Font Size: a A A

Study On Steeping Principles, Process And Intelligence Expert System For Corn Wet Milling

Posted on:2007-06-07Degree:DoctorType:Dissertation
Country:ChinaCandidate:G L PengFull Text:PDF
GTID:1101360185955314Subject:Agricultural mechanization project
Abstract/Summary:PDF Full Text Request
Steeping is one of key steps of wet milling process, steepingaffects starch yield and quality. In the whole wet milling process,steeping needs the longest time and more energy, so unfittingsteeping process parameters not only adds cost but also decreasesproduct quality and yield. It is necessary that searching effectivemethods to decide the most optimizing steeping process parametersto rise yield, reduce energy and improve quality of starch. This paperstudies on steeping principles, processes and builds an intelligenceexpert system for decision of steeping process parameters.1)By Lagrange's method and Euler's method, the water transfertheory was studied for corn steeping process, mathematic models ofmass transfer in steeping process were developed. Lagrange'smethod has clearly theoretic sense, but has some problems inpractical application because of defining covariance and correlativetime. Euler's method is similar to Fick's model, they can describe thecorn steeping process. Based on the water potential concept, newwater transfer theory of corn steeping process was put forwards, It isconsidered that water mobility in corn kernels is caused directly bywater potential difference. Water potential is increased by addingmore steeping solution, so steeping velocity also is increased. Thewater transfer theory on water potential can describe steeping processof corn better.2) On the basis of dynamic analysis of corn steeping,thefollowing conclusions can be drawn: As temperature is increased, thewater amount entering corn kernels is increased and time of gettingsaturation moisture is shortened. But the saturation moisture valuehas no obvious difference. SO2 will accelerate absorption of corn, butthe SO2 concentration has no obvious effect to absorption. When SO2concentration changes from 0.1% to 0.3% , there were no obviousdifference in absorption and time of reach saturation moisture. Timeof reach the saturation moisture in SO2 steeping solution is shorterthan in water, but the saturation moisture has no obvious changes. Asthe temperature becomes lower, the SO2 effect is more distinct.Model of corn steeping was built up by Logstic's equation, theapproach of nonlinear regression is applied to calculate equationcoefficient, and comparing the calculated values to experimentresults, the relative error between them is less than 1.5%, thecorrelation coefficient is more than 0.99. So the steeping model onLogstic's equation is established and it agrees with the waterabsorbing progress better. A model has been built up which has onlyone line direction way to enter steeping solution, it can calculatediffusion coefficient in different conditions. The result of the contrastexperiments is that the measured and the forecasted are fitting. Themathematic model of steeping could calculate moisture content atdifferent time and positions, and it reflects water transfer regular incorn better.3) Through the 2nd orthogonal regression experimental designmethod, a starch yield prediction model has been established to SO2concentration, steeping temperature and steeping time. Theoperational conditions have been optimized by MATLAB: steepingtemperature 51.2℃, SO2 concentration 0.18%, steeping time 52.6h,and yield of starch 67.01% which is accord to 67.3% by testing. Butthe initial moisture, drying conditions and corn categories areunconsidered. In the further research, establish the mathematicalmodel under different categories, different drying conditions andinitial moisture, for developing intelligent expert system of decidingcorn steeping process parameters.4) We investigate the steeping characteristic of the corn grainunder different drying temperature by use of steeping index,diffusion coefficient and saturation moisture. When the drytemperature is low than 80℃, steeping index decreased along withthe raise of the drying temperature, but it was still above of 200, thesaturation moisture and the diffusion coefficient have no obviousdifference. When the dry temperature gets to 80℃, the steepingindex will be lower than 200, the saturation moisture will fall, andthe diffusion coefficient will increase. The yields of starch powderreduce and protein content rise with drying temperature and initialmoisture content go rise. Yield of starch is decreased by 2% at initialmoisture content 32.4%, 80℃, it is 8.83% when is 90℃. Whentemperature is high than 80℃, protein in starch is over 0.4%. Whenthe condition is initial moisture content 25.6%, temperature 90℃and 100 ℃ , yield of starch powder is decreased by 2%, 9%respectively;protein content is over 0.4% when temperature is over90℃.So the operational temperature condition is not more than 80℃.5) Several steeping processes was studied by putting cellulose,protease and compound enzyme into steeping solution. Threeenzymes can reduce the dosage to SO2 and shorten steeping time,increase the yield of starch powder, the yield of starch powder inthree enzymes act rise 1.0%, 2.7%, 2.96% respectively, protease andcompound enzyme are much better by rising yield, but protease madethe progress complicated. The compound enzyme is he best chooseand the best progress is SO2 concentration 0.1%, time 12h forsteeping, enzyme dosage 0.3ml.6) On the studying of the corn starch absorption character thefollowing conclusions can be drawn: Isotherms of corn starch havethe properties of Ⅱstyle, the equilibrium moisture content increasedwith temperature rise under water activity. There was a time-lagbetween the whole water activities, At the same water activity, theequilibrium moisture content of desorption is higher than that ofabsorption. GAB model, peleg model and Henderson model have agood fitting for the isothermal sorption of corn starch powder, whichare better than Ferro-Fontan model and smith model, Oswin modelwas the worst for fitting, Mod-BET model and Halsey model can notat all. Analysis showed that BP neural network model not onlyaccommodated temperature and water activity parameter, but also ismore accurate than other mathematical models what only include oneparameter, and the relative error is 2.96% in the absorption modeland 4.84% in the desorption model. BP network model can be used todescribe sorption isotherms of corn starch powder.7) By the Graph Programme of LabVIEW, the IntelligenceExpert System was developed for deciding steeping processparameters. The reasoning pattern of the expert system is leading bythe example collections and rule collections. Based on the originalinformation of corn, dry information, storage information andmanufacture process information, reasoned the steeping processparameters.
Keywords/Search Tags:Corn starch, Wet milling, Steeping, Principle, Process, Model, Expert System
PDF Full Text Request
Related items