渠系前馈蓄量补偿控制时滞参数算法比较与改进

输水明渠系统控制算法主要由前馈及反馈模块组成,其中前馈模块较大程度上影响了系统控制性能。基于蓄量主动补偿的前馈控制算法受时滞参数影响较大,为比较现有多种算法实际控制性能优劣,并寻求不依赖复杂数值计算的简化时滞参数算法,该文进行了横向的算法比较与优化开发。文章基于蓄量阶跃补偿及蓄量二次补偿2类算法,结合改进比例-积分(proportional-integral,PI)反馈控制器,对典型测试渠系及实际工程渠系建模仿真,选取稳定时间、最大超调流量、绝对值误差积分(integral of absolute magnitude of error,IAE)和绝对流量变化积分(integrated absolute discharge change,IAQ)指标分析了蓄量阶跃补偿、动力波原理、水量平衡模型3种时滞参数算法的控制效果。结果显示,在实际工程渠系中,小流量工况下,蓄量阶跃补偿算法的稳定时间比水量平衡模型算法的稳定时间减小40.42%;大流量工况下,蓄量阶跃补偿算法的最大超调流量最小,仅比目标流量高3%,该算法的稳定时间比水量平衡模型算法的稳定时间减少25.45%。结果表明蓄量阶跃补偿算法控制效果较好,但该算法在推求渠池所需蓄量补偿值时需进行水面线推求,文章依据渠道蓄量变化与流量变化间的线性关系,提出简化的时滞参数显式算法。较传统数值算法,时滞参数显式算法与蓄量阶跃补偿算法的时滞参数差值百分比小于8%,在满足实际工程需求的同时,可明显减少推求所需蓄量补偿值的计算量。文章的比较结论及所提出的简化算法对输配水渠道系统,尤其是大型渠道系统调度具有一定的理论价值和应用前景。

Comparison and improvement of time delay parameter algorithm for feedforward volume compensation control in canal system

Control algorithm of open canal system is mainly composed of feedforward and feedback module. Feedforward control roughly adjusts canal flow according to the water supply plan, while feedback module tries to reduce real-time water level deviation from the target. However, feedforward has great impact on the system''s dynamic performance. Generally feedforward control of canal is using volume compensation method which has one most important parameter, time delay parameter. Actually, the complexity of wave actions in open channel makes it difficult to accurately estimate the time delay parameter, thus it reduces the performance of feedforward control, and sometimes lead to waste of precious water resources. The effect of different time delay parameter algorithms applied to the practical canal system has not been compared and evaluated in living literature. At the same time, volume compensation requires numerical computation of volume change which might be further simplified in terms of the relationship between time delay parameter and water demand change flow. The constant downstream water level operation is usually adopted by many water transfer projects such as the middle route and the east route of the South-to-North water transfer project in China. To compare the effectiveness of different algorithm, this paper use test canal system(ASCE) and practical engineering canal system for modeling and simulation. Based on the volume step compensation and volume secondary compensation algorithm, combined with the integral improved proportional- integral(PI)controller, stable time, maximum overshoot flow, integral of absolute magnitude of error(IAE) and integrated absolute discharge change(IAQ) are selected to quantitative analysis the control effect . The results suggest that the secondary disturbance problem of volume secondary compensation algorithm is inevitable, which adds extra disturbance and leads to longer stable time. When the storage volume of target canal is easy to obtain, the time delay parameter calculated by the volume step compensation algorithm makes the transition faster and more stable. In the practical engineering canal system with small flow change, the stable time of the volume step compensation algorithm is 40.42% less than that of the water balance model algorithm. For large flow change, the maximum overshoot flow of the volume step compensation algorithm is the smallest, only 3% more than the target flow; its stable time is 25.45% less than that of the water balance model algorithm. The results show that the effect of the volume step compensation control is much more fast, gentle and stable than the volume secondary compensation algorithm. Furthermore, it is easier to program. Therefore, this paper recommends the volume step compensation algorithm as the main feedforward method for canal system. But even for this method, numerical computation of volume change is still needed. In order to explore the relation between the volume change, water demand change flow and time delay parameter, we examine time delay parameter by adopting the theory of constant gradual flow in open channel, and propose a new equation of the time delay parameter based on the volume step compensation. The coefficients of time delay parameter explicit algorithm are influenced by channel geometry parameters and initial flow conditions. In real application, it is possible to determine the time delay parameter without estimating the volume change. However, compared with the rigorous mathematical solution, the results need to be corrected by empirical formula. Due to the test data of this paper, the percentage difference of time delay parameter between time delay parameter explicit algorithm and volume step compensation algorithm is less than 8%, the accuracy of the proposed formula is good enough for engineering purpose. The results show explicit algorithm is quite promising for the control of canal systems.