基于时滞模型预测控制算法的网络化AGC研究

发布时间：2018-01-07 23:33

本文关键词：基于时滞模型预测控制算法的网络化AGC研究　出处：《电工技术学报》2014年04期 　论文类型：期刊论文

【摘要】：针对网络化自动发电控制(AGC)系统通信延迟的不确定性和数据丢包问题,提出一种基于时滞模型预测控制(MPC)算法的网络化AGC系统,该方法采用计及通信延时的互联电网时滞动态状态方程作为MPC的原始预测模型,以最优加权算法预测通信延时;并在MPC滚动优化目标函数中增加输出量预测值变化量,以减少发电机组调节次数。利用MPC多步超前预测控制功能,在执行机构信号接收端设置缓存器存储控制量预测序列,作为网络偶发数据包丢失时的备用次优控制信号。通过对两区域互联电网网络化AGC系统的仿真分析,仿真结果表明该方法能确保在随机通信延迟下网络化AGC具有优异的动态响应性能,该系统对通信延迟变化具有良好的适应性和较强的鲁棒性。
[Abstract]:Aiming at the uncertainty of communication delay and data packet loss in networked automatic generation control (AGCs) system, a networked AGC system based on delay model predictive control algorithm is proposed. In this method, the delay dynamic state equation of interconnected power network considering communication delay is used as the original prediction model of MPC, and the optimal weighted algorithm is used to predict the communication delay. In order to reduce the regulation times of generating units, the output prediction value change is increased in the MPC rolling optimization objective function, and the MPC multi-step predictive control function is used. A buffer stores a control quantity prediction sequence at the receiving end of the executive signal. As the backup sub-optimal control signal when the network accidental packet is lost, the simulation analysis of the networked AGC system of the two area interconnected power network is carried out. The simulation results show that this method can ensure the excellent dynamic response performance of networked AGC under random communication delay, and the system has good adaptability and robustness to the change of communication delay.
【作者单位】：重庆大学输配电装备及系统安全与新技术国家重点实验室;国网娄底供电分公司;安徽省电力科学研究院;
【基金】：国家自然科学基金资助项目(51277184) 输配电装备及系统安全与新技术国家重点实验室自主研究项目(2007DA 10512711209)
【分类号】：TM73
【正文快照】： 1引言传统AGC采用专用通信通道,运行设计时不需考虑通信延时对系统的影响[1]。基于综合信息网络的网络化AGC以其开放性、信息共享度高、控制灵活、成本低廉[1-4]等优点,将逐渐成为智能电网环境下二次调频控制方式,但是网络化AGC也带来了通信延迟、数据包错序及丢失等问题。通

【参考文献】