試驗設(shè)計與數(shù)據(jù)處理試驗報告

1、試驗設(shè)計與數(shù)據(jù)處理試驗報告試驗設(shè)計與數(shù)據(jù)處理試驗報告正交試驗設(shè)計1.為了通過正交試驗尋找從某礦物中提取稀土 元素的最優(yōu)工藝條件，使稀土元素提取率最高, 選取的水平如下：水平(A)酸用 量/ml(B)水用 量/ml(C)反應(yīng) 時間/h125201220402需要考慮交互作用有 A X B, A X C , BX C,如 果將A, B, C分別安排在正交表L8 (27)的1,2,4 列上，試驗結(jié)果(提取量/ml)依次是1.01, 1,33,1,13, 1.06, 1.03, 0.08, 0.76, 0.56.試用方差 分析法(a=0.05 )分析實(shí)驗結(jié)果，確定較優(yōu)工藝條件解：(1)列出正交表L8

2、( 27)和實(shí)驗結(jié)果，進(jìn)行方差分析。試驗號ABAX BCAX CBX C空號提取量(ml)i11111111.01211122221.3331221122106521212121.03621221210.8722112210.76822121120.56K14.534.173.663.933.53.663.63K23.153.514.023.754.184.024.05k12.2652.085.1.831.9651.751.831.815k21.5751.755. 2.011.8752.092.012.025極差R 1.38因素主次優(yōu)選方案0.660.360.180.

3、680.360.42A AX C B A X B B X CA1B1C1SS0.238050.054450.0162Q7.7816總和T7.68P=TA2/n7.3728SS-0.4088差異源SSdfMSA0.2380510.23805B0.0544510.05445A*B0.016210.0162C0.0040510.00405A*C0.057810.0578B*C0.016210.0162誤差e0.0220510.02205誤差e20.0364530.01215F 0.05(1 ，3) 10.12796449F 0.01 (1 ，小 34.116223) 156可見A因素對實(shí)驗有顯著性影

4、響0.00405 0.05780.01620.02205F顯著性19.5925 *92594.4814814811.3333333330.3333333334.7572016461.3333333331.814814815優(yōu)方案的確立：由上述分析可知，由于提取率越 高越好，且交互作用影響不顯著，所以優(yōu)方案為A1B1C1，即酸用量25ml，水用量20ml，反應(yīng)時間為1小時2.為了提高陶?；炷恋目箟簭?qiáng)度，考察了 A,B, C, D, E , F六因素，每個因素都有3個水 平，因素水平表如下：水平(A) 水泥 標(biāo)號(B) 水泥 用量/kg(C) 陶粒 用量(D) 含砂 率/%(E) 養(yǎng)護(hù) 方式(E

5、) 攪拌 時間/h130018015038空氣1240019018040水1.5350020020042蒸汽2根據(jù)經(jīng)驗還要考察交互作用 AX B, AX C, BXC。如果將A, B, C , D, E, F依次安排在正 交表 L27（331）的 1, 2, 5, 9, 12, 13 列上，試驗 結(jié)果（抗壓強(qiáng)度/kg）依次為100, 98, 97, 95,96,99,94,99,101,85,82,98,85,90,85,91,89,80,73,90,77,84,80,76,89,78,85,試用方差分析（a=0.05）試驗結(jié)果， 確定較優(yōu)水平組合。解:列出表,進(jìn)行方差分析。試驗號i23456

6、78910 11 1213ABAX B空列CAX C空列BX C D空列空列EF抗壓強(qiáng)度/KG.cmi1111 11 11 11 1 111002111131111412225122261222713338133391333102123112123122123132231142231152231162312172312182312193132203132213132223213233213243213253321263321273321K1879800795807K2785790807790K3732806794799k1293267265269k2262263269263k324426926

7、5266極差14716A1,B3,A*B1317因素主次2,C2,A *C3, B*C3,D1, E1, F23總和 T 2396.002222222298333333339711222333952233311196331112229911333222942211133399332221111012312312385312312318212312312982323131285313121239012123231852331223191311233128912231123803213213273132132139021321321773221332184133211328021132213763

8、2321213891313232178212131328580078878483679179082479479280280079980080478580180480681276180580278780126726326127926426327526526426726726626726826226726826927125426826726226712182875141439 -7231231231231231231231231237968027982652672666214418.00P=TA2/ 212622.n 81SS16.019.8 43.8 312 11.1 12.7 1071231.

9、63 14.52 11.632.078.30 廠5.5294“ 3.6375.19Fo.O5 (2,14)Fo.O5 (4,14)F 0.01 (2,14)F 0.01 (4,14)3.743.116.515.04方差分析表差異源SSdfMSf顯著性A1231.63 2.00615.81214.74*B14.522.007.262.53AX B11.634.002.911.01C2.072.001.040.36AX C8.304.002.070.72BX C43.854.0010.963.82*D312.522.00156.2654.49*E107.192.0053.5918.69*f3.63

10、2.001.810.63誤差e 59.858.007.482.61誤差 e2 40.1514.00 2.87優(yōu)方案的確立：在不考慮交互由圖可知，抗壓強(qiáng)度是越大越好，作用下，優(yōu)方案選取k最大水平，即A1B3C2D1E1F.從方差分析結(jié)果看，交互作用對實(shí)驗結(jié)果影響不大，所以不考慮交互作用，優(yōu)方 案為 A1B3C2D1E1F.3.用某種菌生產(chǎn)酯類風(fēng)味物質(zhì)，為了尋找最優(yōu)發(fā) 酵工藝條件，重點(diǎn)考察了葡萄糖用量x1（50150g/L）和蛋白胨用量x2（210g/L）的影 響，實(shí)驗指標(biāo)為菌體生長量y（ g/L），其他發(fā)酵條件不變。實(shí)驗方案和結(jié)果如下：試驗號z1z2z1z2y11119.6121-1-19.13

11、3-11-19.374-1-118.5751.078009.346-1.078008.97701.078010.2180-1.07809.48900010.241000010.33（1）試用二次回歸正交設(shè)計在試驗范圍內(nèi)建立二次回歸方程；（2）對回歸方程和回歸系數(shù)進(jìn)行顯著性檢驗；（3）失擬性檢驗；（4）試驗范圍內(nèi)最優(yōu)試驗方案的確定。解：試驗號ziZ2z1*z2z1z2yyA2z1A2z2A21.001.001.001.000.370.379.6192.351.001.002.001.00-1.00-1.000.370.379.1383.361.001.003.00-1.001.00-1.000.

12、370.379.3787.801.001.004.00-1.00-1.001.000.370.378.5773.441.001.005.001.080.000.000.53-0.639.3487.241.160.006.00-1.080.000.000.53-0.638.9780.461.160.007.000.001.080.00-0.630.5310.21104.240.001.168.000.00-1.080.00-0.630.539.4889.870.001.169.000.000.000.00-0.63-0.6310.24104.860.000.0010.000.000.000.00

13、-0.63-0.6310.33106.710.000.00SUMMARYOUTPUT回歸統(tǒng)計MultipleR0.99R Square0.98Adjusted0.96R Square標(biāo)準(zhǔn)誤差0.11觀測值10.00方差分析dfSSMSFSignifican ce F回歸分析5.003.020.6048.070.00殘差4.000.050.01總計9.003.07Coefficients標(biāo)準(zhǔn)誤差t StatP-valueLower 95%Upper 95%下限)95.0%上限95.0%Intercep t9.530.04268.630.009.439.629.439.62z10.190.044.

14、250.010.070.310.070.31z20.330.047.330.000.200.450.200.45z1*z20.000.060.001.00-0.160.16-0.160.16z10.000.070.001.00-0.190.19-0.190.19z2-0.250.07-3.670.02-0.44-0.06-0.44-0.06y=9.53+0.19z1失擬性檢驗差異源SSdfMSFF0.1因為失擬（If)0.053.000.023.8153.59ff0.1 ，所以失擬不顯著重復(fù)試驗（ef）0.001.000.00(1)回歸方程的表達(dá)式為 y=9 53+0 19z1+0 33z2-0 25z2z1=x1-100/50z2=z2-6/4 z2 =z2*z2-0.63 代入上式中可知Y=9.53+0.19*(x1-100)/50+0.33*(x2-6)/4-0.25*(x2-6）A整理可知 y=8.07+0.038x1+0.275x2-0.016乂2八2偏回歸系數(shù)顯著性檢驗：由上表“t-Stat”和P-