论文部分内容阅读
针对电解铝厂生产过程中氧化铝输送流量在线测量问题,考虑到样本数据较少的因素,采用粒子群优化的最小二乘支持向量机方法,建立氧化铝超浓相输送中氧化铝粉流量的软测量模型.以粒子群优化的方法选取最小二乘支持向量机的模型参数,克服交叉验证法耗时与盲目性的问题,同时发挥最小二乘支持向量机的小样本学习能力强和计算简单的特点.通过采用实际数据进行仿真研究,结果表明,所建预测模型估计值与实际分析值吻合较好,从而验证了PSO-LSSVM预测模型对氧化铝粉流量准确估算的有效性.
In order to solve the problem of on-line measurement of alumina transport flow during the production of aluminum smelter, particle swarm optimization (LS-SVM) method was used to establish the flow rate of alumina powder Soft-sensing model.The particle swarm optimization method is used to select the least squares support vector machine model parameters to overcome the cross-validation method time-consuming and blind issues, while playing the least squares support vector machine small sample learning ability and simple calculation The simulation results show that the proposed model is in good agreement with the actual value, which verifies the effectiveness of the PSO-LSSVM model in accurately estimating the flow rate of alumina powder.