Research on Artificial Intelligence Platform for Power Dispatching Based on Cloud Native Technology

The power system generates a huge amount of data and requires high real-time performance. At every moment, there is a large amount of data such as power generation, electricity consumption, and equipment operating parameters that need to be processed. Artificial intelligence platforms need to have efficient data processing capabilities to meet the needs of real-time scheduling, otherwise scheduling decisions may be delayed due to data processing delays. Therefore, this article introduces cloud native technology to design an artificial intelligence platform for power dispatching. Cloud native technology supports distributed storage systems, dispersing data across multiple nodes to avoid the problem of insufficient capacity on a single storage device. Using this technology, MapReduce computing, HDFS distributed storage and other services are packaged into an independent and portable container image, and deployed in a Kubernetes cluster to achieve dynamic allocation and scaling of power resources. A data fusion module based on artificial intelligence algorithm design platform. Based on this, design a power dispatch model and introduce particle swarm optimization algorithm to solve the model. The experimental results show that under the application of the research platform, the power grid loss is lower, and the fluctuations in power load and power are more in line with the ideal state.