中文摘要：

以TGA试验数据为基础,利用BP神经网络模型分析并预测了煅烧和碳酸化循环过程、钙基吸收剂活性以及循环反应速率和转化率的变化趋势.结果表明：钙基吸收剂经过多次碳酸化循环,其活性存在明显衰减趋势;钙基吸收剂的活性和化学反应速率不仅受到循环次数的影响,而且还受到试验样品物性、煅烧温度、煅烧时间以及煅烧气氛等的影响;BP神经网络模型可以直观、准确地描述钙基吸收剂的碳酸化循环反应特性,其模型数据可以减少试验次数,并可为模型钙基吸收剂煅烧和碳酸化循环反应过程以及反应活性的变化提供新的方式.

英文摘要：

Based on TGA test data,a detailed thermodynamic analysis has been performed using BP neural network model so as to predict the variation tendency of the activity,the reaction rate and the conversion rate of a calcium-based sorbent in calcination/carbonation cycles.Results show that the activity of the calcium-based sorbent reduces greatly with rising number of calcination/carbonation cycles;the activity and reaction rate may be affected not only by the cyclic number,but also by the physical property of sorbent,the calcination temperature,duration and atmosphere;PB neural network model may visually and accurately describe the carbonation cycles of calcium-based sorbent,and the model data may help to minimize the test frequency,which therefore may serve as a reference in simulation of calcination/carbonation cycles and the reactivity variation of calcium-based sorbents.