AI-based model for Prediction of Power consumption in smart grid-smart way towards smart city using blockchain technology

• The article highlights the significance of smart grids in managing energy demand, addressing environmental issues, and ensuring energy security, emphasizing the need for blockchain technology and AI. • The study discusses data collection from smart city power consumption in a smart grid, utilizing techniques like Z-Score normalization and Spatial Temporal Correlation, and suggests blockchain technology for secure data transmission and storage. • The paper proposes a LSTM-RNN-ISSA for improved load forecasting accuracy and discusses the use of Blockchain-Based Smart Energy trading for effective communication in the smart grid. • Task-Oriented Communication facilitates real-time demand response, balancing electrical load and supply, outperforming existing approaches in smart city energy management. A smart grid (SG) is the financial benefit of a complicated and smart power system that can keep up with rising demand. It has to do with saving energy and being environmentally friendly. Growing populations and new technologies have caused a big rise in energy use, causing big problems for the environment and energy security. It is essential and significant to use blockchain technology and artificial intelligence (AI) to solve problems with power control. Data can be collected using a smart city in a power-consumed smart grid data and pre-process using a Z-Score normalization technique. It can extract features using a Spatial-Temporal Correlation (STC) to assess smart grid power usage within the context of a smart city using large-scale, high-dimensional data. Ensuring data integrity, privacy, and trust among grid applicants, transmit the data securely and reliably to a centralized or distributed cloud platform utilizing blockchain technology—a secure transmission and storage using Distributed Authentication and Authorization (DAA) protocol. To achieve precise load forecasting, a short-term recurrent neural network with an improved sparrow search algorithm (LSTM-RNN-ISSA) is incorporated. The smart grid may then record the projected results. Communication can be done on a smart grid with the users; the Blockchain-Based Smart Energy Trading with Adaptive Volt-VAR Optimization (BSET-AVVO) algorithm can be used for effective communication—a quick balancing electrical load and supply via a task-oriented communication mechanism in real-time demand response. Finally, our proposed method performs successfully better than the existing approaches.