Modelling of A Fluidized Bed Drier Using Artificial Neural Network

ABSTRACT

Proper modelling of a fluidized bed drier (FBD) is important to design model based control strategies. A FBD is a non-linear multivariable system with non-minimum phase characteristics. Due to the complexities in FBD conventional modelling techniques are cumbersome. Artificial neural network (ANN) with its inherent ability to “learn” and “absorb” non-linearities, presents itself as a convenient tool for modelling such systems.

In this work, an ANN model for continuous drying FBD is presented. A three layer fully connected feedfordward network with three inputs and two outputs is used. Backpropagation learning algorithm is employed to train the network. The training data is obtained from computer simulation of a FBD model from published literature. The trained network is evaluated using randomly generated data as input and observed to predict the behaviour of FBD adequately.     

Effects of Degumming and Bleaching on 3-MCPD Esters Formation During Physical Refining

Crude palm oil (CPO) was physically refined in a 200-kg batch pilot refining plant. A study of the possible role of degumming and bleaching steps in the refining process for a possible critical role in the formation of 3-chloropropane-1,2-diol (3-MCPD) esters was evaluated. For the degumming step, different percentages of phosphoric acid (0.02–0.1%) as well as water degumming (2.0%) were carried out. Six different types of bleaching clays, mainly natural and acid activated clays were used for bleaching process at a fixed dose of 1.0%. Deodorization of the bleached oils was performed at 260 °C for 90 min. Analyses showed that 3-MCPD esters were not detected in the CPO. Phosphoric acid degumming (0.1%) in combination with acid activated clays produced the highest levels (3.89 ppm) of 3-MCPD esters in the refined (RBD) oil. The esters were at the lowest levels (0.25 ppm) when the oil was water degummed and bleached with natural bleaching clays. However, the refined oil qualities were slightly compromised. Good correlation of 0.9759 and 0.9351 was obtained when concentration of the esters was plotted against acidity of the bleaching earths for the respective acid and water degumming processes. The findings revealed the contribution of acidic conditions on the higher formation of 3-MCPD esters. In order to lower the esters formation, it is important to reduce acid dosage based on the crude oil qualities or to find alternatives to acid degumming process. Neutralization of the acidity prior to deodorization was effective in reducing the formation of 3-MCPD esters.     

Estimating Oil and Protein Content of Sesame Seeds Using Image Processing and Artificial Neural Network

Image processing has many applications in different fields of agriculture. The present study aimed to use image processing techniques and artificial neural networks (ANN) to estimate oil and protein contents of sesame genotypes without the use of time-consuming and costly laboratory methods. The proposed method accurately estimates the parameters in sesame seeds without destructing the genetically valuable material. In this study, a set of 138 morphological features were extracted from the digital image of 125 sesame seed genotypes. A multilayer perceptron (MLP) ANN was then employed to estimate oil and protein contents and determine the relationship between estimated values and laboratory-measured values. The efficiency of this model was compared to radial bases function (RBF), extended RBF (ERBF), GRNN, M5-Rule, M5-Tree, support vector machine regression, and linear regression models. Results showed that MLP performed better in estimating qualitative parameters of seeds in the sesame germplasm. The model estimated oil content with an root mean square error (RMSE) of 2.13% (the accuracy of 97.87%) and an R² of 0.93. Protein content was estimated by an RMSE of 0.378% (the accuracy of 99.62%) and an R² of 0.96.     

山桐子油脱胶工艺研究

文章比较了不同脱胶剂对山桐子油脱胶效果的影响,重点研究了磷酸法脱胶工艺,研究结果表明:适宜的脱胶工艺为反应温度80℃,搅拌15 min,脱酸剂用量为1.2%,本试验中搅拌速度对脱胶效果没有影响;山桐子油中以非水化磷脂为主,磷脂酸与溶血磷脂酸钙镁盐在毛油中所占比例相当。     

Simulation of Crude Palm Oil Dilution in a Palm Oil Mill Using Computational Fluid Dynamics

Computational fluid dynamics (CFD) simulation was applied to simulate the dilution of undiluted crude oil (UCO) in a dilution tank of a palm oil mill. Fluid flow and mixing characteristics were examined. Considering the mixing behavior, the mixing of dilution water and UCO occurred as soon as these fluids entered the dilution tank, and the oil mass fraction in the mixture decreased gradually towards the outlet of the tank. Meanwhile, the velocity of dilution water and UCO declined as the fluids moved from their respective inlets. The intensity of turbulence flow remained until near the tank outlet. For the parametric study, the oil mass fraction of diluted crude oil (DCO) increased with higher UCO flow rate and oil mass fraction in the UCO but declined with higher dilution water flow rate.     

用人工神经网络方法预测鼓泡塔气含率

A new correlation for the prediction of gas hod up in bubble columns was proposed based on an extensive experimental database set up from the literature published over last 30 years .The updated estimation method relying on artificial neural network,dimensional analysis and phenomenological approaches was used and the model prediction agreed with the experimental data with average relative error less than 10%.     

用神经网络法预测鼓泡塔内的气含率

在 2 0 0 0多组公开发表的鼓泡塔实验数据组成的数据库基础上 ,提出了一个预测鼓泡塔内气含率的关联式 ,介绍了一种新的神经网络回归方法。将该方法与受力分析结合起来得到 4个对鼓泡塔内气含率影响较大的无因次准数 ,对所得到的模型作出评价     

Insight into the Enzymatic Degumming Process of Soybean Oil

An enzymatic degumming trial of soybean oil was carried out at a capacity of 400 tons/day by applying microbial phospholipase A₁ from Thermomyces lanuginosus/Fusarium oxysporum. When the pH was kept in the range of 4.8–5.1, less than 10 mg/kg of phosphorous content of The oil was obtained. The gum and oil were easily separated after centrifugation and the oil loss was minimal under the process conditions. Through analysis of phospholipids compounds in the gum by Electrospray Ionization-Mass Spectrometer and phosphorous content, it could be seen that both glycerophospholipids and lysophospholipids existed with contents of 45.7 and 54.3%, respectively. The performance of enzymatic degumming was found to be related to the production of glycerophospholipids.     

用人工神经网络优化己酸乙酯的制备条件

己酸乙酯生产工艺中采用硫酸做催化剂,会产生大量的工艺废水,对环境不利.本文采用硫酸氢钠F型复合催化剂,以环己烷为带水剂催化合成己酸乙酯,并采用正交实验和人工神经网络对该工艺加以优化,建立三层改进的误差反向传播网络（BP-ANN神经网络）,把正交实验的水平范围扩大后输入建立的网络,得到最佳制备条件：催化剂用量为4 g,反应时间为3.4 h,醇酸物质的量比为1.2,带水剂用量为27 mL,在该条件下己酸乙酯的预测收率为98.60%,五次验证实验的平均收率为98.32%.     

Auxiliary Degumming and Pressurized High Temperature Water Washing to Mitigate the Formation of Monochloropropanediols in Palm Oil

This study reports the benefits of auxiliary degumming (Aux.D) and pressurized high temperature (165 °C) water washing (PHTW) to mitigate the formation of monochloropropanediols (MCPD) during labscale physical refining of palm oil. Water-based degumming in combination with bleaching and deodorization are performed as the selected physical refining process. The mitigation concepts Aux.D and PHTW are integrated into the refining protocol and the ultimately observed MCPD levels in the fully refined oils are determined. Aux.D is performed by extracting the hydratable gum from pressed sunflower oil and using it as a degumming agent to further purify palm oil that has been previously subjected to centrifugation and water degumming. This approach enables the mitigation of 3-MCPD from the water washed reference 2.4–0.9 mg kg⁻¹ in ampoules. Even stronger mitigation is obtained when Aux.D is combined with bleaching and executed twice allowing a mitigation from the reference 1.9–0.6 mg kg⁻¹, in ampoules. PHTW is shown to decrease the 3-MCPD content of the refined oil from the reference 2.4–1 mg kg⁻¹, in ampoules and when combined with bleaching and executed twice showing a decrease from the reference 1.9–0.9 mg kg⁻¹. Practical applications: The benefits of these mitigation concepts are confirmed both in sealed ampoule tests and in deodorizer experiments at the lab scale. A combined application of Aux.D or PHTW with physical refining may represent new insights that can help to potentially further mitigate the formation of MCPD in physically refined palm oil beyond the limits achievable with current refining practices.     

人工神经网络研究取代芳烃的定量构效关系

采用误差反传前向人工神经网络（artificial neural network,ANN）建立了24种取代芳烃的结构与其对发光菌的急性毒性之间的定量关系模型（ANN模型）。以24种取代芳烃的量子化学参数作为输入,急性毒性作为输出,采用内外双重验证的办法分析和检验所得模型的稳定性和外部预测能力。所构建网络模型的相关系数为0.9834、交叉检验相关系数为0.9780、标准偏差为0.11、残差绝对值≤0.33,应用于外部预测集,外部预测集相关系数为0.9955;而多元线性回归（multiple linear regression,MLR）法模型的相关系数为0.9786、标准偏差为0.12、残差绝对值≤0.36,外部预测集相关系数为0.9904。结果表明,ANN模型获得了比MLR模型更好的拟合效果。     

The Total Degumming Process – Theory and Industrial Application in Refining and Hydrogenation

The degumming of vegetable oils prior to physical refining is a crucial preliminary step. The degumming process is not only largely responsible for the quality of the final product, but it also determines the amount of bleaching earth to be used, which has a substantial effect on the yield improvement which can be attained by this route. Investigations show clearly that iron, as a pro-oxidant, strongly influences the stability of refined oils, and that oil, degummed before bleaching and physical refining, may contain a maximum of 0.2 ppm Fe, if it is to yield a stable product. The Total Degumming Process has been developed on the basis of these findings, to make it possible to degum oil to a residual Fe-level below 0.2 ppm and a residual phosphorus content below 10 ppm. The principles and industrial application of the process have been considered. The results of industrial production using different raw materials of various qualities have been used to make a comparison between the conventional refining process (neutralization – bleaching – deodorization) and the Total Degumming Process in combination with physical refining. The combination of the Total Degumming Process and a simplified caustic refining process, and the use of Totally Degummed Oil for hydrogenation have also been considered.     

Modeling Oil Content of Sesame (Sesamum indicum L.) Using Artificial Neural Network and Multiple Linear Regression Approaches

Sesame (Sesamum indicum L.) is an important ancient oilseed crop with high oil content (OC) and quality. The direct selection to improve OC of sesame (OCS) due to low heritability leads to a low profit. The OCS modeling and indirect selection through high‐heritable characters associated with OCS using advanced modeling techniques is a beneficial approach to overcome this limitation that allows breeder to get a better idea of the plant properties that should be monitored during breeding experiments. This study, carried out in 2013 and 2014, compared the potential of artificial neural network (ANN) and multilinear regression (MLR) to predict OCS in the Imamzadeh Jafar plain of Gachsaran, Iran. Principal component analysis (PCA) and stepwise regression (SWR) were used to evaluate 18 input variables. Based on PCA and SWR, the 6 traits of number of capsules per plant (NCP), number of days from flowering to maturity (NDFM), plant height (PH), thousand seed weight, capsule length, and seed yield were chosen as input variables. The network with the sigmoid axon transfer function and 2 hidden layers was selected as the final ANN model. Results showed that the ANN predicted the OCS with more accuracy and efficacy (R² = 0.861, root mean square error [RMSE] = 0.563, and mean absolute error [MAE] = 0.432) compared with the MLR model (R² = 0.672, RMSE = 0.742, and MAE = 0.552). These results showed the potential of the ANN as a promising tool to predict OCS with good performance. Based on sensitivity tests, NCP followed by NDFM and PH, respectively, were the most influential factors in predicting OCS in both models. It seems that a breeding program to select or create long sesame genotypes with a long period from flowering to maturity can be a good approach to address OCS in the future.     

利用神经网络动态模拟污水生物脱氮过程

用以活性污泥数学模型 (ASM)为基础的EFOR软件的模拟输入输出数据训练设计好的神经网络 ,然后同时变化输入数据 ,比较输出数据。结果表明神经网络用于污水生物脱氮过程的动态的模拟 ,预测出水COD的误差可控制在± 3%以下 ,出水TN的误差可控制在± 6 %以下 ,出水SS的误差可控制在± 10 %以下。     

Refining and Bleaching of Peanut Miscella

The effects of type of crude miscella, oil content in miscella, concentrations of caustic soda solutions, method of mixing and temperature on the refining of high f. f. a. dark coloured peanut miscellas were investigated. Very effective removal of free fatty acids and decolorisation of the peanut miscellas were achieved by treating at 45-60% oil content with 16°-20° Bé caustic soda solution at room temperature (ca. 32 ± 1° C). Good bleaching of refined miscella samples also at room temperature with commercial acid treated earth and active charcoal was possible. Refined oils had 0.02-0.06% f. f. a. with 94-99% colour removed. Successful commercial possibilities are indicated.     

基于遗传算法的人工神经网络在油水层识别中的应用

油水层识别是测井二次解释的重要内容。基于遗传算法的人工神经网络方法具有自学、自适应和抗干扰等特点，能有效地区分油水层。实际应用表明，其误判率远低于传统方法。     

基于自适应神经网络的芳烃异构化过程建模

针对芳烃异构化过程(AHIP)中影响对二甲苯(PX)产率的因素众多且复杂等特点,提出一种自适应神经网络(Adaptive ANN)以建立AHIP的各因素与PX产率的关联模型.Adaptive ANN将样本分成训练样本和校验样本,并设计过拟合判据参数.通过训练样本对网络进行训练,训练过程中以模型对校验样本的预测性能为指标,通过过拟合判据参数的计算自适应地在获得具有最佳预测性能模型时终止网络训练,克服了传统的神经网络以模型的拟合精度为指标,造成训练时间过长和过拟合等缺点.     

Optimization of the Degumming Process for Aqueous-Extracted Wild Almond Oil

Wild almond (Amygdalus scoparia) oil is rich in oleic acid and, considering both nutritional and stability points of view, it can be utilized for future food applications. In the current study, acid degumming was investigated based on a method by response surface methodology using four degumming parameters, namely the amount of phosphoric acid (0.0–0.2%, w/w), the amount of water (1.0–5.0%, w/w), degumming temperature (30–70 °C), and degumming time (10–50 min). Optimum conditions for the minimum phosphorus level in the oil were found to be 0.15% phosphoric acid, 3.0% water, 40 °C degumming temperature, and 28 min degumming time, resulting in an almost complete removal of phosphorus. The final degummed wild almond oil had less than 1 mg kg⁻¹ phosphorus (reduced from an original value of 206 mg kg⁻¹). The experimental value of phosphorus reduction at optimum conditions agreed well with that predicted by the model. Peroxide value, anisidine value, iron, copper, and lead contents, phytosterols, unsaponifiable matter, and color of the oil decreased significantly during the degumming process; however, the fatty acid composition did not change. Also, degumming did not significantly impact the free fatty acid level, refractive index, density, iodine value, and the saponification value of the oil. However, tocopherols and the oxidative stability of the oil increased during degumming. Crude wild almond oil contained a trace level of amygdalin, which was completely eliminated during the degumming process.     

基团贡献人工神经网络法估算有机物常压沸点

基于基团贡献应用人工神经网络法对有机物的常压沸点进行了估算 ,输入参数为有机物的基团数 ,输出为常压沸点 ,本方法具有很好的灵活性 ,并能考虑到基团的相互作用 ,189个样本估算的平均相对误差仅为 1.5 9%