网络环境下基于多智能体陶瓷梭式窑群信息融合与混杂智能控制方法研究

The project is about the research on information fusion and hybrid intelligent control methods for ceramic shuttle kiln groups based on multi-agent under network enviroment.The intelligent optimized information fusion algorithms which can identify the real conditions of shuttle kiln burning zone are obtained based on the information fusion theory and the flame image recognition method.In addition, the hybrid intelligent control methods and algorithms of shuttle kiln groups under the network environment are got based on the network control theory,intelligent control theory, muliti-agent theory and hybrid system theory.The above research is with important theoretical significance and engineering application value for the promotion of the progress of control science and technology and the development of Chinese ceramics industry.. At first, the hybrid system model of single shuttle kiln is established by using the hybrid modelling method of mechanism analysis method, expert experience method, fuzzy identification method and hybrid sysytem modelling theory, and so on; Secondly, the hybrid system model of ceramic shuttle kiln groups under the network environment is established on the basis of the above; Again, the intelligent optimized information fusion algorithms which can identify the real conditions of the shuttle kiln burning zone are obtained based on the information fusion theory and the flame image recognition method; In the end, the multi-agent-based hybrid intelligent optimized control methods and algorithms of shuttle kiln groups under the network environment are got based on networked control system model of shuttle kiln groups; In the meantime,the system simulation platform is built to verify the theoretical results obtained.

本课题研究网络环境下基于多智能体陶瓷梭式窑群信息融合和混杂智能控制方法。以信息融合理论和火焰图像识别方法为基础，研究得到梭式窑烧成带工况识别智能优化信息融合算法。另外，以网络控制理论、智能控制理论、多智能体理论和混杂系统理论为基础，研究得到网络环境下基于多智能体梭式窑群智能优化控制方法和算法。以上的研究对于促进控制科学技术进步以及中国陶瓷工业发展具有重要的理论意义和工程应用价值。. 本项目首先利用机理分析、专家经验、模糊辩识和混杂系统建模理论等相结合的混合建模方法建立单个陶瓷梭式窑混杂系统模型；其次，在以上基础上建立网络环境下陶瓷梭式窑群混杂系统模型；再次，基于信息融合理论和火焰图象识别方法研究得到识别梭式窑烧成带工况智能优化信息融合算法；最后，基于梭式窑网络控制系统模型，研究得到网络环境下基于多智能体陶瓷梭式窑群混杂智能优化控制方法和算法；同时，构建系统仿真平台，验证所得理论结果。

陶瓷梭式窑是陶瓷工业应用最广的现代间歇式窑炉之一，其检测与控制方式直接影响着陶瓷产品质量和能源利用效率。本课题针对陶瓷梭式窑的热工检测精度低、生产自动化水平低、控制方法落后以及陶瓷梭式窑网络控制系统存在数据传输延迟、数据包丢失等问题研究能够提高陶瓷梭式窑烧成工况检测精度、提升陶瓷梭式窑生产自动化水平、改善控制效果并提高梭式窑网络控制系统稳定性的方法。. 本项目针对网络环境下陶瓷梭式窑群的建模、检测与控制开展相关的研究工作。首先针对陶瓷梭式窑利用机理分析、专家经验、频率子空间辨识方法和数据驱动建模理论等相结合的混合建模方法建立了其模型；其次，基于信息融合理论和火焰图象识别方法设计了陶瓷梭式窑火焰图像模式识别分类器、提出了基于K均值聚类陶瓷窑炉烧成工况火焰图像分割方法以及识别陶瓷梭式窑烧成工况智能优化信息融合方法和算法；再次，基于已有的网络控制系统和智能控制的研究成果，研究了并提出网络环境下陶瓷梭式窑群智能混杂控制方法，同时讨论了网络控制系统存在数据传输延迟、数据包丢失等的控制问题以及设计了基于基于观测器的反馈控制器，获得了相应的控制方法和算法。同时，构建了系统仿真平台，仿真结果表明所提出的方法是有效和可行的。以上的研究对于促进控制科学技术进步以及中国陶瓷工业发展具有重要的理论意义和工程应用价值。