【作者】 韩延成；

【导师】 高学平；

【作者基本信息】 天津大学 ， 水力学及河流动力学， 2007， 博士

【摘要】 在长距离调水工程中,渠系输水控制运行是最重要任务之一,特别是对复杂的大型长距离调水工程,其特点是距离长,控制站点多,分布不均,要求实行不间断供水,调度和控制十分复杂,任何调度运行的失误都可能造成严重的后果。研究新的、高效的输水控制新技术及非恒定流仿真模拟系统是非常必要的。(1)对带有梯级泵站的大型长距离调水工程,电费是最大的运行成本之一,为了降低输水成本,本文利用电网峰荷期和基荷期不同的电价及最优化原理,提出了一种针对长距离复杂调水工程的最优输水控制模型。从仿真模拟结果看,优化控制方法不仅能有效的降低运行电费,也能有效控制水位的变化,而且可以以最少泵站开停机次数,有效控制整个渠道运行。(2)针对大型自流型调水工程,根据控制蓄水量法的基本原理、最优化理论和多年调度运行的实践经验,提出了一种使渠道内水流在最短时间内恢复到目标水位的最优输水控制的二步法模型。较传统输水方法:一是可以在规定的时间内,以最快的速度使渠道水位恢复到目标水位,传统方法需要几天才时间完成调控的渠道,用本优化控制方法,可在几小时内使渠道水位达到目标水位;二是模型中对渠段的蓄水量变化速率、流量变化等进行了约束,因此能使渠道内水流平稳运行,也可避免由于水位降落速度过快产生的渠堤滑坡。(3)根据明渠输水特点,提出了将传统渠道下游常水位输水PID控制和RBF人工神经网络控制方法结合的非线性输水控制模型,使输水控制具有自学习、自适应功能,容错性、鲁棒性强的特征。通过非恒定流仿真表明,基于RBF网络的PID输水控制方法,能够通过不断学习,自动调整控制参数,使输水控制过程超调量小,响应速度快,具有不需要特意选择或计算控制参数的优点。(4)根据模糊控制的优点,结合PID控制结构简单、稳定的特点,建立了渠道输水控制参数自适应模糊控制输水模型。模型通过模糊推理,吸纳渠道输水自动控制的专家经验,实现参数的优化,使控制过程始终处于一个较优的状态。水力模拟结果表明,用模糊自适应控制渠道输水,能明显减少超调量,输水控制的动态特性也得以改善。(5)根据渠道输水具有非线性的特点,提出了渠道输水控制的非线性PID数学模型;分析了渠道输水PID控制响应曲线,各个阶段控制参数对控制过程的影响,根据参数相对较优的变化趋势确定了比例、积分、微分参数的非线性函数曲线。进行了非线性PID控制的非恒定流模拟,仿真结果表明,应用实时调整增益参数的非线性模型较常规线性PID控制结构简单、响应速度快,超调量小。(6)提出了步进式PID渠道输水控制模型。在步进式PID渠道输水控制过程中,目标水位不是一次设定的静态水位,而是动态的、多步的,使输入指令一步一步的逼近所要求的最终目标水位。步进式控制可使渠道输水更稳定,使水位下降速度可控,具有简单实用、可控性强的特点。仿真结果表明,较常规PID控制,步进式PID控制渠道输水,水位波动小,超调量小,较适合需要严格控制水位变化的明渠输水或冰期输水。(7)研究并开发了专门针对长距离调水工程调度运行及自动控制过程的非恒定流仿真模拟系统,系统可用特征线法也可用隐格式法计算。可用于实际输水调度、输水控制模型研究、河网调度等。系统针对渠道输水控制,编写了常用的渠道输水自动控制模型及RBF神经网络法、模糊控制法、二步法、优化法及步进法的控制程序。用户也可自定义控制过程。更多还原

【Abstract】 The operation and control is one of the most important work for long-distance water conveyance project, especially for complex large-scale canal system, distance is long, control points is more, distributing asymmetrical, and water supply need uninterrupted. So the control is not easy. Any mistake can lead to serious results such as liner-plates damaged or dam banks slipped. The research about new, high quality and needs-satisfying control models and unsteady flow simulation system for long-distance water conveyance canals is important and necessary.To large-scale water conveyance systems with stair pumps, the power cost is high. For reducing the cost of transmitting water, a new optimal methodology used control flow in canal was presented based on optimal methods and different power-network prices in peak-load and low-load periods for a long distance water conveyance project. The result of an application example shows that the cost can be reduced largely, the minimum frequent of switch can be achieved, flow in canal can be stabilized and the safety of the canal can be ensured.A two-step model, which can make the deviant water levels restore to the normal position in minimum duration was present for long and large-scale water conveyance canal based on the volume control method, optimum technique and experiences in operation and control works. The merits using this model are (1) Water levels arrive to the goal level during minimum time. The simulations show that the time that the water levels reach the target levels and keep them stable can be reduced from a few days using classic method to a few hours using this optimum method. (2) The drop and rise of flow rate is constrained, so the flow is stable which can avoid landslip because of water level rapid fall.A method combining normal PID and RBF neural network method was presented according to the characteristic of canal transmitting water. It has merits of self-studying, self-adapting, strong fault tolerant and robustness. Through hydraulic simulating, the results show that in the process of RBF PID control, the method can adjust the parameters at a best state,the water level overshoot is smaller and the response is quicker. It not require to choose and computer parameters complexly, and can be used to the real-time control of nonlinear system such as canal transmitting water.The paper presents a parameter self-adapting canal flow fuzzy PID control combining fuzzy control merits with PID control features of simple structure and stability. Through fuzzy inference method, the model uses the canal automation history control experiences to realize the optimization of parameters and make the control process at the best state. The simulation results of an adaptive fuzzy sample show that using the adaptive fuzzy PID control canal flow, the overshoot is reduced, the dynamic behaviors are improved. This paper presented a new step-size PID control method for canal flow according to the actual canal-flow operation for canal of cool area. In the step-size PID canal control process, it is not once time to set goal water level statically, but set many temp goal that approach final goal step by step. This can control the water drop speed, and make the flow in canal steady. The Step-PID has the characteristics of simple, better controllability and operability. The simulations show that the water fluctuation and overshoot are smaller using Step-PID. The method is suitable to open channel flow with drawdown speed is hard limited or canal flow under ice.A nonlinear open channel PID control method was proposed according to the nonlinear and hysteretic nature of canal flow control. The nonlinear functions of proportional and integral parameters were created based on the relatively best tendency. The unsteady flow of canal nonlinear PID control process is simulated last. The simulation results show that the response of nonlinear PID control speeds is faster, the overshoot is less rather than that of normal liner PID control method.The unsteady flow simulation systems of canal operation and automation control have been researched and developed for long distance and large-scale water conveyance canals. Users can use either characteristic line method or implicit format to simulate. It can be used to simulate unsteady flow processes of operation and control, to simulate, verify and develop auto control models, to simulate the hydraulics and flood of river network. The systems include some classic canal control models and RBF+PID, fuzzy+PID, Two-Step optimum control for gravity canal, cost optimum control for water conveyance system with pumps.更多还原