【作者】 彭霜；

【导师】 聂松林；

【作者基本信息】 华中科技大学 ， 机械电子工程， 2009， 硕士

【摘要】 控制液压系统污染的关键问题之一是合理地设计和维护过滤系统,将各关键点的污染物浓度控制在一定范围内。不必要的过滤器的安装及过分地维护势必会增加系统成本。本研究引入系统优化的思想,寻求系统性能和成本之间的折衷,通过Lingo求解出最优决策。由于在优化建模中存在许多随机、模糊、区间等不确定性的参数,采用多种优化方法结合的不确定性优化理论和方法对过滤器进行优化配置研究具有十分重要的意义。本文首先总结了国内外液压系统污染控制和两阶段随机、机会约束以及模糊优化方法方面的研究成果,阐述了采用优化思想和多种优化方法对过滤器进行配置的必要性。其次,在典型单回路液压系统的基础上,发展出带旁路的过滤系统。在考虑各种不确定性参数的情况下,建立了基于多种优化参数相结合的区间两阶段随机非线性规划(Interval Two-Stage Stochastic Nonlinear Programming缩写为ITSNP)、区间机会约束非线性规划(Interval Chance-Constrained Nonlinear Programming缩写为ICNP)以及区间模糊机会约束非线性规划(Interval Fuzzy chance-Constrained Nonlinear Programming缩写为IFCNP)等三种非线性规划方法的液压系统过滤器配置决策模型。本文将ITSNP、ICNP和IFCNP方法分别求得的最优解从理论上进行了对比和分析研究。研究结果表明:(1)改进的模型真实的模拟了污染物侵入产生规律的动态特性;(2)在污染物侵入产生率重度水平下,过滤器的纳垢容量和液压元件的污染敏感度的膨胀(即在一定的概率水平下,过滤器的纳垢容量和元件能够忍受的污染物浓度比标定的浓度偏大。)对最终的优化结果有着显著的影响;(3)IFCNP方法成功的得到了系统最优成本,但区间模糊因子为1,即各过滤器的纳垢容量约束为紧约束,故IFCNP方法的经济性是以牺牲设计风险为代价的;该方法在实际生产中可作为一种参考手段。更多还原

【Abstract】 One of the key problems of hydraulic contamination control is to design and maintain filtration system reasonably reducing the contamination level of key positions as low as possible. However, unnecessary allocations for filters and excessive maintenance for filtration system will definitely increase the system cost at the same time. This research focuses on the compromise between the quality of system performance and cost by introducing optimization theories, and finally obtaining the optimal results with the aid of Lingo. In optimization model, many parameters can be expressed as stochastic, fuzzy and interval numbers which are all uncertain forms. It is meaningful that planning research would be conducted on the allocation of system filters by combinating several advanced uncertain optimization methods.Firstly, the latest researches on contamination control for hydraulic system and some optimization methods like two-stage stochastic, chance-constrained and fuzzy optimization are concluded and the necessity of adopting optimum ideas and multi-optimization theories to allocate filters. Secondly, on the basis of typical single circuit hydraulic system, the research developed a filtetration system with bypass filters. With consideration of some uncertain parameters, three optimum decision-making models, including interval two-stage stochastic nonlinear programming (ITSNP)、interval chance-constrained nonlinear programming (ICNP)、interval fuzzy chance-constrained nonlinear programming (IFCNP), for allocation of the filters, were established based on combinations of multi-optimization parameters.Those optimal results of the three models were compared with each other. Analysis and study were made at the same time. The results demenstrat that: (1) the developed model truly mimic the dynamic characteristic of the generation/ingression principles for contaminants; (2) at high contaminant generation/ingression rates, the expansion of the contaminant contain capacity and contamination sensitivity (The contaminant retaining capacity and the sensitivity of the hydraulic components will be expanded at some probability level.) has great influence on the optimized results; (3) IFCNP successfully achieved some optimized results withλbeing 1, which means the constraints are tight ones and some risk exists. This method can be adopted as a referrence during the evaluation of the system.更多还原