8 results about "Resource leveling" patented technology

The invention discloses an artificial intelligence platform computing power resource scheduling management method and system, and relates to the technical field of artificial intelligence computing power scheduling, and the method comprises the specific steps: hardware topology collection traverses heterogeneous resources, and generates and updates a graph; the collaborative strategy decision is based on the atlas, and optimal precision and parallel combination are screened; load balancing deploying split tasks, matching hardware capability and planning a transmission path; executing and monitoring multi-dimensional real-time tracking; the quantization deviation is dynamically adapted and adjusted, and the strategy is timely optimized; according to the invention, through hardware topology acquisition and collaborative strategy decision, a computing power resource optimal configuration scheme is formed, and the training task starting efficiency and stability are improved; load balancing deployment and dynamic adaptation adjustment are combined, resource balancing allocation and whole-process dynamic optimization are achieved, the problems of resource imbalance and response lag are effectively solved, the computing power utilization efficiency is improved to the maximum extent, and efficient and stable training is guaranteed.

The application relates to the technical field of big model computing power scheduling, in particular to a computing power cooperative intelligent scheduling method and system based on a big model application, which comprises the following steps: obtaining a global computing power task request containing multiple big model inference tasks, extracting multi-dimensional features according to task constraint conditions and forming a task demand vector. An improved greedy algorithm is adopted to match the task demand vector and a state vector of a dynamic heterogeneous computing power resource pool, a preliminary scheduling instruction set is generated through weighted distance, a multi-step forward-looking evaluation model with computing power resource balance and task communication overhead as joint targets is constructed, the running state of tasks and resources in multiple time slices is simulated, the scheduling instruction is corrected, and finally the instruction is issued to a hardware computing unit to complete task execution. The method can improve the matching accuracy of tasks and computing power resources, optimize the balance of resource allocation, and reduce the data communication overhead between tasks.

The present application relates to cloud computing resource scheduling technical field, especially in a kind of real-time virtual machine deployment method and system based on buffer queue ant colony algorithm, comprising: complete computing period is divided into several sub-periods, each sub-period includes scheduling time period and deployment time period;In scheduling time period, the optimal deployment result of current sub-period is solved using buffer queue ant colony algorithm;In deployment time period, according to the optimal deployment result of current sub-period, virtual machine is deployed.The present application can significantly reduce the number of working servers in different scale scenarios and improve resource utilization, effectively solve the energy consumption optimization and resource balance problem under the real-time load of cloud computing.

The application discloses a thermal power building engineering construction optimization method based on BIM technology and belongs to the technical field of thermal power building engineering construction optimization. The method comprises the following steps: collecting BIM geometric parameters and construction time sequence constraints, and constructing a construction information network diagram with a topological connection relationship; extracting multi-dimensional space-time characteristics of each component, wherein the multi-dimensional space-time characteristics comprise position coordinates, construction time length, resource demand quantity and preposition dependency relationship; initializing a construction sequence population by using a differential evolution algorithm, wherein each individual is an integer permutation code meeting dependency constraints; performing construction conflict detection based on the network diagram on the population individuals, calculating an adaptive value fused with a construction total time length, resource balance and conflict penalty; performing mutation, crossover and legality repair operations according to the adaptive value, iteratively updating the population to convergence, and outputting an optimal construction sequence scheme. The method deeply couples BIM space interference analysis and differential evolution, automatically avoids space conflicts while optimizing the construction period and resources.

This invention relates to the field of smart grid inspection and edge computing technology, specifically to a cloud-edge collaborative computing resource balancing scheduling system for unmanned vehicles used in converter station inspections. It includes modules for collaborative management and control, risk mapping, demand prediction, resource scheduling, and feedback. The system identifies computing power sensitive areas and passage zones through physical entropy value correlation analysis. Its core is to combine scenario risk coefficients with visual semantic pre-simulation to determine instantaneous computing power peaks and demand types. When strong real-time, heavy computation, or low-risk signals are generated, edge exclusive locking, cloud high-priority slice offloading, or energy-saving degradation strategies are executed respectively. This invention changes the current scheduling situation that relies solely on CPU load, ensuring that resources are preferentially tilted towards high-risk areas, and realizing dynamic reconfiguration of computing power resources.

The application relates to a regional education resource balanced allocation method, device, equipment and medium. The method comprises the following steps: firstly, obtaining shared resource use time length data of each education institution to form an original log sequence; then, carrying out time sequence analysis on the original log sequence to generate an occupation rate distribution mode of the shared resource in different time periods; based on the occupation rate distribution mode, determining the idle state of the shared resource through a preset idle threshold, adopting a clustering algorithm to perform grouping processing on the resource supply and demand relationship to obtain a demand difference classification result; analyzing and planning a shared resource cross-institution flow path according to the demand difference classification result, and then generating a preliminary flow plan; finally, taking the preliminary flow plan as input, constructing a resource flow simulation model, adopting an iterative optimization algorithm to adjust and optimize the model parameters, and finally outputting a shared resource balanced flow allocation scheme covering cross-institution and cross-time period. The method can realize efficient activation and balanced allocation of regional education shared resources.

This invention discloses a dual-tunnel scheduling optimization method based on a hybrid genetic algorithm and rolling time domain, belonging to the field of tunnel engineering construction management technology. It includes: constructing a multi-objective scheduling model for dual-tunnel construction with the objectives of minimizing the total project duration and minimizing the weighted resource peak value; using a constraint method to transform the multi-objective scheduling model into a resource-balanced single-objective optimization model under a fixed project duration constraint; using a hybrid genetic algorithm to perform static global optimization on the single-objective optimization model; establishing a dynamic monitoring mechanism based on rolling time domain through multi-dimensional IoT sensing; updating the parameters or status of affected processes according to the type of interference event when interference events are detected; using a heuristic local scheduling strategy to reschedule unscheduled tasks after interference, predicting the final project state, and generating a dynamically adjusted construction scheduling scheme. This invention combines static global optimization with dynamic real-time adjustment to achieve dual optimization of project duration and resources.

The application discloses a kind of based on logic guidance and double target ant colony's maintenance scheduling method and system, the method includes: based on maintenance project analysis cluster data obtains discrete queue;Using iterative self-organizing data analysis algorithm to discrete queue iterative clustering obtains elevator team queue;The benchmark solution that elevator team queue satisfies hard constraint logic rule is calculated;To benchmark solution iterative addition soft constraint adjustment rule obtains initial scheduling strategy;Initial scheduling strategy is as hot start pheromone input scheduling optimization model;Management efficiency matrix and geographical efficiency matrix are set to scheduling optimization model and iteratively optimized;Scheduling optimization model outputs Pareto optimal solution set.Thereby, iterative self-organizing data analysis algorithm realizes maintenance resource balanced allocation, with initial scheduling strategy hot start, make convergence time reduce to second level, satisfy high concurrent scene under instant output demand, multidimensional constraint guidance mechanism has excellent global optimization performance, realizes the balanced allocation of manpower and maintenance work order.