Parametric optimisation through the use of Box-Behnken design in the Co-gasification of oil palm trunk and frond for syngas production

An experimental and response-surface-based optimization approach towards production of producer gas in a circulating fluidized bed gasifier using blends of renewable fibre-based biomass mixtures

“…Stochastic parameters such as the F -value, correlation coefficient ( R 2 ), and adjusted R 2 for all three responses are represented in Table 6. All the values were found to be satisfactory and were found to be in-line with statistical recommendations and the literature 55 Statistical indicators such as the F -value and P -value assist in assessing the model adequacy and the relevance of each contributing variable; a lower P -value advocates the variable significance. 56 An efficient and improved model for space navigation implies that model sufficiency is supported by correlation coefficient ( R 2 ) values greater than 0.8 and close to 1.0.…”

Ethylene Glycol Dry Reforming over La‐Co/Al2O3: Optimization and Regeneration Studies

“…3 Results and Discussion Additionally, it is advantageous that the model's lack of fit is not substantial. This indicates that the RSM model adequately fits the experimental data [26,27].…”

A Critical Review of Data Science Applications in Resource Recovery and Carbon Capture from Organic Waste

“…MLR and PLSR are both linear regression models, where PLSR models are able to resolve the colinearity among the input variables and provide unbiased input-output relationships . On the contrary, MPR is nonlinear in nature and typically represents a second- or third-order polynomial regression model in the RRCC literature. − The outputs of MPR are further utilized for process optimization through response surface methodology. Among ML models, ANNs are black-box models that mimic biological neurons by attempting to replicate information transfer between biological neurons through electrical signals using mathematical functions. , ANN models have been highly successful in combining complex nonlinear input data to predict outputs .…”

“…Of the reviewed studies that applied data-driven methods to model RRCC technologies, 20% (36% statistical and 64% ML methods) represented models that predicted syngas yield through the gasification of various feedstocks (Figure a and Table S9). Gasification was most frequently modeled by applying MPR ,− using primary data and ANN ,− using both primary and secondary data. These MPR and ANN models respectively, comprised 33% and 36% of the data-driven gasification models (Figure a and Table S9).…”