基于改进TS模糊的直流电弧炉闪变抑制优化

为提高静止无功补偿器（Static Var Compensator, SVC）应对直流电弧炉等冲击性负载的闪变抑制性能，本文在改进Takagi-Sugeno（TS）模糊算法的基础上，提出了一种SVC滚动预测控制方法。首先，建立了直流电弧炉电气模型并仿真分析其无功特性；然后，针对经典TS模糊预测算法应用于波动负荷时，出现的输出异常置0情况，提出一种范围自适应修正的改进方法，该方法能消除一类算法应用机理导致的异常值，从而提高TS模糊算法对波动负荷无功功率预测的可靠性和准确性；最后，基于模型训练时间约束，建立无功功率半周期滚动预测控制模型，提前10ms预测无功功率，改善了SVC传统控制系统响应的滞后特性。仿真结果表明，所提方法相比于SVC传统控制方法，平均闪变改善率提高了54.17%，对闪变现象的抑制效果提升显著。

Flicker suppression optimization of DC arc furnace based on improved TS fuzzy algorithm

In order to improve the flicker suppression performance of static var compensator (SVC) in response to impact loads such as DC electric arc furnace, this paper proposes a SVC rolling predictive control method based on the improved Takagi-Sugeno (TS) fuzzy algorithm. Firstly, the electrical model of DC electric arc furnace is established and its reactive power characteristics are simulated and analyzed. Then, an improved method of range adaptive correction is proposed to address the issue of abnormal output set to 0 when the classical TS fuzzy prediction algorithm is applied to fluctuating loads. This method can eliminate the outliers caused by the application mechanism of a class of algorithms, so as to improve the reliability and accuracy of TS fuzzy algorithm for reactive power prediction of fluctuating loads. Finally, based on the model training time constraint, a reactive power half-cycle rolling predictive control model is established to predict the reactive power 10 ms in advance. It is competent to improve the lag characteristics of the traditional SVC control system response. The simulation results show that the average flicker improvement rate of the proposed method is increased by 54.17 % compared with the traditional SVC control method, and the suppression effect of flicker phenomenon is significantly improved.