Scholarlink

Demand Response alleviates demand curves in intelligent grids during challenges in the power systems. Smart grids with the use of Artificial Intelligence (AI) optimize load prediction, grid stability, fault detection and security. The future power systems are based on these technologies which are powered by artificial intelligence, big data and IoT. Nevertheless, in detecting smart grids AI has some challenges in achieving full potential. The present literature includes standalone AI approaches that are not combined to optimize their effectiveness and system reliability is also an issue. In this paper, the review is done on AI technologies, deep learning and machine learning of smart grids to comprehend their efficiency in energy generation, distribution, and consumption optimization. The review studied the use of AI in load forecasting, stability analysis, fault detection, and demand side control, considering such methods as stochastic optimization and robust optimization and power system expansion simulation. Maintenance prognostic and real time decision making are methods of improving system durability in energy systems. Deep learning methods are effective in optimization of operations. Findings indicate that AI models that are supported by smart grid infrastructure enhance effectiveness and reliability. AI frameworks of secure systems should be adopted by energy managers. The researchers have not validated AI frameworks in real-time, which is a limitation of the study. Quantitative analysis has found out that the number of citations per document is 32.74 and 14.32/year with an increasing impact of the field. As the scientometric mapping shows, hybrid AI-neuro-evolutionary algorithms are much better than conventional optimization frameworks, with more than 20% higher benchmarked stud-ies predictive accuracy and 15% less cost. Future studies are important to deploy adaptive neuro-evolutionary models in live smart-grid systems to evaluate scalability and interoperability and resiliency to dynamic demand conditions. The combination of cybersecurity, explainable AI, and ethical frameworks into the future systems may increase reliability and trust in the AI-enabled energy management.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

AI- Driven Demand Response Optimization in Smart Grids Using Self-Adaptive Neuro-Evolutionary Algorithms

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • R. Archana Reddy,
  • D. Raja Babu,
  • P. Pramod Kumar,
  • Sai Anjani Manda

摘要

Demand Response alleviates demand curves in intelligent grids during challenges in the power systems. Smart grids with the use of Artificial Intelligence (AI) optimize load prediction, grid stability, fault detection and security. The future power systems are based on these technologies which are powered by artificial intelligence, big data and IoT. Nevertheless, in detecting smart grids AI has some challenges in achieving full potential. The present literature includes standalone AI approaches that are not combined to optimize their effectiveness and system reliability is also an issue. In this paper, the review is done on AI technologies, deep learning and machine learning of smart grids to comprehend their efficiency in energy generation, distribution, and consumption optimization. The review studied the use of AI in load forecasting, stability analysis, fault detection, and demand side control, considering such methods as stochastic optimization and robust optimization and power system expansion simulation. Maintenance prognostic and real time decision making are methods of improving system durability in energy systems. Deep learning methods are effective in optimization of operations. Findings indicate that AI models that are supported by smart grid infrastructure enhance effectiveness and reliability. AI frameworks of secure systems should be adopted by energy managers. The researchers have not validated AI frameworks in real-time, which is a limitation of the study. Quantitative analysis has found out that the number of citations per document is 32.74 and 14.32/year with an increasing impact of the field. As the scientometric mapping shows, hybrid AI-neuro-evolutionary algorithms are much better than conventional optimization frameworks, with more than 20% higher benchmarked stud-ies predictive accuracy and 15% less cost. Future studies are important to deploy adaptive neuro-evolutionary models in live smart-grid systems to evaluate scalability and interoperability and resiliency to dynamic demand conditions. The combination of cybersecurity, explainable AI, and ethical frameworks into the future systems may increase reliability and trust in the AI-enabled energy management.