that a traditional system is single in task issuing mode, generally, only the same task content can be set for all trainees, and personalized training requirements cannot be met are solved.

The application discloses a cognitive risk stratification method based on a dual-task gait, which comprises the following steps: acquiring single-task and dual-task walking gait data of a target subject in a longitudinal tracking population cohort; calculating gait loss data of the dual task compared with the single task; performing prognosis correlation analysis in combination with cognitive function decline events in a preset time period, screening target features, determining an optimal cutoff value and a risk stratification determination standard; and completing cognitive risk grade division and outputting a result based on the standard. The method solves the problems of traditional cognitive evaluation, such as complicated operation and difficulty in large-scale application, realizes precise and non-invasive early screening of cognitive risk, and has important clinical application value.

The embodiment of the application provides a kind of algorithm warehouse management method and system based on service tool chain, belong to algorithm warehouse management technical field.The method is matched resource demand by task category identification, generates hardware resource available parameter sequence in combination with the dynamic resource occupation prediction of parallel task of execution terminal, the first screening action is based on the resource fitting degree calculation matching degree per time, substantially improve the resource adaptation accuracy of algorithm model and execution terminal, reduce resource waste, the second screening action is combined with the historical execution information of target service terminal and non-target service terminal, both guarantee the confidence of single task to meet standard, also through service terminal group confidence distribution adjustment execution terminal selection, ensure the overall execution stability of system.Based on the confidence ratio of service terminal group, the execution terminal selection range is dynamically adjusted, the intelligent management of system confidence and computing power reserve is realized, extreme imbalance is avoided, and the adjustability of long-term operation is guaranteed.

The invention discloses an order type task scheduling optimization method and system, and relates to the technical field of task scheduling and control optimization, and the method comprises the steps: constructing an order type task scheduling control model; by constructing a dual-objective function of total delay time minimization and overall execution efficiency maximization, global optimization of task resource allocation is achieved, it is clear that task execution time resources do not exceed the maximum processing duration, and the situation that due to excessive optimization, resources are occupied by a single task for a long time, and the task execution efficiency is improved is prevented. Two objective functions of total delay time and average processing efficiency are constructed, the total delay time objective function can reflect the influence of the task queuing length on the execution efficiency, a penalty term is introduced into a scheme violating constraints, the evaluation weight of the scheme is reduced, it is ensured that the generated scheduling scheme meets both the optimization objective and the actual constraint condition, and the scheduling efficiency is improved. The frontier solution set is generated through iterative solution, conversion of task planning from static rules to dynamic optimization is achieved, and an efficient and reliable solution is provided for an intelligent task management system.

The invention relates to the technical field of intelligent manufacturing, in particular to a vehicle repair store operation system and method using an intelligent station, and the system comprises a scheduling data processing module, a delay evaluation module, a scheduling adjustment module, a priority decision module, and a scheduling execution module. According to the method, quantitative evaluation and grading of the station delay state are achieved by calculating the handover time difference between the operations and comparing the handover time difference with the standard period, the chain reaction range possibly caused by single task delay can be recognized in a prospective mode, and then the adjustment requirement is focused on the affected follow-up tasks; unnecessary global disturbance on an overall plan is avoided, meanwhile, the emergency degrees of the tasks and the time proximity after adjustment are comprehensively evaluated, dynamic station time slot occupation weights are generated, all the tasks to be adjusted are globally sorted, the limitation of a traditional fixed sequence is broken through, and the task scheduling efficiency is improved. And finally, according to the dynamic priority of the task and the actual operation duration, automatically searching and matching an optimal idle station time slot.

The application discloses a kind of multi-task method and system based on multi-modal continuous learning, the method includes: obtaining multi-modal data and constructing multi-modal dataset;Multi-modal dataset is mapped to unified language model representation space, and unified semantic feature representation is obtained;Unified semantic feature representation is input into sparse mixed expert layer, and the previous expert is based on token level sparse activation by router, and forward calculation is carried out by self-attention and expert feedforward neural network;In continuous learning, expert activation frequency and average gradient norm are regularly counted, and low active cold experts or new experts are dynamically pruned, and sparse mixed expert layer is optimized;Decoupling low-rank adaptation mechanism is used, historical task direction is frozen, and only new task learning direction and amplitude are updated;Finally, the prediction result of classification, question answering or generation task is outputed.The application uses the above method, and breaks through the bottleneck of poor modality adaptability and single task of traditional multi-modal model.

Systems and methods described herein provide a computer-implemented approach for process simulations using multi-agent systems. Software agents are trained using historical process data associated with a process. Each software agent represents a single task in the process. Based on a candidate process model, the process is simulated to generate simulation results by executing a multi-agent system. The software agents operate autonomously within the multi-agent system based on trained behavior. The process is also executed using a runtime configuration that is based on the candidate process model. The execution of the process generates runtime results for comparison with the simulation results. In some examples, the candidate process model is adjusted based on the comparison, and the simulation and the execution of the process are repeated using an adjusted multi-agent system and an adjusted runtime configuration, respectively. A deployment process model is generated based on the adjusted candidate process model.

The application provides a multi-role convention and exhibition scene-oriented intelligent reception scheduling and information collaboration method and system, aiming to solve the problems of information fragmentation, high dependence on manual scheduling and difficulty in multi-role collaboration in the prior art convention and exhibition reception. The application includes constructing a multi-role account and a resource pool; converting collected guest itinerary information and personalized needs into a structured form of a multi-tuple; intelligently scheduling based on the multi-tuple to generate a reception task scheme containing vehicle pick-up, accommodation and catering; visualizing the reception task scheme and differentially assigning it to corresponding executors and guests, supporting task state updating and rescheduling mechanism. The application not only schedules a single task, but also links the reception arrangement and state change of the three business boards of pick-up, check-in and catering, and realizes information synchronization of multiple roles.

The invention discloses an airspace task cooperative processing method and system based on multi-source target analysis, and belongs to the technical field of multi-source target analysis. Task execution information of task units is continuously collected in each data sampling period, a spatial position coordinate data sequence and a speed data sequence are constructed, and time weighted dynamic trajectory center parameters of a single task unit in a single data sampling period are calculated; calculating the dynamic space coupling degree of the adjacent task units in the data sampling period; calculating the task weight and the comprehensive coupling strength of a single task unit; acquiring a speed change rate of a single task unit, and calculating a task dynamic stability coefficient and a task hierarchy division threshold value; based on the task hierarchy division threshold value and the task dynamic stability coefficient, the stability of the task unit is analyzed, and cooperative processing is performed, so that an intelligent task processing system capable of adapting to airspace dynamic change in real time is realized, and the accuracy, robustness and overall operation efficiency of airspace task scheduling are remarkably improved.

The invention discloses an MCP-based question and answer system dynamic collaborative scheduling method and system, and belongs to the technical field of artificial intelligence. The method comprises the following steps: starting a task through a first workflow, and calling a standardized server module to perform multi-source data acquisition to form a double-source data set; the intelligent agent receives the data for reasoning, dynamically decides whether to call a tool for iterative optimization or not, and generates a final answer; and judging and executing memory updating through the second workflow to realize a knowledge closed loop. The system adopts a client-server architecture, standardized communication is realized through an MCP protocol, and a workflow execution mode and an intelligent agent execution mode are fused in a client. According to the method, the problems of low module cooperation efficiency, poor compatibility and single task mode in the prior art are solved, the expansibility, the complex task processing capability and the answer accuracy of the system are remarkably improved, and the method is suitable for enterprise-level large-scale application scenes.

The invention discloses a traffic scheduling method and system in a distributed training task scene, which are used for distributed training tasks between a first device end and a second device end. The method comprises the following steps: according to a communication demand of a distributed training task, performing bipartite graph abstraction and graph coloring processing on a first device end, a second device end and network resources between the first device end and the second device end in sequence, establishing a communication model, and establishing a traffic scheduling optimization problem in combination with a first traffic scheduling target and a second traffic scheduling target; the first flow scheduling target selects paths flowing through different core switches for communication flows with the same source address or target address in any communication stage of a single task, and the second flow scheduling target selects paths flowing through different core switches for communication flows with the same source address or target address among different tasks; and solving a traffic scheduling optimization problem to obtain a path set and complete routing. According to the method, the flow conflicts in the tasks and among the tasks in the distributed training task scene are reduced.

The application discloses an accident investigation report crawling method and system based on multi-agent dynamic cooperation, which comprises the following steps: receiving a target website to generate a task queue, mapping a global task description to a global semantic representation vector; based on the vector, the task is disassembled, a sub-task exclusive vector is generated, and a localized multi-agent cluster covering a heterogeneous crawling whole process is constructed; based on an event-triggered communication mechanism and dynamic switching of three cooperation modes, the interaction of each agent is controlled to complete single task execution; the core fields of the report are extracted, a standardized file name is generated, a data and system double-layer consistency constraint mechanism is constructed for verification and closed-loop feedback optimization; and the structured archiving of the verified file is completed to generate a result account book. The application realizes complete localization of sensitive data security processing, effectively breaks through the limitation of heterogeneous webpage structure, and improves the accuracy and full automation level of safety production data acquisition.

The application discloses a task resource coordination and distribution method for unmanned cluster command control, and belongs to the technical field of unmanned cluster command control, and comprises the following steps: extracting all unmanned cluster tasks and corresponding requirements of each task, extracting task activity sequences of the tasks and resource requirements of each task activity in sequence based on respective meta-task models of the tasks, calculating feasible unmanned clusters of each single task based on the resource requirements and available task resource sets, obtaining a multi-task candidate unmanned cluster list, when there is a resource contention, realizing automatic coordination and distribution of the contended resources among unmanned clusters based on overall benefit maximization of formation tasks, and when there is no resource contention, selecting an unmanned cluster with the highest comprehensive evaluation value as a task resource distribution result. The application can realize a task resource distribution method suitable for unmanned aerial vehicle clusters, unmanned ship clusters and unmanned underwater vehicle clusters in a multi-domain and multi-task condition.

The invention relates to a task scheduling method and device, electronic equipment and a storage medium, and relates to the technical field of task scheduling. The method comprises the steps that a task flow set is obtained, and the task flow set comprises a plurality of task flows with a dependency relationship; according to the dependency relationship, determining an upstream and downstream relationship of each task flow; for a single task flow, in response to determining that the execution state of the upstream task flow of the task flow is successful execution, executing the task flow; in the execution process of the task flow, for any subtask in the task flow, determining whether a task instance of the subtask exists or not; in response to determining that the task instance of the subtask exists, determining whether the subtask is successfully executed or not according to the task instance of the subtask; and in response to determining that the subtask is successfully executed, directly obtaining an execution result of the subtask. By applying the method, the task execution efficiency is effectively improved, and resources are saved.

The invention discloses an unmanned cluster cooperative task allocation method, and the method comprises the steps: constructing a multi-unmanned-aerial-vehicle cooperative task allocation model through constructing a three-objective optimization function which comprehensively considers the task execution efficiency, the resource consumption cost and the task comprehensive income, and combining a task resource demand and a single-machine single-task constraint; a dynamic task importance evaluation mechanism is introduced, and task priorities are updated in real time based on a distance attenuation model and target motion vector analysis; an improved genetic algorithm which introduces multi-modal initialization, a hybrid selection strategy, a self-adaptive crossover mutation operator and a constraint repair mechanism is adopted to carry out iterative solution on the allocation model, and an optimal task allocation scheme which meets multi-target and multi-constraint even conditions is efficiently generated. According to the method, an efficient, feasible and stable cooperative task allocation scheme can be provided for the unmanned cluster under the conditions of limited resources and dynamic task change.

This invention discloses a method, device, and medium for pooled scheduling of small-scale intelligent computing resources, relating to the field of resource management technology. The method includes: semantic parsing and resource requirement identification of real-time business requests based on an initial resource status list, and organizing task attributes to generate a task profile list; real-time comparison and optimization of candidate scheduling schemes to form a target execution scheme; task distribution and process control based on the target execution scheme and the task profile list, with dynamic adjustments during execution to obtain an execution result package; resource reclamation and scheduling feedback updates of the target execution scheme based on the execution result package, and correction of subsequent scheduling strategies to form the next round of resource status list. This invention improves the success rate and timeliness of single tasks and even mixed task flows by continuously monitoring the task execution process.

The application discloses a kind of based on production condition perception's advertisement material production scheduling method and system, it is related to advertisement material manufacturing management technical field.The kind of based on production condition perception's advertisement material production scheduling method and system, including S1, collection and pre-processing advertisement material production process in execution time related data and resource queue data, constructs standardization production condition dataset;S2, analysis rework influence in production process in the degree of conduction, adjustment execution arrangement in production process;S3, assesses the rhythm stability degree of production process, linkage constraint production execution arrangement in time window occupancy and task release rhythm;S4, assesses the adjustable degree of task time landing point, corrects task plan start time.It solves the problem that existing advertisement material production management relies on single task average duration scheduling, lacks the depiction of rework timing aggregation and cross-node conduction disturbance, leading to scheduling deviation accumulation and causing resource occupancy imbalance.

The invention discloses a tool request calling method for an agent, which comprises the following steps: acquiring a tool calling request of the agent, and if the request belongs to a request of a single task, acquiring the current LLM calling quota number of the task, the LLM calling quota number being used for limiting the total number of times of calling LLM by the agent in the single task; if the LLM calling quota number is not zero, initiating LLM calling to generate a tool calling instruction; the tool calling instruction is responded, the current tool calling quota number is detected during response, and if the tool calling quota number is not zero, a tool corresponding to the tool calling request is called. By means of two-dimensional limitation, LLM calling and tool calling are limited at the same time, all-around cost control is achieved, compared with an unlimited system, the cost can be controlled within a budget range, and abnormal consumption is avoided.

The embodiment of the invention provides an algorithm warehouse management method and system based on a servitization tool chain, and belongs to the technical field of algorithm warehouse management. According to the method, resource requirements are matched through task category identifiers, dynamic resource occupation prediction of parallel tasks of an execution terminal is combined, a hardware resource available parameter sequence is generated, and the matching degree is calculated based on the moment-by-moment resource fitting degree through a first screening action, so that the resource adaptation precision of an algorithm model and the execution terminal is greatly improved, and resource waste is reduced; the second screening action is combined with historical execution information of the target service terminal and the non-target service terminal, it is guaranteed that the confidence coefficient of a single task reaches the standard, execution terminal selection is adjusted through service terminal group confidence coefficient distribution, and the overall execution stability of the system is guaranteed. And the selection range of the execution terminal is dynamically adjusted based on the confidence ratio of the service terminal group, intelligent management of system confidence and computing power reserve is realized, extreme imbalance is avoided, and adjustability of long-term operation is guaranteed.