Active decision-making method and system for resource capacity expansion in deep learning

Abstract

The invention provides an active decision-making method and system for resource capacity expansion in deep learning, and the method comprises the steps: S1, obtaining pre-reserved prediction data and historical operation and maintenance data corresponding to the prediction data; S2, comparing the pre-reserved prediction data with the corresponding historical operation and maintenance data, and determining whether the pre-reserved prediction data is in a normal range or not according to a comparison result; S3, when the pre-reserved prediction data is in a normal range, calling real-time business index data, and optimizing the pre-reserved prediction data according to the real-time business index data; and S4, establishing a simulation model according to the optimized prediction data, and performing active decision on resource capacity expansion through the simulation model to obtain a resource capacity expansion result. According to the method, the simulation model is utilized to enable the resource expansion system to realize intelligent decision making, so that the expected designed hardware architecture can meet the algorithm requirement, the resource optimization capability is strong, the prediction capability is better, and the user experience and the use efficiency are improved.

Description

technical field [0001] The invention relates to the technical field of the Internet, in particular to an active decision-making method and system for resource expansion in deep learning. Background technique [0002] With the continuous development of technology, in the operation and maintenance of power enterprise information systems, they are faced with serious problems of slow fault location and long time-consuming fault processing. For example, nearly a hundred invalid alarms are generated on average every day. It takes an average of more than 30 minutes, and the average fault handling is about 1 hour. In addition to the time-consuming work order approval and dispatching on the dispatch system, the overall processing time is about 90 minutes. [0003] In order to solve the above time-consuming problems, specific systems are generally designed to solve them, such as intelligent operation and maintenance technology to realize the intelligent operation of the data center of...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to propose an active decision-making method and system for resource expansion in deep learning to solve the problem that in existing deep learning algorithms, when making resource expansion decisions, the predicted data is different from the actual operation and maintenance data, which affects the system. Technical issues in normal operation and maintenance

Images