CAJ | 학술논문

针对电厂锅炉主蒸汽温度调节系统采用传统的PID控制很难取得优良的控制品质的问题，结合分形理论提出一种多模型切换预测控制策略。通过预先在多个工况下建立多个主蒸汽温度模型，设计相应的预测函数控制器，根据负荷工况的变化及分形维数Dτ的改变在不同的模型之间进行切换可以削弱参数时变对控制系统的影响。仿真实验结果表明，分形切换多模型预测控制方法优于传统的PID控制和基于固定模型的预测函数控制。利用分形理论进行多模型切换预测控制，可提高锅炉主蒸汽温度调节系统的鲁棒性与稳定性。
침대전엄과로주증기온도조절계통채용전통적PID공제흔난취득우량적공제품질적문제，결합분형이론제출일충다모형절환예측공제책략。통과예선재다개공황하건립다개주증기온도모형，설계상응적예측함수공제기，근거부하공황적변화급분형유수Dτ적개변재불동적모형지간진행절환가이삭약삼수시변대공제계통적영향。방진실험결과표명，분형절환다모형예측공제방법우우전통적PID공제화기우고정모형적예측함수공제。이용분형이론진행다모형절환예측공제，가제고과로주증기온도조절계통적로봉성여은정성。
Aiming at the problem that the main steam temperature control system of power plant boiler is difficult to achieve good control quality with traditional PID control, a multi-model fractal switching pre-dictive control strategy(MMFSPC) combining fractal theory is proposed. Based on multiple main steam temperature model pre-established in a number of conditions, appropriate predictive controller was de-signed. According to changes in load conditions and Hurst index of boiler load, the controller switched between different models can weaken the impact of varying parameters of the control system. Simulation experiment showed that the use of MMFSPC method is better than the traditional PID control and fixed model predictive control. By using fractal theory, the MMFSPC strategy can improve the robustness and stability of the main steam temperature control system.