In recent years, the burgeoning amount, speed, and variety of data in multiple energy fields have generated new requirements that far exceed those of traditional database systems and especially pose new challenges related to sustainable energy analytics. Classical relational database architectures are usually not capable of satisfying the needs of performance, scalability, and energy efficiency of emerging energy platforms for smart grids, renewable energy forecasting, and real-time monitoring. The study proposes a high-level database optimization framework targeted at improving the performance of large-scale energy analytics infrastructures. Specifically, the system combines the Adaptive Query Optimization (AQO) and Partition-Aware Load Balancing (PALB) to address the issues of query delay, storage limit, and energy consumption. AQO dynamically optimizes execution plans by considering workload statistics and planner feedback, while PALB balances query distribution in concurrent scenarios based on system resource metrics. We demonstrate via experiments on a high-performance computing platform that our DWPRF achieves significant improvements in various aspects, such as query execution time reduction, advanced compression to realize higher storage efficiency, reduced energy consumption, and better scalability of the system under high concurrency. The experimental results indicate the feasibility of scheduling and allocating plans when addressing these challenges and show the power of using intelligent planning with resource-aware execution to optimize databases for energy informatics. Moreover, the study outlines potential avenues for future extensions, including machine learning-driven optimizers and distributed deployment at the edge and cloud. The proposed approach provides a solid basis for constructing high-performance, energy-efficient data management systems, which are a fundamental requirement for sustainable energy systems.
Alshareefi et al. (Thu,) studied this question.