有色冶金过程系统的不确定动态优化算法

项目名称： 有色冶金过程系统的不确定动态优化算法

项目编号： No.61463013

项目类型： 地区科学基金项目

立项/批准年度： 2015

项目学科： 自动化技术、计算机技术

项目作者： 牛林

作者单位： 红河学院

项目金额： 42万元

中文摘要： 实施先进控制和实时优化是现代过程工业保持最大效益化的关键所在，其核心技术是高效的动态优化算法和软件。目前，有色冶金过程不确定动态优化的研究尚在探索中。本项目针对有色冶金过程，提出一种不确定实时优化算法┄基于机会约束规划的动态优化算法。以有色金属锡熔炼过程为研究对象，采用数值分析、相似特性装置模拟等手段，探究有色冶金过程系统的优化问题。利用重要性采样和回归拟合构建机会约束的描述；基于多目标优化机制，寻求不确定动态行为影响系统优化经济指标的关键函数，建立机会约束优化算法。研究机会约束优化算法的求解：对机会约束逆向映射机制进行扩展，在满足特定的单调性条件下，将机会约束转化为与变量不确定方差相关的普通约束，再运用马尔可夫链蒙特卡尔理论，通过序列模拟算法，得到目标函数期望的求取。探寻过程系统的优化与控制一体化策略。项目完成后开发出具有自主知识产权的软件系统，实现部分理论成果在实际生产过程中的应用验证

中文关键词： 随机系统；不确定性；随机控制

英文摘要： The implementation of advanced control and real-time optimization is the key to the maximum efficiency of the modern process industry, which core technology is a dynamic algorithm and efficient software. The dynamic optimization of the non-ferrous metallurgical process is still studied at present. Aiming at the non-ferrous metallurgical process, an uncertain optimization algorithm -- dynamic optimization algorithm based on chance constrained programming is proposed in the project. Non-ferrous metal tin smelting process is used as the research object to explore the problem of optimization Non-ferrous metallurgical production process be means of numerical analysis and simulation in a Similar characteristics instrument. Important sampling and regression fitting form the description of the chance constraint. Based on the multi-objective optimization mechanism, the key function of the economic performance effected by uncertain dynamic behavior so as to build a constrained optimization algorithm. The solution of the opportunities constrained optimization algorithm is studied. That is, the chance constrained inverse mapping mechanism is extended. When the monotonicity is satisfied, the constrained constraints is transformed into the ordinary Constraint related to the variance of the variable uncertainty. Then using the Markov Chain Monte theory, the expected distribution of the objective function is obtained through sequence simulation algorithm. Finally, the integration of optimization and control in a system is explored. The software system will be developed with independent intellectual property rights after the completion of the project, a part of the theoretical results is verified in the actual production process.

英文关键词： stochastic system;uncertainty;stochastic control