基于PID神经网络的MMC-BESS并网控制  

MMC-BESS grid connected control based on PID neural network

作　　者：梁帅 李晓英[1] LIANG Shuai;LI Xiaoying(Lanzhou University of Technology,Lanzhou 730050,China)

机构地区：[1]兰州理工大学电气工程与信息工程学院,兰州730050

出　　处：《自动化与仪器仪表》2023年第8期91-95,共5页Automation & Instrumentation

摘　　要：并网控制作为模块化多电平变流器-电池储能系统的应用基础,可采用功率外环,电流内环的控制方法。双闭环控制稳定性较高,但电流内环解耦较为依赖系统的电路参数。对此,将PID神经网络应用于电流解耦。同时,为避免电池工作电流过大,将功率限幅加入功率控制。然后,针对实际需求,加入无功调压和有功一次调频功能。最后,通过仿真验证了电流环采用PID神经网络进行解耦,超调量更小,调节时间更短;功率控制则可以稳定并网点电压,减少电网频率波动。As the application basis of modular multilevel converter-battery energy storage system,grid-connected control can adopt the control method of power outer loop and current inner loop.The double closed-loop control has high stability,but the current inner loop decoupling is more dependent on the circuit parameters of the system.Therefore,PID neural network is applied to current decoupling.At the same time,in order to avoid excessive working current of the battery,power limiting is added to power control.Then,according to the actual demand,the reactive voltage regulation and active primary frequency regulation functions are added.Finally,the simulation results show that the current loop is decoupled by PID neural network,the overshoot is smaller and the adjustment time is shorter;Power control can stabilize the voltage at the grid connection point and reduce the frequency fluctuation of the grid.

关 键 词：PID神经网络 解耦控制 功率限幅 无功调压 一次调频 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]