土石坝漫坝风险分析计算方法研究

发布时间：2018-08-03 16:28

【摘要】：水库大坝是人类社会发展中的巨大财富,在当今社会起着至关重要的作用。一方面,水库通过汛期蓄水减轻了非汛期水资源紧缺的现状,并通过水库调节降低了下游遭受洪灾的风险；另一方面,其存在又不可避免地带来了大坝失事的新风险,尤以土石坝最为突出。水库发生漫坝失事的主要诱因是洪水,因此对于洪水特性的研究显得极为重要。本文在风险分析概要介绍的基础上,对土石坝漫坝风险进行了分析,提出了一种基于改进蒙特卡罗法的计算漫坝风险值的新方法,同时对与漫坝风险计算密切相关的水文频率分析方法进行了研究,针对已有方法存在的不足,提出了一种新的水文频率分析计算方法。最终,以辽宁省大伙房水库为例对上述计算方法进行了实际应用。本论文主要研究内容如下：(1)基于风险分析相关理论及计算方法,提出了采用改进的蒙特卡罗方法计算漫坝风险值。该方法通过采用重要抽样与拉丁超立方抽样相结合的抽样法替代了直接蒙特卡罗法中的简单抽样。通过对其与直接蒙特卡罗法的抽样效果及收敛效果进行分析,可知改进的蒙特卡罗法在计算漫坝风险过程中有效提高了抽样效率,减少了计算工作量。(2)由于随机变量频率分析误差的存在会对随机变量设计值计算结果产生影响,本文提出在漫坝风险分析时应考虑频率分析的不确定性,并给出了考虑该不确定性时漫坝风险的计算方法。(3)在水文频率分析计算中,为避免参数法的线型限制问题及非参数法复杂的核函数选取问题,提出了一种新的水文频率分析方法——基于概率密度演化法的水文频率分析方法。在对概率密度演化法原理介绍的基础上,建立了水文频率分析模型,并给出了模型求解方法及水文频率设计值推求的具体流程。(4)采用蒙特卡罗模拟,对所建立的基于概率密度演化法的水文频率分析模型鲁棒性进行了研究,发现与常用的参数法相比该方法具有较好的鲁棒性。为进一步研究所提出方法的特性,以台湾石门水库及嫩江大赉水文站作为实例进行了数值分析。结果表明,同常用的参数法相比,基于概率密度演化法的洪峰流量频率曲线计算结果可以更好地与样本经验频率点据拟合,是一种可行且有效的水文频率曲线计算方法。此外,概率密度演化法还可以作为频率分析参数法估计结果的对比,用来分析检验参数法所采用的假设总体分布是否合理。(5)以辽宁省大伙房水库为例,对其漫坝风险进行了分析,分别给出了不考虑频率分析不确定性及考虑频率分析不确定性时,由洪水引起的漫坝风险值和由洪水与风浪联合作用引起的漫坝风险值。数值分析结果表明,风浪及频率分析不确定性均对大伙房水库漫坝风险值产生一定的影响,特别是风浪作用。根据计算结果,对大伙房水库的调度运行提出了一些建议,以期在保证大坝安全的前提下提高水库兴利效益。
[Abstract]:Reservoir dam is a great wealth in the development of human society and plays a vital role in today's society. On the one hand, the reservoir reduces the shortage of water resources in non flood season through the flood season, and reduces the risk of flood in the lower reaches through reservoir regulation. On the other hand, the existence of the reservoir inevitably brings about the new dam failure. Risk, especially the earth rock dam is the most prominent. The main cause of the flood failure in the reservoir is the flood, so the study of the flood characteristics is very important. On the basis of the summary of the risk analysis, this paper analyses the overflow risk of the earth dam, and puts forward a new formula for calculating the risk value of the overflow dam based on the modified Monte Carlo method. At the same time, the method of hydrological frequency analysis, which is closely related to the calculation of dam risk, is studied. A new method of hydrologic frequency analysis and calculation is put forward in view of the shortcomings existing in the existing methods. Finally, the practical application of the above calculation method is taken as an example of Dahuang Fang reservoir in Liaoning province. The main contents of this paper are as follows: (1) Based on the related theory and calculation method of risk analysis, an improved Monte Carlo method is proposed to calculate the risk value of the overflow dam. By using the sampling method combining important sampling with Latin hypercube sampling to replace the simple sampling in the direct Monte Carlo method, the sampling effect and convergence of the direct Monte Carlo method are obtained. The effect is analyzed. It is known that the improved Monte Carlo method can effectively improve the sampling efficiency and reduce the calculation workload in the process of calculating the overflow risk. (2) because the existence of the frequency analysis error of the random variable will affect the result of the design value of the random variable, this paper proposes that the frequency analysis should be taken into account in the analysis of the overflow risk. A new method of hydrologic frequency analysis, based on the probability density evolution method, is proposed. (3) in hydrologic frequency analysis and calculation, a new method of hydrologic frequency analysis, based on the probability density evolution method, is proposed in order to avoid the problem of line type restriction and the complex kernel function selection of non parametric method in the calculation of the hydrologic frequency. On the basis of the introduction of the principle of probability density evolution method, a hydrological frequency analysis model is established, and the concrete flow of the model solving method and the calculation of hydrologic frequency design value is given. (4) the Monte Carlo simulation is used to study the robustness of the established hydrological frequency analysis model based on the probability density evolution method. The parameter method has better robustness than the method. In order to further study the characteristics of the proposed method, the numerical analysis is carried out with the Shimen reservoir in Taiwan and the Dalai hydrological station of Nenjiang as an example. The results show that the calculation results of the flood peak flow frequency curve based on the probability density evolution method can be better compared with the common parameter method. This empirical frequency point is a feasible and effective method for calculating the hydrological frequency curve. In addition, the probability density evolution method can also be used as a comparison of the estimation results of the frequency analysis parameter method. It is used to analyze the rationality of the overall distribution of the hypothesis used in the test parameter method. (5) take the Liaoning Dahuang reservoir as an example, the risk of its overflow. The analysis shows that the risk value of the flood caused by the flood and the risk value of the overflow caused by the combined effect of flood and wind wave, which do not consider the uncertainty of frequency analysis and the uncertainty of frequency analysis, show that the uncertainty of wind wave and frequency analysis all have a certain value on the flood risk value of the big Huang reservoir. According to the results of the calculation, some suggestions are put forward for the dispatching and operation of the Dahuofang reservoir, in order to improve the benefit of the reservoir on the premise of guaranteeing the safety of the dam.
【学位授予单位】：大连理工大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TV641;TV122

【相似文献】