【作者】 庄海宁；

【导师】 金征宇；

【作者基本信息】 江南大学 ， 粮食、油脂及植物蛋白工程， 2011， 博士

【摘要】 双螺杆挤压法比单螺杆挤压法加工性能稳定、机械化、自动化程度高,在“人造米”生产中已逐渐得到应用。但是,目前有关双螺杆挤压法生产人造米或营养米的研究,仍然存在以下不足:双螺杆挤压机挤压模型仍待建立,人造米挤压成型技术的研究仍待完善,人造米的营养强化仍需扩展。因此,本课题主要从双螺杆挤压机的操作参数与系统参数及目标参数之间的关系研究入手,计算了人造米挤压过程中的滞留时间分布(RTD),利用神经网络建立双螺杆挤压机人造米理化性质与系统参数的预测模型,在此基础上,设计出一种高仿真米形开孔模头,优化了基于该模头的人造米成型工艺,最后,开发了一种添加有凉粉草胶的营养米挤出物,并对其理化性质及其体外抗氧化功能进行了评价。本课题主要结论如下:(1)通过旋转中心组合设计,考察了挤压人造米在生产过程中挤压加工变量对系统参数的影响,结果表明:在人造米生产过程中,物料温度、扭矩、压强、比机械能(SME)、产品水分含量这些系统参数都非常明显的受到挤压加工变量的影响,其中受螺杆转速和进料水分的影响最为显著(P<0.001)。合适螺杆转速和进料水分含量的控制对于人造米生产节能和品质稳定来说意义重大。(2)通过旋转中心组合设计,考察了挤压人造米在生产过程中挤压加工变量对产品物理性质及糊化特性的影响,结果表明:本试验条件下,人造米膨胀度平均为1.23±0.10,说明本实验取值范围属于较为理想的挤压人造米加工条件。在本实验所设定的操作参数范围内,能够对产品膨胀度、容积密度、糊化度、峰值黏度、终值黏度和回复值这6个指标建立起合适的响应面预测模型,这些模型的建立亦为人造米生产中实现质量预测和在线自动化控制提供参考;但颜色指标L,a,b,ΔE以及水溶性指数(WSI)这5个指标没有能够建立起合适的响应面预测模型。因而,有必要寻找更为科学合理的方法,为各项指标建立起更为合适的模型而用于指导生产实践。(3)本实验中,米粉挤压的平均滞留时间在2.09-3.99min,σ2值在1.5-4.45min2之间。D/uL值在0.081-0.255之间,本试验的扩散数模式属于混合流中的大量扩散模式(D/uL>0.2)。四个操作参数对挤压机机筒内物料的平均滞留时间(MRT)都有显著影响,其中进料速度对MRT影响最显著。操作参数与MRT间,可利用回归分析得到较为精确的RTD预测数学模型。(4)为了寻找更为科学合理的方法,为各项指标建立起更为合适的模型而用于指导生产实践,本课题开发了基于BP神经网络的PTW-24/25D实验型双螺杆挤压机的人造米理化性质与系统参数预测模型。该模型是由4个输入神经元(螺杆转速、进料速度、水分含量、机筒温度),1个隐含层(包括36个神经元)和18个输出神经元(吸水性指数(WAI),水溶性指数(WSI),亮度(L),红度(a),黄度(b),色差(dE),峰值黏度peak visc),终值黏度(Final visc),回落值(setback),糊化度(DG),堆积密度(BD),膨胀率(ER),水分(M.C.),平均滞留时间(MRT),人造米挤出温度(TM-E3)模头压(p),扭矩(M),比机械能(SME))构成,实现了对人造米挤压时的系统参数和目标参数的很好预测;在此基础上,开发了基于Matlab GUI的友好可视化用户交互界面,该界面操作简单、使用方便。(5)以挤压机物料流量和大米形状数值描述计算为基础,结合双螺杆挤压机螺杆和模口特性线的绘制,设计了具有高仿真米形开孔的模头,进行了人造米成型切割试验,所得人造米的米形与自然大米非常接近,表明模头设计非常成功;以千粒重、容重指标对人造米质量进行优化,成功的建立了以螺杆转速(A),进料速度(B),切割速度(C)预测人造米干态千粒重的数学模型,其中,千粒重=45.06+0.88×A+9.24×B-6.33×C+2.1×AC-1.77×BC + 1.03B2,R2=0.9429。(6)为了扩展大米中营养添加剂的开发领域,首次将我国特有中草药资源凉粉草胶添加入米粉中,并考察了营养米挤出物的理化性质和体外抗氧化功能,结果表明:随着MBG添加量的增加,挤出物灰分显著增加,蛋白质含量下降,糊化度(DG)呈先下降后上升的趋势;吸水指数(WAI)与水溶性指数(WSI)呈上升的趋势;膨化率相对米粉均下降;容积密度先略微减小而后增加;L*(亮度)下降;a*(红绿度)与b*(黄蓝度)先上升,达到最高点后下降;硬度先上升后下降。不同比例混合的凉粉草胶-大米挤出物的萃取液中,添加量为15%的大米挤出物的多酚含量最大,为15.22μg/ml。其中,添加量为15%的大米挤出物对DPPH自由基清除能力最强,半清除率(IC50)浓度为0.015μg/ml;添加量为10%大米挤出物对羟基自由基的清除能力较强。挤出物对超氧阴离子自由基O2-·并没有显示出清除作用;添加量为15%的大米挤出物对Fe2+的螯合能力最强。更多还原

【Abstract】 Twin-screw extrusion has been gradually applicated in fabricated rice because of its more stable processing capacity, higher automation than those of single-screw extrusion. However, there are still some drawbacks in the current research of producing fabricated rice or nutrient rice by twin-screw extrusion. These problems are unefficient extruding mode of twin-screw extrusion, unmature technology of rice molding and expansion of fabricated rice nutrition fortification.Thus, the present thesis is studied from the following aspect.The relationship between operation parameter, systematic parameter and objective parameter of twin-screw extruder has been set up; neuro network prediction model based on operation parameter was also set up, the residence time distribution of twin-screw extruder was studied, a die with rice shape hole has been designed,and the fabricated rice molding technical process has been optimized, a man-made nutrient rice with Mesona Blumes gum addtition has been developed and the physicochemical properties and its antioxidation capacity in vitro has been evaluated. The main conclusion was following like these:(1) During the fabricated rice processing, systematic parameters such as extrudates temperature, torque, pressure, SME, moisture content and product expansion ratio are obviously affected by the operation parameters of extrusion. Extrudates temperature, SME, product expansion ratio increased with the screw speed increment, while die pressure, torque and product moisture content decreased with screw speed increment. Increasing the moisture content of feed materials could obviously reduce extrudates temperature, torque, pressure, SME and product expansio ratio, the control of appropriate screw speed and feed moisture content is very significant for fabricated rice processing in energy saving and quality stablility.(2) In this study, expansion ratio of fabricated rice is 1.23±0.10, which is a relatively ideal processing condition of extruded fabricated rice. Expansion ratio of product increased with screw speed, feed rate and final barrel temperature increment, but decreased with feed moisture content increment. The predicted mathematic model of systematic parameter and expansion ratio of product was set up during fabricated rice extrusion by response surface methodolgy (RSM), which could provide reference to quality prediction and on-line control during fabricated rice processing. Within the operation parameters, the prediction model of RSM of ER, BD, DG, PV, FV and setback could be set up. These models can be used to provide reference for quality prediction and on-line automation control during fabricated rice processing. The appropriate RSM predicted model couldn’t be set up for L, a, b,ΔE and WSI. Thus, there is necessary to find a more scientific and reasonable method for guiding the pratical production with a more proper model for each parameter.(3) In this study, MRT of rice extrusion is between 2.09-3.99min,σ2 is between 1.5-4.45min2, D/uL is between 0.081-0.255. The diffusion mathematic model belongs to the large scale diffusion model of mixed flow (D/uL>0.2). The influence of operation parameters on mean residence time (MRT) of extrudates within barrel is very significant, among them feed rate has the most significant effect on MRT. MRT of extrudates within barrel remarkably decreased with the feed rate increment. The appropriate mathematic regression model of operation parameter, MRT and RTD can be obtained by regression analysis.(4) In order to find a more scientifically sensible method, more appropriate model of each parameter could be set up to direct processing. The BP neuro network of PTW-24/250 experimental twin-screw extruder was built up. The systematic and objective parameters of fabricated rice could be predicted from operation parameters by this BP neuro network. The BP network is composed of 4 input neurons (screw speed, feed rate, moisture content of feed and barrel temperature), 1 hidden layer(36 neurons) and 18 output neurons including WAI, WSI, L, a, b, dE, PV, FV, setback, DG, BD, ER, M.C., MRT, TM-E3, P, M, SME. The systematic and objective parameters of fabricated rice extrusion could be well predicted by this network. On the basis of the network, friendly visual user interactive interface from Matlab GUI has been developed. This interface is very simple and convenient for user to predict systematic and objective parameters of fabricated rice through extrusion on PTW-24/25O experimental extruder.(5)Based on material flow and descriptive calculation of rice shape, combined with drawing of characteristic curve of screw speed and die, the special die with rice-shape hole has been designed. The die is used into the moulding and cutting experiment of fabricated rice. It was found that the shape of fabricated rice was very close to that of natural rice, which indicated that the design of such a die is very successful. The thousand seed weight and bulk density were taken as the quality level of fabricated rice and the screw speed (A),feed rate (B) and cutting speed (C) were taken as operation parameters. Prediction model of thousand seed weight of dry fabricated rice has been built up successfully through central combined design(CCD) of RSM, which provides a useful exploration for setting product standard of fabricated rice industry. Among them, TSW=45.06+0.88×A+9.24×B-6.33×C+2.1×AC-1.77×BC+1.03B2, R2=0.9429.(6) In order to expand the developing field of rice nutrient additives, Mesona Blumes gum as our traditional herbal resource was firstly added into rice flour, and then the physicochemical and antioxidant activities in vitro of nutrient rice extrudates were explored. The results indicated that ash content increased remarkably with MBG addition, protein content decreased but the influence was not very obvious, DG decreased firstly and increased lately, WAI and WSI always increased. Compared with rice flour, ER reduced, but with MBG addition, ER increased, BD reduced slightly and then increased. L*(lightness) decreased. a* (redness) and b*(greenness) firstly increased and then decreased, hardness increased firstly and then decreased. The phenol content of 15% MBG-rice extrudate 15.22μg/ml is the largest among the MBG-rice extrudates within the MBG addition levels. DPPH radical scavenging capacity of 15% MBG-rice extrudate is the strongest; IC50 concentration is 0.015μg/ml. Hydroxyl radical scavenging capacity of 10% MBG-rice extrudates is the strongest. There is no superoxide anion radical scavenging capacity of all MBG-rice extrudates. Ferrous ion chelating ability of 15% MBG-rice extrudates is the strongest. From above, it was concluded that MBG-rice extrudates is a potential antioxidant food.更多还原