基于模仿者动态算法的虚拟发电厂内部分布式发电资源均衡调度方法

Replicator Dynamics Algorithm Based Equilibrium Dispatching Strategy for Distributed Energy Resources in Virtual Power Plant

为优化整个虚拟发电厂（virtual power plant，VPP）内部分布式发电资源（distributed energy resources，DER）的调度成本。基于VPP的多代理结构形式建立了VPP内部DER的均衡调度模型，以实时满足系统负荷需求为条件，利用模仿者动态算法处理VPP内部调度资源分配问题。通过分析不同种类DER的适应度系数与可调容量，提出多方共赢的VPP内部DER的均衡调度策略：适应度系数越小（即调用成本越低1、可调容量越大的DER参与调度的机会越多，反之越少。算例分析证明了所提模型和算法的合理性和有效性。与传统基于市场机制的VPP调度方法相比，模仿者动态算法在解决VPP内部DER调度问题方面能有效控制调度成本，同时具有简便、易拓展的特点。

To optimize the dispatching cost for distributed energy resources （DER） within whole virtual power plant （VPP） , based on the multi agent structure of VPP an equilibrium dispatching model for DER inside VPP is established. Under the condition of meeting load demand of power system in real time the replicator dynamics algorithm （RDA） is utilized to allocate the DER inside VPP. By means of analyzing the fitness coefficients and dispatchable capacities of different types of DERs, a win-win equilibrium dispatching strategy among multi DERs within VPP is proposed; the smaller the fitness coefficient, namely the lower the invoking cost, the higher opportunity of DER to participate the dispatching is, and if the former is higher, the lower the latter will be. The reasonability and effectiveness of the proposed model and algorithm are validated by simulation results of a VPP containing six generating units. In the solution of DER problem within VPP, the dispatching cost can be controlled more effectively by RDA than traditional market mechanism based VPP methods, besides, RDA possesses the advantages such as convenient to use and easy to expand.