论文部分内容阅读
采用基于机理分析获得的状态空间模型和非线性系统的自适应推广的Kalman滤波算法实现了对铝电解槽Al_2O_3浓度的直接估计。用虚拟噪声补偿了简化模型包含的时变误差和过程中的时变噪声。用实测数据证明了模型与算法的有效性。
A direct estimation of the concentration of Al_2O_3 in aluminum reduction cell is achieved by using the Kalman filter algorithm based on the state space model obtained through the mechanism analysis and the adaptive extension of the nonlinear system. The virtual noise compensates for the time-varying error contained in the simplified model and the time-varying noise in the process. The measured data prove the validity of the model and the algorithm.