基于蒙特卡罗和最大熵方法的水泵测试不确定度研究

发布时间：2018-05-25 23:47

本文选题：水泵测试 + 不确定度　；参考：《沈阳工业大学》2011年硕士论文

【摘要】：自从《测量不确定度评定指南》颁布以来,不确定度理论在众多领域得到广泛的应用。近些年来,不少人致力于寻求最优评定方法的研究,取得了不少成果,但是在水泵测试不确定度评定的领域,研究者却很少,加之我国的水泵行业整体水平相对落后,技术人才较为缺乏,寻求一种有效的评定手段势在必行。本文就水泵测试不确定度评定普遍存在的问题给出一种不确定度评定新方法,基于蒙特卡罗方法和最大熵方法开展的对单级离心泵的性能测试参数的不确定度评定方法有效结合了两种方法的优点,展开进行不确定度评定。课题研究主要做了以下几个方面的工作：不确定度评定的方法的可行性分析以及各个方法的优劣性及适用性比较；对水泵性能测试系统进行分析,寻找不确定度来源及其消除措施,并结合实际的单级清水离心泵的性能测试进行实时数据采集,并对利用贝塞尔公式进行不确定度评定与分析；对收集的数据利用最大熵方法进行分布模拟,模拟得出这些数据分布类型,根据求得的分布类型,用舍选法产生相应分布的一定数量的随机数,然后利用蒙特卡罗方法进行不确定度的评定,并将结果与传统方式评定结果作比较,得出结论；针对最大熵问题求解过程普遍存在的初值选择困难的问题,进行算法研究,给出一种基于粒子群算法进行求解的方案,不再需要人工给定初始值,而只需给定一个范围即可,程序便可以自行在空间中搜索最优解,评定程序得到简化,避免一些随机因素尤其是人为选择初始值不当造成的不收敛问题,并将测试数据代入计算,与常规算法求解结果进行对比,验证算法的有效性。
[Abstract]:Since the publication of the Guide to the Evaluation of uncertainty in Measurement, uncertainty Theory has been widely used in many fields. In recent years, many people have devoted themselves to the research of the optimal evaluation method, and achieved a lot of results, but in the field of uncertainty assessment of pump testing, the researchers are few, and the overall level of the pump industry in China is relatively backward. There is a shortage of technical talents, so it is imperative to seek an effective means of evaluation. In this paper, a new method for evaluating uncertainty of water pump testing is presented. Based on the Monte Carlo method and the maximum entropy method, the uncertainty evaluation method for the performance test parameters of single-stage centrifugal pump is developed, which effectively combines the advantages of the two methods, and carries out the uncertainty evaluation. The main work of the research is as follows: The feasibility analysis of the methods of uncertainty evaluation and the comparison of the advantages and disadvantages and applicability of each method; Based on the analysis of the pump performance test system, the source of uncertainty and its elimination measures are found, and the real-time data collection is carried out in combination with the performance test of the actual single-stage clean water centrifugal pump, and the uncertainty evaluation and analysis are carried out by using the Bessel formula. The data collected are simulated by the maximum entropy method, and the distribution types of these data are obtained. According to the distribution types obtained, a certain number of random numbers of the corresponding distributions are generated by the selection method. Then Monte Carlo method is used to evaluate the uncertainty, and the result is compared with the traditional evaluation method, and the conclusion is drawn. In order to solve the maximum entropy problem, which is difficult to select the initial value, the algorithm is studied, and a solution based on particle swarm optimization is proposed. It is no longer necessary to give the initial value manually, but only to a range. The program can search the optimal solution in space by itself, simplify the evaluation procedure, avoid the problem of non-convergence caused by some random factors, especially the improper selection of the initial value, and substitute the test data with the calculation. Compared with the results of conventional algorithm, the validity of the algorithm is verified.
【学位授予单位】：沈阳工业大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：TH38

【相似文献】