摘要
实时电价是需求侧管理策略的重要手段,是解决智能电网供需平衡的理想需求响应机制,能起到削峰填谷的作用。为此,为智能电网设计了一种分布式实时电价算法:基于改进对偶分解的近端中心算法,求解用户总效用与电能供应商成本之差最大的优化问题。在此算法中,对偶问题的拉格朗日乘子即为实时电价,通过迭代更新拉格朗日乘子,形成电能供应商的实时电价与用户的实时能耗水平之间的互动,算法最终为每个用户找到最优的能耗水平(即用户的总效用最大化),同时使得电能供应商的成本最小化。所提算法既保留了问题的可分离性,又加快了收敛速度,克服了基于对偶分解的次梯度法求解该优化问题在用户规模较大时收敛慢甚至不收敛的缺点。仿真结果充分表明了所提算法具有快速收敛的特性。
Real-time pricing scheme is an important means of demand side management strategies and is an ideal demand response mechanism to achieve the supply and demand balance in smart grid. It can bring out peak clipping and valley filling. Thus, we propose a distributed real-time ＇pricing algorithm for the smart grid, that is, proximal center algorithm based on an improved dual decomposition, which can solve the optimization problem that maximizes the difference between the utilities of all users＇ and the cost of the energy provider. In the algorithm, Lagrangian multiplier of the dual problem is the energy price. Through updating the real-time pricing the interactions are activated between real-time pricing of the energy provider and real-time energy consumption of users. The algorithm finds the optimal energy consumption levels for each user （the maximization of the users＇ utilities）, and meanwhile the minimum of the cost for the energy provider. It not only preserves the separability of the problem but also accelerates the convergence rate. When the network system is large, this algorithm overcomes the weakness of the subgradient algorithm based on dual decomposition, which is convergent slowly or even not. The simulation results testify the quality of fast convergence for the proposed algorithm.
出处
《电力系统保护与控制》
EI
CSCD
北大核心
2012年第21期42-47,共6页
Power System Protection and Control
基金
国家自然科学基金资助项目(60702081
61002016)
教育部科学技术研究重点项目(212066)
浙江省自然科学基金资助项目(LY12F02042)
浙江省留学人员科技活动择优资助项目~~
关键词
智能电网
需求侧管理
削峰填谷
实时定价
次梯度算法
近端中心算法
smart grid
demand side management
peak clipping and valley filling
real-time pricing
subgradient algorithm
proximal center algorithm
