多尺度系统能耗优化的双PWM变频调速系统动态协同预测控制

Compared with the system without active front end (AFE), the dual-PWM variable frequency system, in which rectifier and inverter accomplish the energy conversion control on grid side and the load side respectively, has a more complex control structure and feature of energy flow. The buffer of dynamic process mainly depends on energy storage device with large capacity in DC circuit, which is lack of effective synchronized operation mechanism leads to negative impact on system performance and efficiency. This project tends to launch a research on improving the control performance and running efficiency of dual-PWM variable frequency system based on real-time optimization of system-level loss and bilateral instantaneous power dynamic synchronization. In this paper, the model predict control algorithm is combined with the control schemes on both sides, and the dynamic process which is required in coordinated operation and the delay compensation of the system will be achieved via adopting multi-step prediction and objective function optimization. Meanwhile, the model of system-level loss is established, which contains grid, power electronic conversion, motor and its load, and the multidimensional real-time optimization of loss during dual-PWM variable system energy flowing is gained by using dynamic optimization method. Finally, the system level loss minimization control is introduced into the dual-side power balance theory, and the energy fluctuation in the central loop is suppressed through accurate control of active power in dual-PWM system, thus the bilateral system energy dynamic coordinated control will be implemented in the condition of acceptable system performance, and the aim of improving the energy efficiency of frequency variable system will be achieved.

双PWM变频系统中整流侧与逆变侧分别完成电网侧和负载侧的能量变换控制，相对于不带主动整流前端的系统具有更为复杂的控制结构和能量流动特征，其动态过程主要依靠中间回路的大容量储能元件进行缓冲，缺乏有效的同步运行机制，系统性能与运行效率往往难以兼顾。本项目拟站在系统级多尺度损耗实时优化与双侧瞬时功率动态协同的角度，对双PWM变频系统的控制性能与运行效率提升开展研究。将模型预测控制算法与双侧各自控制策略相结合，采用多步预测和目标函数优化方法实现系统延时补偿并获得协同运行所需的动态性能。同时，建立包含电网、电力电子变换、电机及其负载的多尺度损耗模型，采用实时在线寻优方法对双PWM变频系统能量流动过程中的损耗进行动态优化，进而将能耗优化控制与基于双侧功率平衡的预测控制相结合，通过对系统有功功率的精准预报抑制中间环节的能量波动，达到同时提升PWM变频调速系统综合能效指标和动态性能指标的目的。

高性能大功率双PWM交直交系统中整流侧与逆变侧分别完成电网侧和负载侧的能量变换控制，相对于不带主动整流前端的系统具有更为复杂的控制结构和能量流动特征，其动态过程主要依靠中间回路的大容量储能元件进行缓冲，缺乏有效的同步运行机制，系统性能与运行效率往往难以兼顾。项目拟站在系统级多尺度损耗实时优化与双侧瞬时功率动态协同的角度，对双PWM变频系统的控制性能与运行效率提升开展研究，包括：(1)将模型预测控制算法与双侧各自控制策略相结合，采用多步预测和目标函数优化方法实现系统延时补偿并获得协同运行所需的动态性能；(2)建立包含电网、电力电子变换、电机及其负载的多尺度损耗模型，采用实时在线寻优方法对双PWM 变频系统能量流动过程中的损耗进行动态优化；(3)将能耗优化控制与基于双侧功率平衡的预测控制相结合，通过对系统有功功率的精准预报抑制中间环节的能量波动；(4)双PWM变频系统实验平台的建立与运行，并应用于大功率交直交中压轧机主传动系统。通过研究，探索双PWM变频系统多尺度损耗实时优化与双侧瞬时功率动态协同控制方法，达到提升PWM变频调速系统综合能效指标和动态性能指标的目的。.项目按计划完成了以上全部内容，实验结果验证了所提出策略可以提升PWM 变频调速系统综合能效指标和动态性能指标。结合该项目的研究，已培养从事高性能交流变频控制系统与电力电子技术应用领域的青年教师2位，培养硕博士研究生前后总计达12人，其中已毕业6人，在读6人。共计发表学术论文10篇，国家授权专利11项。