|
| 基于改进粒子群优化算法的动态数据校正技术研究 |
| Study of dynamic data reconciliation based on a novel particle swarm optimization |
| |
| DOI： |
| 中文关键词: 粒子群优化 不确定知识 动态数据校正 离群值 连续搅拌反应槽系统 |
| 英文关键词:PSO algorithm uncertain knowledge dynamic data rectification outlier continuous stirred tank reactor |
| 基金项目: |
|
| 摘要点击次数: 2482 |
| 全文下载次数: 2137 |
| 中文摘要: |
| 动态数据校正是根据过程模型的数据校正机制降低测量值中离群点的影响,消除随机误差,使得到的估计校正值满足能量守恒和质量守恒定律.提出了一种基于不确定知识的粒子群优化算法的动态数据校正和离群值检测方法,改进的算法在粒子群优化算法的基础上引入不确定知识,在粒子寻优迭代过程中限定粒子的搜索中心和搜索范围,提高了粒子可信度,克服了粒子早熟问题,使得粒子以更快、更有效的方式向最优值靠近.连续搅拌反应槽系统(CSTR)仿真实验证明了改进方法的有效性. |
| 英文摘要: |
| Dynamic Data reconciliation (DDR) is the adjustment of a set of process data based on models of the process so that the derived estimates conform to natural laws of energy and mass conservation.A novel particle swarm optimization based on uncertain knowledge (PSOUK) for DDR and outlier detection is proposed in this paper.The uncertain knowledge is introduced in particle swarm optimization (PSO) algorithm,which limits the search center and range of particles,so that the confidence of particles is improved and the premature problem can be overcome.The particles fly more quickly and effectively toward the direction close to the optimal solution.The simulation of a continuous stirred tank reactor (CSTR) verifies the effectiveness of the proposed algorithm. |
| HTML 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
