A Genetic Algorithm Driven Linear Programming for Battery Optimal Scheduling in nearly Zero Energy Buildings

Co‐optimization of active power curtailment, load shedding and spinning reserve deficits through hybrid approach: Comparison of electrochemical storage technologies

“…Robust unit commitment formulations account for the worst‐case scenario and involve different uncertainty factors relating to load forecasting, renewable power output and unintentional generation outages. In this work, we utilize the advances of priority‐list schemes to provide the optimal UC schedule and genetic algorithm (GA) as a tool to drive the optimization and optimally define the ES parameters of charging and discharging power output [32]. In contrast to other mathematical approaches, priority‐list method does not suffer from the identical heat‐rate sensitivity.…”

Enhanced Lagrange relaxation for the optimal unit commitment of identical generating units

“…This way, the computation time of exact methods becomes impractical calling for some heuristic approaches for which optimality is not given such a high priority, but the emphasis is on producing near optimal solutions in a lower computational burden [12]. The dominant methods extensively presented in the literature are genetic algorithms [13], simulated annealing [14], particle swarm optimisation [15], fuzzy LOGIC [16], expert systems [17], ant colony [18], Tabu search [17], evolutionary programming [19] and artificial neural networks [17]. Owing to their main disadvantages of local optimum trapping, untraceable transition rules and unreliable findings, hybrid metaheuristic approaches have been proposed in order to employ more rigorous mechanisms.…”

A Hybrid Optimization Approach for Autonomy Enhancement of Nearly-Zero-Energy Buildings Based on Battery Performance and Artificial Neural Networks

“…This was ensured with the aid of Equation ( 25) and, in this regard, the feasible exploration space was further decreased, mitigating the consequent computational burden. The algorithm shown above constitutes an amelioration of a previously applied algorithm presented in Reference [26].…”

“…The main aim of the study was to minimize the operational costs for electricity, space heating, and hot water, and simultaneously maintain the investment costs as low as possible, based on the initial investment and maintenance costs of the system. Georgiou et al [26] made a first attempt to use a hybrid optimization approach using LP and genetic algorithm (GA) for optimal dispatching of a battery in a building with PV installed. The authors demonstrated the benefits of using such approaches for maintaining the building's net grid energy at low levels.…”

“…The purpose of the current study was the nearly zeroing of the net grid electrical energy in buildings, in real time, in an optimal manner. This is a problem that, to the best of our knowledge, has only been dealt with on two occasions (References [26,30]), but not in the holistic manner of the current paper, which utilized and integrated convex optimization, heuristic optimization, load and PV forecasting, and realistic battery dispatch software to nearly zero the building's daily net-grid energy. This target was achieved by using LP to optimally dispatch the battery operation in an adaptive manner, assisted by equally powerful methods such as artificial neural networks (ANN) used to forecast both the next 24 hours' PV generation and load demand, and GA to drive LP to optimal solutions based on daily forecasts relating to PV generation and demand.…”