【摘 要】
:
According to the smooth operation and automatic control problem for the distillation column with side reactors (SRC), the integrated structure design, process simulation calculation and automatic cont
【机 构】
:
Nanjing University of Technology School of Automation & Electrical Engineering School of Chemistry a
论文部分内容阅读
According to the smooth operation and automatic control problem for the distillation column with side reactors (SRC), the integrated structure design, process simulation calculation and automatic control system of SCR process for chorobenzene production are discussed in the paper. The mechanism models and process simulation systems are established respectively, and then multivariable control structure schemes are designed with the traditional control strategy. The dynamic simulation system including closed-loop control system is developed, in which the different types of disturbances are added, such as feed flow rate disturbances, composition disturbance, the boiling steam heat value disturbance and setting-value change. Through analyzing quantitatively performance of closed-loop system again different disturbances, the effectiveness and controllability of the proposed control structure are verified.
其他文献
针对含有复杂约束条件的非线性最优控制问题,提出了一种改进的Gauss伪谱法(IGPM)。这类问题难以得到解析解,特别是有些问题不存在解析的模型,一些参数只能通过查表得到,使得传统方法难以求解。在传统的Gauss伪谱法的基础上,将非线性的终端状态积分约束等价地转化为线性形式,提出了IGPM,通过协态映射定理可以计算出协态变量,检验最优性,使得IGPM具有间接法一样的精度。并且给出了初始时刻协态变量和
在网络控制系统中,当传感器数据通过网络进行传输时,时间延迟几乎是不可避免的。本文针对具有随机一步延迟的多传感器系统,通过引入随机变量并利用扩维的方法将随机一步延迟系统转化为带随机参数的系统,然后针对新系统利用最小化估计误差协方差矩阵的准则设计融合估计器,给出了序贯式滤波估计融合算法。单传感器的随机一步延迟滤波算法在已有的研究中已经表现出良好的效果,本文的主要工作是将其推广到多传感器融合,从而通过信
在多模型软测量建模中,对于新的数据以及异常样本点,传统的聚类方法没有充分考虑它们的特性,因而所属类别往往不能反映其真实属性,最终导致模型精度不高。本文提出一种基于最小环路能量聚类的算法,该方法将样本聚类转化为寻找一个最小能量环问题, 通过模拟退火算法搜索一条经过所有样本点的最小能量环实现样本集的聚类;对侦破出的异常样本点和新的测试数据根据其能量值确定其所属属性,从而提高聚类和分类精度;然后利用支持
基本多目标粒子群算法在收敛性和分布性上都表现一般,且容易陷入局部最优。为此,本文提出了一种新的用于求解多目标问题的粒子群算法,该算法采用一种新的全局极值和个体极值选取策略,提升了种群逼近Pareto最优前沿的稳定性和精度,同时为了提升种群跳出局部最优的能力,提出两步变异操作。此外本算法还采用了外部存档存储每一代产生的非支配解,并且使用动态更新的拥挤距离来维持外部存档的规模。然后,通过典型的ZDT系
The nonlinear dynamical model of wind turbine generator drive system is presented considering nonlinear stiffness of the drive system. The frequency characteristics of wind turbine drive system incent
A model predictive controller is proposed for path following of survey ships with respect to control constraints. The control system is a typical nonlinear system which has characters such as strong n
太阳能-相变蓄热新风系统是利用太阳光的热量和相变蓄热装置的潜热蓄热特性为室内提供具有一定温度新风的辅助供暖系统。本文首先介绍了新风系统的原理和三种运行模式;然后通过对新风系统运行特性和模糊控制算法的特点分析,提出把模糊算法应用于系统的新风控制,并设计基于多输入单输出(MISO)的Mamdani型模糊控制器,实现了通过自动调节热媒水的流速对风机盘管的新风出口温度的模糊控制。经过运行实际的新风系统实验
Advanced feedback control for optimal operation the mineral grinding process has been based on the model predictive control (MPC) dynamic optimization for years. Due to the MPC does not handle the dis
In this paper, an improved hybrid differential evolution algorithm (IHDE) is proposed for nonlinear and mixed-integer nonlinear programming models (NLPs and MINLPs) in engineering optimization fields.
考虑复杂制造系统调度的特点,设计并实现了蚁群优化算法仿真平台。与其他类似平台相比,该平台首次实现了工件的多重入特性与设备的批加工能力。以半导体生产线批加工设备调度为例,给出了平台的应用效果。仿真结果表明,该平台具有充分的柔性,适合复杂制造系统调度问题求解。