中国有色金属学报
中國有色金屬學報
중국유색금속학보
THE CHINESE JOURNAL OF NONFERROUS METALS
2010年
2期
323-328
,共6页
蔡从中%温玉锋%朱星键%裴军芳%王桂莲%肖婷婷
蔡從中%溫玉鋒%硃星鍵%裴軍芳%王桂蓮%肖婷婷
채종중%온옥봉%주성건%배군방%왕계련%초정정
7005铝合金%力学性能%支持向量机%粒子群算法%留一交叉验证法%回归分析
7005鋁閤金%力學性能%支持嚮量機%粒子群算法%留一交扠驗證法%迴歸分析
7005려합금%역학성능%지지향량궤%입자군산법%류일교차험증법%회귀분석
7005 Al alloys%mechanical properties%support vector machines%particle swarm optimization%leave-one-out cross validation%regression analysis
根据7005铝合金在不同工艺参数(挤压温度、挤压速度、淬火方式和时效条件)下的力学性能(抗拉强度σ_b、屈服强度σ_(0.2)和硬度HB)实测数据集,应用基于粒子群算法(PSO)寻优的支持向量回归(SVR)结合留一交叉验证(LOOCV)的方法,对7005铝合金力学性能进行建模和预测研究,并与偏最小二乘法(PLS)、反向传播人工神经网络(BPNN)和两者结合的PLS-BPNN模型的预测结果进行比较.结果表明:基于SVR-LOOCV法的预测精度最高,对3种力学性能(σ_b、σ_(0.2)和HB)预测的均方根误差(RMSE)分别为4.531 9 MPa、14.550 8 MPa和HB1.414 2,其平均相对误差(MRE)分别为0.72%、2.61%和0.66%,均比PLS、BPNN和PLS-BPNN方法预测的RMSE和MRE要小.
根據7005鋁閤金在不同工藝參數(擠壓溫度、擠壓速度、淬火方式和時效條件)下的力學性能(抗拉彊度σ_b、屈服彊度σ_(0.2)和硬度HB)實測數據集,應用基于粒子群算法(PSO)尋優的支持嚮量迴歸(SVR)結閤留一交扠驗證(LOOCV)的方法,對7005鋁閤金力學性能進行建模和預測研究,併與偏最小二乘法(PLS)、反嚮傳播人工神經網絡(BPNN)和兩者結閤的PLS-BPNN模型的預測結果進行比較.結果錶明:基于SVR-LOOCV法的預測精度最高,對3種力學性能(σ_b、σ_(0.2)和HB)預測的均方根誤差(RMSE)分彆為4.531 9 MPa、14.550 8 MPa和HB1.414 2,其平均相對誤差(MRE)分彆為0.72%、2.61%和0.66%,均比PLS、BPNN和PLS-BPNN方法預測的RMSE和MRE要小.
근거7005려합금재불동공예삼수(제압온도、제압속도、쉬화방식화시효조건)하적역학성능(항랍강도σ_b、굴복강도σ_(0.2)화경도HB)실측수거집,응용기우입자군산법(PSO)심우적지지향량회귀(SVR)결합류일교차험증(LOOCV)적방법,대7005려합금역학성능진행건모화예측연구,병여편최소이승법(PLS)、반향전파인공신경망락(BPNN)화량자결합적PLS-BPNN모형적예측결과진행비교.결과표명:기우SVR-LOOCV법적예측정도최고,대3충역학성능(σ_b、σ_(0.2)화HB)예측적균방근오차(RMSE)분별위4.531 9 MPa、14.550 8 MPa화HB1.414 2,기평균상대오차(MRE)분별위0.72%、2.61%화0.66%,균비PLS、BPNN화PLS-BPNN방법예측적RMSE화MRE요소.
The support vector regression (SVR) approach based on the particle swarm optimization (PSO) for its parameter optimization, combined with leave-one-out cross validation (LOOCV), was proposed to predict the mechanical properties (tensile strength σ_b, yield strength σ_(0.2) and hardness HB) of 7005 Al alloys under different processing parameters including extrusion temperature, extrusion velocity, quenching type and aging time. The results strongly support that the prediction precision of SVR-LOOCV method is superior to those of partial least squares (PLS), back-propagation neural networks (BPNN) and their combination PLS-BPNN model by applying the identical dataset. The root mean square errors (RMSE) for σ_b, σ_(0.2) and HB achieved by SVR-LOOCV are 4.531 9 MPa, 14.550 8 MPa and HB 1.414 2, respectively, and their mean relative errors (MRE) are 0.72%, 2.61% and 0.66%, respectively, which are less than those predicted by PLS, BPNN or PLS-BPNN approach.