Neuro-fuzzy modeling to predict physicochemical and microbiological parameters of partially dried cherry tomato during storage: effects on water activity, temperature and storage time

Modelling of ultrasonic assisted osmotic dehydration of cape gooseberry using adaptive neuro-fuzzy inference system (ANFIS)

In the present investigation, the cape gooseberry (Physalis peruviana L.) was preserved by the application of osmotic dehydration (sugar solution) with ultrasonication. The experiments were planned based on central composite circumscribed design with four independent variables and four dependent variables, which yielded 30 experimental runs. The four independent variables used were ultrasonication power (X_P) with a range of 100-500 W, immersion time (X_T) in the range of 30-55 min, solvent concentration (X_C) of 45-65 % and solid to solvent ratio (X_S) with range 1:6-1:14 w/w. The effect of these process parameters on the responses weight loss (Y_W), solid gain (Y_S), change in color (Y_C) and water activity (Y_A) of ultrasound assisted osmotic dehydration (UOD) cape gooseberry was studied by using response surface methodology (RSM) and adaptive neuro-fuzzy inference system (ANFIS). The second order polynomial equation successfully modeled the data with an average coefficient of determination (R²) was found to be 0.964 for RSM. While for the ANFIS modeling, Gaussian type membership function (MF) and linear type MF was used for the input and output, respectively. The ANFIS model formed after 500 epochs and trained by hybrid model was found to have average R² value of 0.998. On comparing the R² value the ANFIS model found to be superior over RSM in predicting the responses of the UOD cape gooseberry process. So, the ANFIS was integrated with a genetic algorithm (GA) for optimization with the aim of maximum Y_W and minimum Y_S, Y_C and Y_A. Depending on the higher fitness value of 3.4, the integrated ANFIS-GA picked the ideal combination of independent variables and was found to be X_P of 282.434 W, X_T of 50.280 min, X_C of 55.836 % and X_S of 9.250 w/w. The predicted and experimental values of response at optimum condition predicted by integrated ANN-GA were in close agreement, which was evident by the relative deviation less than 7%.

Effects of an alternative bleaching method for yerba mate leaves on total phenolic content: evaluation and prediction with soft-computing approaches

BACKGROUND

The effects of an alternative bleaching method on the total phenolic content (TPC) and antioxidant activity (AA) of yerba mate extracts were evaluated. Traditional bleaching ('zapeco') was compared with submerging the leaves in water followed by a hot air oven-drying process. Seven different approaches, i.e. linear model (LM), response surface model (RSM), Mamdani, Larsen, adaptive neuro-fuzzy inference system (ANFIS) with the product (Prod) and the minimum (Min) operators, and ANFIS with automatically membership functions (Auto), were employed to compare the TPC of yerba mate extracts based on drying temperature and AA assays.

RESULTS

The results showed that if leaves were bleached followed by drying at higher temperatures, we obtained higher AA and TPC values. For submerging bleaching treatment, RSM model delivered the best accuracy measures with a mean absolute error (MAE), average absolute percentage error (MAPE), and mean squared error (MSE) of 0.128, 0.006, and 0.028, respectively. The ANFIS Auto model was the best for traditional bleaching treatment, with MAE, MAPE, and MSE of 0.490, 0.013, and 0.612, respectively.

CONCLUSION

The results suggest a second-order linear relation between drying temperature, AA assays to TPC, and a high level of relation complexity of drying temperature, AA assays, and TPC. The evaluated soft-computing approaches have the excellent ability to estimate TPC from bleached leaves. © 2022 Society of Chemical Industry.

Adaptive neuro-fuzzy interface system and neural network modeling for the drying kinetics of instant controlled pressure drop treated parboiled rice

Hot air drying kinetics of paddy grains during instant controlled pressure drop (ICPD) assisted parboiling process and its impact on the quality and micro-structural properties of milled rice were investigated. Among five mathematical models, Midilli model showed best fitted outcomes for prediction of adequate drying behavior. For the mapping of moisture ratio (MR) as a function of treatment pressure (TP), decompressed state duration (DD) and drying time (DT), artificial neural network (ANN) and adaptive neuro-fuzzy interface system (ANFIS) were applied. ANFIS model (5-5-5) with Gaussian membership function demonstrated best performance when contrasted with 3-5-1 ANN architecture. Effective diffusivity of the drying process varied from 2.8 × 10^-09 to 7.0 × 10^-09 m²/s with the increase of TP and DD. In comparison of quality parameters with the variation of TP and DD, positive impacts on head rice yield (HRY), redness (a*) and yellowness (b*) values and negative consequences on cooking time (CT) and brightness (L*) value were observed. The outcomes additionally uncovered that parboiled rice obtained at 0.6 MPa TP, indicated best quality in terms of improved process performance, HRY, CT, color and micro-structural properties.