基于产生式知识表示的铝用炭素煅烧过程专家控制系统的设计

将基于知识的规则模型和解析的数学模型相结合，建立两级结构的实时专家控制系统(ECS)。由具有专家自学习和自适应功能的实时专家系统(ES)，以专家优化方式寻找煅烧过程中煅烧带的最优温度、长度和位置，由实时控制系统(RTCS)采用基于CARMA模型和最小二乘法的自校正调节算法和经典PID算法分别进行给料量和负压的控制，保证生产过程在最优生产条件下进行。系统于2003年底正式投入运行，长期运行结果表明系统可靠性高、实用性好，既稳定了生产，又提高了产品质量，实现了节能。

Design of expert control system based on productive knowledge representation for carbon calcinations process in aluminum production

By integrating knowledge rule model and analysis mathematical model, the two-leveled real-time expert control system (ECS) was established. With real-time expert system (ES) with self-learning and self-adaptation ability, the optimal temperature, length and location of calcinations’ zone were obtained through expert optimization way. And real-time control system (RTCS) adopted self-adjusting algorithm based on CARMA model and least square method and classicPID algorithm to control the flux and negative pressure. The system put into operationat the end of 2003 and the long-term operation showed that this system was reliable and practical; besides, the product quality was improved with the reduction of energy consumption.