基于知识模糊迁徙的城市污水处理膜污染决策

针对城市污水处理膜污染难以精准决策的问题,提出一种基于知识模糊迁徙的膜污染决策方法。首先,结合城市污水处理运行过程数据和运行经验,利用模糊规则的形式实现膜污染决策知识的表达;其次,提出一种知识重构机制(knowledge reconstruction mechanism,KRM),动态平衡源域与目标域之间的准确性和多样性,并采用知识迁徙的方法完成决策知识重构;最后,建立一种基于数据和知识驱动的区间二型模糊神经网络(data-knowledge-driven interval type-2 fuzzy neural network,DK-IT2FNN)的决策模型,利用模糊规则设计模型参数,采用迁徙梯度下降算法动态调整网络权值,提高决策精度。实验结果表明,该模型能够实现膜污染的精准决策。

KNOWLEDGE RECONSTRUCTION AND JUSTIFICATION FOR REGIONAL VITALIZATION

This paper considers the regional vitalization problem and discusses the methodology to create regional vitalization plans, which include activating the local economy, enriching people＇s lives, and activating the feelings of people, by new initiatives. Activily underlying the methodology is the experience of implementing several actual projects with the local residents, and theory underlying the methodology is the knowledge construction and justification theory based on knowledge management and systems thinking. Introducing an actual vitalization project as an illustrative example, the paper proposes a knowledge reconstruction and justification procedure for regional vitalization.