Energy optimization design and implementation of building engineering based on intelligent decision-making algorithm

Abstract This paper presents an intelligent decision-making algorithm that integrates Q-learning, CNN, and LSTM to optimize energy consumption in building design and operations. The proposed method is evaluated against GA, PSO, SVR, and LSTM, demonstrating superior performance in energy reduction, response time, and computational efficiency. Experimental results show that the method reduces energy consumption by up to 25% in residential buildings and responds efficiently even under high computational loads. The system is further enhanced through a hybrid edge-cloud architecture to support real-time decision-making. While the findings are promising, limitations include reliance on simulated environments and the need for broader real-world validation. The proposed approach offers valuable insights for energy-efficient building design and smart energy management.