National Energy Group Wuhai Energy Co.,Ltd.,the Nei Monggol Autonomous Region
Guoneng Zhishen Control Technology Co.,Ltd.

重介质选煤过程干扰多、工况动态变化复杂,传统的PID控制对灰分的跟踪控制效果差,无法满足生产需求,迫切需要更先进、更智能的控制方法来改善灰分跟踪控制性能,更好地实现节能降耗。分析了重介质选煤过程的工艺流程及重介质选煤灰分控制特性,针对重介质选煤过程的强非线性和复杂的动态特性,利用低阶线性模型和高阶未建模动态项的组合模型描述重介质选煤过程灰分控制系统的模型。利用投影算法和自适应模糊系统(AdaptiveNetwork-basedFuzzyInferenceSystem,ANFIS)交替辨识线性模型的参数和未建模动态项,设计了一种未建模动态补偿驱动的重介质选煤自适应控制系统,抵消了未建模动态对闭环系统稳定性产生的影响。通过重介质选煤试验对比未建模动态补偿驱动的重介质选煤自适应控制与线性自适应控制的控制效果,结果表明:未建模动态的补偿的重介质选煤过程的灰分跟踪控制的响应更快,其平均输出误差绝对值为0.3165%,较线性自适应控制降低13.67%,具有更好的稳态性能,验证了所提的灰分控制系统模型估计算法和自适应控制方法的有效性。

The heavy medium coal preparation process is fraught with many uncertainties and disturbances, and its working conditions undergo complex dynamic changes. The traditional PID control falls short in effectively tracking and controlling ash content, which cannot meet production requirements. There is an urgent need for more advanced and intelligent control approaches to improve the performanceof ash tracking and controlling, thereby facilitating improved energy conservation and consumption reduction. The technological process ofheavy medium coal preparation process and the control property of coal ash content in heavy medium coal preparation process were analyzed. Aiming at the strong nonlinear and complex dynamic characteristics of heavy medium coal preparation process, a combination modelof low-order linear model and high-order unmodeled dynamic term was used to describe the ash control system model of heavy medium coal preparation process. By using the projection algorithm and the adaptive fuzzy system ( Adaptive Network-based Fuzzy Inference System, ANFIS) to alternately identify the parameters of the linear model and the unmodeled dynamic terms, an adaptive control system of heavy medium coal dressing driven by unmodeled dynamic compensation was designed to offset the influence of the unmodeled dynamic on the stability of the closed-loop system. Through heavy medium coal selection experiments, the control effects of unmodeled dynamic compensation driven heavy medium coal selection adaptive control and linear adaptive control were compared. The results show thatthe ash tracking control of the unmodeled dynamic compensation heavy medium coal selection process has a faster response, with an average output error absolute value of 0.316 5%, which is 13.67% lower than linear adaptive control, and has better steady-state performance.The study validates the effectiveness of estimation algorithm and adaptive control method of the proposed model for ash content control system.

国家自然科学基金资助项目(42030807)

李伟涛,苟晓东.未建模动态补偿驱动的重介质选煤自适应控制[J].洁净煤技术,2023,29(12):128-135.

LI Weitao,GOU Xiaodong.Adaptive control of heavy medium coal preparation without modeling dynamic compensation drive[J].Clean Coal Technology,2023,29(12):128-135.