Design of an autonomous intelligent Demand-Side Management system using stochastic optimisation evolutionary algorithms

作者：Galvan Lopez Edgar^*; Curran Tom; McDermott James; Carroll Paula

来源：Neurocomputing, 2015, 170: 270-285.

DOI：10.1016/j.neucom.2015.03.093

• Demand-Side Management systems
• Evolutionary algorithms
• Electric vehicles
• Peak-to-average ratio
• Electricity costs
• Smart grid time-of-use pricing

Demand-Side Management systems aim to modulate energy consumption at the customer side of the meter using price incentives. Current incentive schemes allow consumers to reduce their costs, and from the point of view of the supplier play a role in load balancing, but do not lead to optimal demand patterns. In the context of charging fleets of electric vehicles, we propose a centralised method for setting overnight charging schedules. This method uses evolutionary algorithms to automatically search for optimal plans, representing both the charging schedule and the energy drawn from the grid at each time-step. In successive experiments, we optimise for increased state of charge, reduced peak demand, and reduced consumer costs. In simulations, the centralised method achieves improvements in performance relative to simple models of non-centralised consumer behaviour.

• 出版日期2015-12-25