CAJ | 학술논문

考虑电弧炉终点参数既受定量因素的影响,又受非定量因素的影响.将灰色马尔可夫 (grey Markov)与支持向量机 (SVM)相结合.建立了电弧炉终点参数grey Markov-SVM预报模型,其中grey Markov模型反映非定量因素对电弧炉终点参数预测值的影响,SVM模型反映电弧炉各种定量输入对终点参数预测值的影响.建立grey Markov-SVM模型的方法是:首先建立反映非定量因素的GM (1,1)模型,然后用Markov链修正其预测值:由于grey Markov模型对定量输入的影响无法准确反映,因此grey Mar-kov模型必然存在预测偏差,此预测偏差通过建立反映定量输入与终点参数预测偏差之间关系的SVM模型方法加以补偿,并采用粒子群算法 (PSO)对SVM的参数进行寻优,最终得到电弧炉终点参数的预报值,同时实现滚动预测.仿真实验表明grey Markov-SVM模型与grey-SVM模型、Markov-SVM相比较,具有很高的终点预报精度.
고필전호로종점삼수기수정량인소적영향,우수비정량인소적영향.장회색마이가부 (grey Markov)여지지향량궤 (SVM)상결합.건립료전호로종점삼수grey Markov-SVM예보모형,기중grey Markov모형반영비정량인소대전호로종점삼수예측치적영향,SVM모형반영전호로각충정량수입대종점삼수예측치적영향.건립grey Markov-SVM모형적방법시:수선건립반영비정량인소적GM (1,1)모형,연후용Markov련수정기예측치:유우grey Markov모형대정량수입적영향무법준학반영,인차grey Mar-kov모형필연존재예측편차,차예측편차통과건립반영정량수입여종점삼수예측편차지간관계적SVM모형방법가이보상,병채용입자군산법 (PSO)대SVM적삼수진행심우,최종득도전호로종점삼수적예보치,동시실현곤동예측.방진실험표명grey Markov-SVM모형여grey-SVM모형、Markov-SVM상비교,구유흔고적종점예보정도.
Because the end-point parameters of an electric arc furnace (EAF) are affected by both quantitative factors and non-quantitative factors, we combine the grey Markov model with support-vector-machines (SVM) to produce a grey Markov-SVM prediction model for estimating the end-point parameter values of an EAR The effects from the nonquantitative factors on the prediction values of end-point parameters are reflected by the grey Markov model; while the effects from the quantitative inputs are reflected by the SVM. The GM (1,1) model that reflects non-quantitative factors is established firstly, and then, its prediction values are revised by the Markov chain. Because the effect from the quantitative inputs can not be reflected by the grey Markov model, the grey Markov-model is certainly not free from prediction errors from the quantitative inputs. These prediction errors are compensated by the SVM model with parameters optimized by particle swarm optimization (PSO) algorithm. The final prediction values of the end-point parameters in EAF are thus obtained. Meanwhile, the rolling forecasting is realized. Experiments show that the grey Markov-SVM model has the best prediction precision in comparison with the grey SVM model or the Markov-SVM model.