基于ID预测模型的明渠系统控制参数在线优化研究

Open Canal System Control Parameters Online Optimization Based on ID Prediction Model

渠系控制算法在输水工程中的成功应用,大大提高了渠系的运行调度水平,但现有控制算法依然存在渠池间耦合和时间滞后、传统PID控制算法应对突变敏感性较高等问题。为进一步提高渠系的响应速度,减少水位波动,现将模型预测控制引入渠道自动化控制中,采用线性化的ID模型作为预测模型,通过在线滚动优化目标函数,以设计可实时调节反馈控制参数的PID控制器。以湖北省漳河灌区三干渠三分干为工程背景,建立单渠段控制模型并进行仿真。仿真结果表明,MPC控制方法具有实时在线滚动优化的特点,针对渠道中不断变化的需求情况,可以较好地预测未来系统状态并以此选择合适的控制器参数。

The successful application of canal system control algorithm in water conveyance project has greatly improved canal system operation, but the existing canal system control algorithms still have some drawbacks, such as the coupling between the canal and pool time lag, as well as traditional PID control algorithm dealing with high sensitivity mutation and so on. In order to further improve the response speed of the canal system and reduce the water level fluctuation, the model predictive control is introduced into the channel automation control. The linearized ID model is used as the forecasting model, and the objective function is optimized through on-line scrolling to design a PID Controller that can adjust feedback control parameters in real time. Based on the engineering background of three-trunk canals in Zhanghe Irrigation District of Hubei Province, a single canal control model is established and simulated. The simulation results show that MPC control method has the characteristics of a rolling real-time online optimization, demand for changing channels, and the future state of the system can better predicted and used to select the appropriate controller parameters.