基于约束卡尔曼滤波的区域多能源系统鲁棒状态估计

Robust State Estimation for Regional Multi-energy Systems Based on Constrained Kalman Filter

为了保障区域多能源系统(regional multi-energy system,RMES)的可靠运行,亟需提出面向RMES的全面、实时的状态感知方法,为此计及异质多能流动态特性,提出基于约束卡尔曼的RMES鲁棒状态估计(state estimation,SE)方法。该方法首先采用Lax-Wendroff差分法对配气网动态过程进行数值求解;随后构建配电网状态空间模型,并以燃气轮机和电转气为接口,建立电气双向耦合环节;最后,以卡尔曼滤波(Kalman filter,KF)为基础,添加量测噪声自适应算法,并计及配气网时序状态约束及电气边界耦合约束修正估计结果。采用改进的IEEE 33节点配电网与10节点配气网构建MATLAB仿真算例,仿真结果表明该方法能够有效跟踪RMES的状态变化,相较常规KF算法,配气网估计精度提升2个数量级,且在坏数据干扰下保持较好的鲁棒性。

In order to ensure the reliable operation of the regional multi energy system(RMES),it is urgent to propose a comprehensive and real-time state perception method for RMES.Therefore,this paper proposes a robust state estimation(SE)method for RMES based on constrained Kalman filter,which takes dynamic characteristics of heterogeneous multi energy flows into account.The method firstly uses the Lax-Wendroff difference method to numerically solve the dynamic process of the gas distribution network.Then,it constructs a state-space model of the power distribution network and establishes a two-way coupling link between electricity and gas,using gas turbine and power-to-gas as the interface.Finally,based on the Kalman filter,the method adds a measurement noise adaptive update algorithm and corrected SE results by considering the temporal state constraints of the gas network and the electricity-gas boundary coupling constraints.By using an improved IEEE 33 node distribution network and a 10 node gas distribution network,the paper constructs a MATLAB simulation example.The simulation example demonstrates that this method can effectively track the state changes of RMES.Compared to the traditional KF algorithm,it improves the accuracy of gas network SE by two orders of magnitude,maintaining good robustness under bad data interference.