A Hybrid Marine Predators Algorithm with Particle Swarm Optimization Using Renewable Energy Sources for Energy Scheduling Problem-Based IoT

The energy scheduling problem (ESP) is a complex NP-hard optimization scheduling problem of finding the best schedule for the energy consumed through a time horizon by smart appliances according to several constraints and dynamic pricing scheme(s). The primary goal of addressing ESP is to optimize electricity bills for users, the highest energy consumption through a time horizon, and the comfort level of users. Several communication and controlling techniques were utilized to interconnect and centralize the smart appliances to be scheduled, where the Internet of Things (IoT) technique is the most helpful. Several studies have proposed different optimization methods to address the ESP optimally. Unfortunately, due to the low performance of the utilized methods and the high number of the problem’s constraints, the best schedules were not achieved in many cases. In this paper, a new improved optimization method on the basis of a hybridization approach between two well-established optimization methods, called the marine predators algorithm and particle swarm optimization, is proposed to handle ESP optimally. The proposed method is mainly established and utilized to enhance the schedules with the worse fitness values to improve them to be more acceptable. Furthermore, to avoid the high complexity of the ESP constraints, Renewable Energy Sources (RESs) based on real-world dataset is utilized along the proposed method. Also, the IoT technique is utilized to establish the connection and build the scheduling system. Finally, to optimize the ESP objectives simultaneously, ESP is formulated as a multi-objective problem. In the evaluation stage, the proposed hybrid method with RESs is tested among four different comparison studies and phases, including the adapted methods with and without RESs, hybrid methods with and without RESs, and based state-of-the-arts. The proposed hybrid method with RESs proves its efficiency and high performance in all comparison phases, where it achieved the best results among all compared methods.