Model scheduling optimization method based on artificial intelligence and related equipment

Abstract

The invention provides a model scheduling optimization method and device based on artificial intelligence, electronic equipment and a storage medium, and the method comprises the steps: carrying out the statistics of historical data of edge computing equipment, wherein the edge computing equipment comprises a plurality of algorithm models; counting population data of the edge computing device radiation area; calculating a confidence of the historical data based on the demographic data; integrating the confidence coefficient by using a self-defined integration model to obtain the global confidence coefficient of the historical data; respectively calculating the basic importance and the time directivity of the algorithm model based on historical data, and correcting the basic importance through global confidence; inputting the basic importance and the time directivity of each algorithm model into a user-defined aggregation model to obtain a scheduling value; and comparing the scheduling value with a preset threshold value to obtain a scheduling optimization scheme. According to the method, the scheduling scheme of the algorithm model can be flexibly adjusted according to the alarm times and alarm moments of the algorithm model in historical data, so that the working efficiency of the model is improved.

Description

technical field [0001] The present application relates to the technical field of artificial intelligence, and in particular, to an artificial intelligence-based model scheduling optimization method, device, electronic device and storage medium. Background technique [0002] As the society pays more and more attention to environmental protection issues, various environmental protection policies have been issued in many regions, and the policies related to garbage disposal have received extensive attention from residents because they are closely related to life. [0003] At present, cameras, edge computing devices and image recognition algorithm models are usually combined to supervise the behavior of residents at garbage disposal sites. However, due to the diverse behaviors that need to be supervised, it is usually necessary to deploy multiple algorithm models in the same edge computing device to monitor various behaviors of illegally throwing garbage, and run all algorithm m...

AI Technical Summary

Problems solved by technology

However, due to the diversity of behaviors that need to be supervised, it is usually necessary to deploy multiple algorithm models in the same edge computing device to monitor various illegal garbage dumping behaviors, while continuously running all the algorithm models to monitor the edge computing devices. The performance impact is greater, and it is easy to cause hardware loss, which will result in waste of resources and increase supervision costs.

Images