Digital twin-enabled hybrid deep evolutionary framework for smart building sustainable infrastructure management

Yunbo Xu,Jiachao Zhang,Heyu Qin,Hao Zhou,Zikai Yang

DOI: https://doi.org/10.1016/j.seta.2024.103773

IF: 7.632

2024-04-24

Sustainable Energy Technologies and Assessments

Abstract:Combining services has become a major challenge in the use of service-based technology in smart building. The composition of services has numerous problems, specifically the quality of service (QoS), becoming increasingly difficult as dynamic environments change. The other concern is selecting services with a higher level of reliability based on trust value. A number of metaheuristic algorithms are employed to reduce the NP-hard complexity of service composition. Thus, the study adopts a hybrid gray wolf with cuckoo search optimization (HGWCSO) algorithm to achieve its objectives. The trust challenge is additionally addressed by using a trust-driven clustering algorithm using deep convolutional neural network (CNN) and k-means. The synergy between k-means and CNNs can result in improved clustering accuracy and better representation of underlying patterns in the data. According to the outcomes, the suggested algorithm is more effective in small-scale problems than the gray wolf optimization and particle swarm optimization algorithms. As the search space is clustered and reduced, response times are enhanced; it is also easier to select trusted services. The outcomes on the large-scale digital twin of smart building indicate that the suggested approach outperforms the average outcomes of similar studies in terms of computational time.

energy & fuels,green & sustainable science & technology