Article View

A Hybrid Quantum Differential Evolution Algorithm for Cost-Optimal Scheduling of Appliances and Energy Storage in PV-Equipped Smart Homes

DOI: 10.66838/ma.228

Gursleen Kaur Department of Computer Science, Punjabi University, Patiala, Punjab 147002, India
Rajesh Kumar Bawa Department of Computer Science, Punjabi University, Patiala, Punjab 147002, India

Abstract

Minimising the daily electricity cost of a household equipped with rooftop photovoltaic (PV) panels and a stationary energy storage system (ESS) requires appliance scheduling and storage dispatch to be optimised jointly under a day-ahead tariff. The resulting problem is of a mixed discrete–continuous nature: it combines integer appliance start times with continuous charge and discharge commands, and it is coupled across successive hours through the stored energy. This paper proposes a hybrid quantum differential evolution (HQDE) algorithm for this cost-minimisation problem. HQDE directs its differential mutation towards an elite subset of the population using control parameters that are scheduled over the run, and it additionally explores the appliance variables in a quantum-inspired angle space while retaining the storage commands in their physical units. Each candidate solution is decoded into an appliance schedule that always satisfies the required operating hours, and a fixed allocation order coordinates PV generation, ESS operation and grid exchange in every hour. Over 50 independent runs on a common simulator, HQDE reduced the mean daily cost of improved differential evolution (IDE), the strongest of four baseline algorithms, by 9.75% for a benchmark household, and even its worst run outperformed the best run of every baseline. HQDE also achieved a lower mean cost than IDE in all six storage configurations examined, and these configurations further revealed that the charge/discharge rate of the ESS influences the minimum cost more strongly than its capacity. A sensitivity analysis over 18 combinations of PV size, selling-price coefficient and storage rate showed that PV size has the greatest impact on cost and that the selling coefficient becomes most influential when PV generation is abundant.

Keywords

Appliance scheduling, cost minimisation, differential evolution, energy storage system, home energy management system, photovoltaic generation, quantum-inspired optimisation.,