论文部分内容阅读
针对钢铁企业高炉煤气发生量的机理模型难以对其进行预测的问题,建立了基于Elman神经网络和最小二乘支持向量机相结合的预测模型。预测前利用概率神经网络对其进行分类,并对分类后的数据进行HP滤波处理,得到趋势序列和波动序列分别预测;预测后引入马尔科夫链的状态转移矩阵,对预测残差进行修正。组合建立的PNN-HP-Elman-LSSVM-MC分类预测模型,减少了训练时间,同时也提高了预测精度。根据企业实际数据应用模型,结果表明,所建模型不同工况分类准确,预测效果良好,为合理调度副产煤气提供操作依据。
Aimed at the problem that it is difficult to predict the amount of blast furnace gas in steel enterprises, a prediction model based on Elman neural network and least square support vector machine is established. It is classified by probabilistic neural network before the prediction, and the classified data are filtered by HP. The trend sequence and the fluctuation sequence are predicted respectively. After the prediction, the state transition matrix of Markov chain is introduced to correct the prediction residual. Combined PNN-HP-Elman-LSSVM-MC classification prediction model reduces the training time, but also improves the prediction accuracy. According to the actual data application model of the enterprise, the results show that the model is classified accurately according to different working conditions, and the forecasting effect is good, which provides the operational basis for the reasonable dispatch of by-product gas.