Response Surface Model Research Articles

Dual approach in textile wastewater treatment: optimisation and comparison of fluidised-bed and conventional Fenton processes

ABSTRACT This study optimises secondary textile dyeing wastewater removal using fluidised-bed Fenton (FBR-Fenton) and standard Fenton processes for chemical oxygen demand (COD) and colour removal efficiency. The Box–Behnken response surface approach optimised pH, iron ion concentration, and hydrogen peroxide concentration. At pH 3.3, [Fe2+] = 2.6 mm, [H2O2] = 6.4 mm in the fluidised bed reactor with 67 g/L quartz sand, with a 50% bed expansion after 40 min, the FBR-Fenton process removed 75.8% COD and 98.0% colour. The standard Fenton method was tuned to pH 3.2, [Fe2+] = 3.1 mm, [H2O2] = 6.6 mm, and COD and colour removal effectiveness of 69.7% and 97.5%, respectively. FBR-Fenton enhanced COD and total iron removal efficiency, reduced chemicals, and reduced sludge by 28.0% compared to conventional Fenton. This study also showed quartz sand reusability in FBR-Fenton treatment cycles. Experimental results matched response surface model predictions, proving regression equation correctness.

Heat transfer characteristic and optimization design of alter diameter array impingement cooling structure

The fine design of blade cooling is one of the leading directions in the development for blade manufacturing technology, attaining optimal and consistent cooling efficiency is the common research concern in the field of fine design. In the current work, the alter diameter hole arrays are applied to reasonably regulate the distribution of cold air flow, which aims to achieve uniform cooling performance of impinging cooling technology. Experiment involving temperature sensitive paint technology and simulation based on SST k-w turbulence model are conducted. Subsequently, the mechanism of cooling performance increasement is investigated; the impact of structural elements on the heat transfer properties is discussed. Moreover, the correlation between the Nu number of the target surface and the comprehensive influence factors was fitted based on response surface method; the optimal structure in the range of the structural parameters is calculated based on NSGA-II and TOPSIS method. The results reveal that the influence of d and x on average surface Nusselt number is remarkable. In addition, the response surface model fitting model is accurate to predict the structure parameter effect on cooling performance of alter impingement cooling, with the prediction error less than 2.67 %. Compared with the basic EDI, the Nu of the ADI structure with ∇T constraint is increased by 5.2 %, the temperature inhomogeneity is reduced by 53.4 %. The significant improvement of temperature inhomogeneity indicates the practicality of ADI structure and optimized procedure used in improving impingement cooling efficiency which shows promising performance increasement.

Construction of controlled hyper-crosslinked nanofibrous tubes for Cr(VI) removal: Response surface, kinetics, and isotherm

Porous organic polymers (POPs) exhibit significant potential for adsorbing toxic metal ions in wastewater. Developing POPs with controlled morphologies is a pivotal direction in this field. This study synthesized a series of novel hyper-crosslinked nanofibrous tubes designated HCNT-Cn (n = 4, 8, 12, 16) via Friedel-Crafts alkylation and quaternization reactions. These reactions were fine-tuned through a post-synthetic strategy involving varying alkyl chain lengths. These materials were characterized using FT-IR, SEM, N₂ adsorption-desorption isotherms, among others, and they were specifically evaluated for their ability to adsorb Cr(VI). Among the variants, HCNT-C₄ exhibited the highest specific surface area (495.26 m2 g−1), superior hydrophilicity (CA = 48.7°), and optimal adsorption performance. The adsorption kinetics of HCNT-C₄ conformed to a pseudo-second-order model, while its adsorption isotherm aligned with the Langmuir model. An investigation into the impact of Cr(VI) removal was conducted using three independent variables in a Central Composite Design (CCD) response surface model, revealing that under optimal conditions, the Cr(VI) removal efficiency reached 98%. Additionally, a mechanism for Cr(VI) adsorption on HCNT-C₄ was proposed. It was also found that HCNT-C₄ could be reused up to four times, maintaining a removal efficiency of 70%. This study suggests potential applications for removing Cr(VI) from contaminated wastewater.

Applications of mixture methods in epidemiological studies investigating the health impact of persistent organic pollutants exposures: a scoping review.

Persistent organic pollutants (POPs) are environmental chemicals characterized by long half-lives in nature and human bodies, posing significant health risks. The concept of the exposome, encompassing all lifetime environmental exposures, underscores the importance of studying POP as mixtures rather than in isolation. The increasing body of evidence on the health impacts of POP mixtures necessitates the proper application of statisticalmethods. We aimed to summarize studies on the overall effects of POP mixtures, identify patterns in applications of mixture methods-statistical methods for investigating the association of mixtures-and highlight current challenges in synthesizing epidemiologic evidence of POP mixtures on health effects as illustrated through a case study. We conducted a systematic literature search on PubMed and Embase for epidemiological studies published between January 2011 and April 2023. We included 240 studies that met our eligibility criteria. 126 studies focused on per- and polyfluoroalkyl substances (PFAS) mixtures only, while 40 analyzed three or more classes of POPs in mixture analyses. We identified 23 unique mixture methods used to estimate the overall effects of POP mixtures, with Bayesian Kernel Machine Regression (BKMR), a type of response-surface modeling, being the most common. Additionally, 22.9% of studies used a combination of methods, including response-surface modeling, index modeling, dimension reduction, and latent variable models. The most extensively explored health outcome category was body weight and birth sizes (n = 43), and neurological outcomes (n = 41). In the case study of PFAS mixtures and birth weight, 12 studies showed negative associations, while 4 showed null results, and 2 showed positive associations. This scoping review consolidates the existing literature on the overall effects of POP mixtures using statistical methods. By providing a comprehensive overview, our study illuminates the present landscape of knowledge in this field and underscores the methodological hurdles prevalent in epidemiological studies focused on POP mixtures. Through this analysis, we aim to steer future research directions, fostering a more nuanced comprehension of the intricate dynamics involved in assessing the health effects of POP mixtures. Our work stands as a significant contribution to the ongoing exploration of the chemicalexposome.

Experimental study of toxic effects of cadmium against background of physical exercise

Introduction. Nowadays, working conditions in industry are often associated with chemical exposure along with manual labour. Given the evidence of the combined effect of heavy metals and muscular work, an increase in cadmium toxicity can be expected. However, an experimental study and actual assessment of exposure to both factors are necessary to avoid underestimation or overestimation of real risks. The purpose of our study was to establish health effects of subchronic cadmium toxicity against background of physical exercise. Materials and methods. In a subchronic animal experiment, four groups of outbred male rats, 14 animals each, were exposed to cadmium chloride at a single dose of 0.77 mg/kg body weight, forced physical exercise (treadmill running), both factors combined, or acted as a control group. Hematological, biochemical, and morphometric indices in rodents were assessed at the end of the experiment. The statistical significance of differences between the groups was established using Student’s t-test (p<0.05). For mathematical modelling, we used the Response Surface Methodology (RSM) with the construction of Loewe isoboles. Results. We observed significant unfavorauble shifts in 27.5% of the assessed morphometric, hematological, biochemical, cytological, and histomorphometric indices in rats following the exposure to cadmium only and in combination with exercise. The mathematical analysis based on the response surface modelling showed the typology of the combined impact of chemical and physical factors to vary depending on the specific effect by which it is assessed, as well as on the dose and level of effect. Limitations. Laboratory animals of only one biological species and sex were used for the experiment, exposed to a single concentration of cadmium and physical exercise of only one intensity level. Conclusion. The principles of health effects of cadmium and physical stress as risk factors of the workplace environment specified in the in vivo experiment may serve as a basis for scientific justification of risk assessment criteria to avoid under- or overestimation of real risks in the occupational setting.

Prediction of Microstructure Evolution in Ball Mill Liner Forging Process

The liner is affixed to the inner side of the ball mill cylinder to protect the cylinder. Through isothermal compression experiments, Arrhenius constitutive models, peak strain models, critical strain models, dynamic recrystallization dynamic models, and grain size models suitable for the forging process of Mn–Cr–Ni–Mo steel used in ball mill liners were established. By utilizing Deform software, a 3D thermo‐force‐structure coupling model for the hot forging process of ball mill liners was constructed, and the volume fraction of dynamic recrystallization and average grain size during forging was predicted. The response surface model was employed to investigate how process parameters interacted with each other and affected microstructure uniformity in ball mill liners. After optimization, the optimal parameters were determined: initial forging temperature at 1200 °C, forging speed at 30 mm s−1, and friction coefficient at 0.3. Subsequently, a hot forging experiment on ball mill liners was conducted using these optimized parameters; samples were analyzed through backscattered electron diffraction device experiments and microscopic tissue observations. Results demonstrated that microstructural changes observed during actual forging processes aligned with numerical simulation results—thus verifying both the accuracy of the Mn–Cr–Ni–Mo steel material model and numerical simulation method.

Environmental factors affecting accelerated carbonation of recycled concrete aggregates using response surface methodology

To further enhance CO2 capture and sequestration in concrete waste, this study investigates the interaction of four environmental factors, i.e., temperature, relative humidity (RH), CO2 concentration and pressure on the accelerated carbonation of recycled concrete aggregates. 29 groups of accelerated carbonation tests are performed based on the Box-Behnken design. The effects of four factors on the carbonation percentage of recycled concrete aggregates (RCAs) are analyzed using a response surface methodology. A response surface model is obtained to characterize the experimental results. Additionally, three single-factor scenarios and three multi-factor coupled scenarios are employed to show the mechanisms affecting RCA carbonation and carbonation environments. Single factor analysis indicate that RH and CO2 concentration can significantly affect the carbonation of the RCA and the significance of the interactions follows the order of temperature - RH, RH - CO2 concentration and temperature - CO2 concentration. The microstructure of RCAs during carbonation is further studied and presented.

Optimization and modeling of process parameters for nutrient recovery from sewage wastewater

ABSTRACT Nitrogen (N) and phosphorus (P) contamination in wastewater pose significant environmental challenges. Recovering these elements as struvite not only mitigates environmental pollution but also aligns with sustainable development goals by recycling valuable resources. This research hypothesizes that optimized recovery methods can enhance the efficiency and effectiveness of struvite crystallization, addressing existing challenges in conventional techniques. To achieve optimal removal and recovery of N and P from sewage, a response surface model was employed. This model allowed for the identification of optimal process conditions and the elucidation of interactions among various components. Key variables impacting struvite recovery were identified using the Plackett–Burman design, while the central composite design was used for further optimization. The study determined the optimized parameters for phosphate recovery to be an Mg:P ratio of 1:2, pH of 10.5, additive concentration of 350 ppm, and a precipitation time of 30 min. Thermogravimetric analysis indicated that the residual amounts were below 50%. Additionally, the size and surface morphology of the final product were influenced by the process parameters, particularly the Mg:P ratio and pH. An inexpensive, quick, and efficient method to recover struvite fertilizer with a minimum demand of time and chemicals is established toward SDG 2 and 6.

Evaluation of orthogonal composite designs for second‐order model in presence of missing observation

AbstractOrthogonal‐array composite designs (OACDs) and orthogonal‐uniform composite designs (OUCDs) are orthogonal composite designs that combine two‐level full or fractional factorial and three‐level orthogonal‐array/uniform designs for estimation of the linear, bilinear, and quadratic effects in a second‐order response surface model. In this study, the effects of missing one observation in the various design portions (factorial (f) axial (a) and center (c)), on the precision of parameter estimates, prediction variance and design efficiency of OACDs and OUCDs for 5 ≤ k ≤ 9 factors at different values of α (the distance of a non‐zero co‐ordinate in an additional design point from the center) are evaluated. The results showed that missing a factorial and an axial point have adverse effect on the precision of parameter estimates of OACDs and OUCDs, while missing a center point has little effect. Missing an axial point caused the highest effect on the prediction variance and design efficiencies. The FDS plots showed OACDs to be better designs for k ≤ 7 and OUCDs for k = 8 and 9 factors.

Silver nanoparticles modified with MoS2 and β-cyclodextrin as SERS substrate for rapid determination of cysteamine hydrochloride in meat products

An efficient Surface-enhanced Raman scattering (SERS) method for the detection of cysteamine hydrochloride (CSH) was developed by synthesizing a composite substrate comprising silver nanoparticles (AgNPs) functionalized with MoS2 and β-cyclodextrin (β-CD). The enhanced Raman signals of CSH by β-CD/MoS2/AgNPs substrate were the contribution of electromagnetic enhancement (EM) as well as chemical enhancement (CM), and the enhancement factor (EF) can reach up to 3.11 × 106 (peak at 633 cm−1). Various instrumental techniques were used to characterize the substrate, such as X-ray diffraction (XRD), thermogravimetric analysis (TGA), transmission electron microscopy (TEM), high-angle annular dark field scanning transmission electron microscopy (HAADF-STEM) and ultraviolet visible (UV–vis). The binding of β-CD/MoS2/AgNPs and CSH was confirmed by UV–vis and Fourier transform infrared (FT-IR). The optimal experimental conditions were determined by single factor experiments as well as response surface model. The influences of different metal ions and analogous drugs on the detection of CSH were investigated. Under optimum conditions, a good linear correlation (R = 0.9997) was established for CSH in the range of 10.00–1000.00 nmol/L, and the limit of detection (LOD) was as low as 0.78 nmol/L (S/N = 3). The contents of CSH in meat samples were detected. The recovery was 96.6–103.1 %, and the relative standard deviation (RSD) of the measurement was 0.7–3.9 % (n = 7).

Optimization Analysis of Response Surface for Ultrasonic Welding of Battery Tabs

Abstract The response surface methodology was used to optimize the ultrasonic welding process parameters of battery tabs. The three process parameters of welding pressure, welding energy and welding amplitude were selected as the influencing factors, the peel strength and welding time were selected as the response variables, a three-factor and three-level experimental matrix was designed. The experimental results were analyzed by multiple regression fitting and the regression model was established. The optimal parameters derived from the model were verified. The optimum combination of process parameters is pressure (0.25Mpa), energy (649J) and amplitude (75%) with the goal of maximum peel strength and minimum welding time. The test results show that the relative error is within a reasonable range, and the established response surface model has a higher precision, which provides a certain reference for reasonable selection of ultrasonic welding process parameters for battery tabs.

PROCESSING PARAMETERS OPTIMIZATION OF VACUUM-BAGGING PREFORMING IN OVEN CURE FOR INTER-LAMINAR SHEAR STRENGTH (ILSS) IN GLASS/EPOXY COMPOSITE LAMINATE

Autoclave had been restricted to immense production expenses and overabundant residual stress. These disadvantages had prompted development on the alternative of out-of-autoclave (OoA) processing. The optimization on vacuum-bagging-only (VBO) pre-forming process in producing high inter-laminar shear strength (ILSS) of composite laminate was proposed. The effects of individual and combined pre-forming parameters of VBO-oven cure processing for the conventional low-cost glass/epoxy composite material towards ILSS of cured laminates were quantified. 20 composite panels were manufactured following the designated parameter combinations based on central composite design in fractional factorial for response surface model. Three factors of vacuum debulk duration, edge breather number of sides and intensifier weight were investigated. For validation, two laminates without additional processing parameters were produced via oven (baseline) and autoclave. ILSS test was conducted based on ASTM D 2344. The interaction between combined parameters was analyzed using analysis of variance (ANOVA). Lowest ILSS was found in laminates where intensifier was absent, while highest ILSS was measured with edge breather, debulk and intensifier at different levels. An optimum combination of 30 minutes debulk, each sides of edge breather and 1kg intensifier were validated to produce laminate with highest ILSS of 38.96 MPa, which was 23.53% higher than baseline laminate.

The Application of Converter Sludge and Slag to Produce Ecological Cement Mortars.

The paper presents an analysis of the effective use of a mixture of steel sludge (S1) and slag (S2) from the converter process of steel production for the production of cement mortars. Metallurgical waste used in the research, which is currently deposited in waste landfills and heaps near plants, posing a threat to groundwater (possibility of leaching metal ions present in the waste), was used as a substitute for natural sand in the range of 0-20% by weight of cement (each). The obtained test results and their numerical analysis made it possible to determine the conditions for replacing part of the sand in cement mortars with a mixture of sludge and slag from a basic oxygen furnace (BOF) and to determine the effects of such modification. For the numerical analysis, a full quadratic Response Surface Model (RSM) was utilized for two controlled factors. This model was subsequently optimized through backward stepwise regression, ensuring the inclusion of only statistically significant components and verifying the consistency of residual distribution with the normal distribution (tested via Ryan-Joiner's test, p > 0.1). The designated material models are helpful in designing ecological cement mortars using difficult-to-recycle waste (i.e., sludge and converter slag), which is important for a circular economy. Mortars modified with a mixture of metallurgical waste (up to 20% each) are characterized by a slightly lower consistency, compressive and flexural strength, and water absorption. However, they show a lower decrease in mechanical strength after the freezing-thawing process (frost resistance) compared to control mortars. Mortars modified with metallurgical waste do not have a negative impact on the environment in terms of leaching heavy metal ions. The use of a mixture of sludge and steel slag in the amount of 40% (slag/sludge in a 20/20 ratio) allows you to save 200 kg of sand when producing 1 m3 of cement mortar (cost reduction by approx. EUR 5.1/Mg) and will also reduce the costs of the environmental fee for depositing waste.

Parameter optimisation of marine four-stroke engine EGR systems using RSM and NSGA-III

ABSTRACT To reduce NOx emission of marine engines, an engine parameter optimisation method combining response surface method (RSM), Bayesian neural network and non-dominated sequence genetic algorithm (NSGA-III) is proposed. First, a simulation model of a marine four-stroke engine with an external exhaust gas recirculation (EGR) system was established and validated in AVL-BOOST software. Then, the five parameters of EGR valve opening, ignition advance angle, compression ratio (CR), intake valve opening (IVO) and exhaust valve closing (EVC) are selected as the decision parameters, and the three parameters of engine power, brake specific fuel consumption (BSFC) and NOx emission are taken as the target parameters to be optimised. In Design-Expert software, the response surface model is established by using the experimental design approach of central composite design (CCD), and the impact of decision parameters on target parameters are discussed by using response surface analysis. NSGA-III was used for the multi-objective optimisation of parameters, and the accuracy of NSGA-III was improved by the Bayesian neural network. The optimal solution is selected and verified by the technique for order of preference by similarity to the ideal solution (TOPSIS) approach. The results show that when the EGR valve opening is 47.5%, ignition advance angle is −18.99°CA, CR is 14.995, IVO is 15.27°CA and EVC is 7.68°CA, compared with the standard setting, the optimised power is increased by 4.17%, BSFC is reduced by 4.27%, and NOx emissions are reduced by 12.37%. Therefore, the combination of the Bayesian neural network and NSGA-III multi-objective optimisation can improve engine performance while reducing fuel consumption and NOx emissions.

AutoRL-Sim: Automated Reinforcement Learning Simulator for Combinatorial Optimization Problems

Reinforcement learning is a crucial area of machine learning, with a wide range of applications. To conduct experiments in this research field, it is necessary to define the algorithms and parameters to be applied. However, this task can be complex because of the variety of possible configurations. In this sense, the adoption of AutoRL systems can automate the selection of these configurations, simplifying the experimental process. In this context, this work aims to propose a simulation environment for combinatorial optimization problems using AutoRL. The AutoRL-Sim includes several experimentation modules that cover studies on the symmetric traveling salesman problem, the asymmetric traveling salesman problem, and the sequential ordering problem. Furthermore, parameter optimization is performed using response surface models. The AutoRL-Sim simulator allows users to conduct experiments in a more practical way, without the need to worry about implementation. Additionally, they have the ability to analyze post-experiment data or save them for future analysis.

A methodology to enhance ye'elimite in calcium sulfo-aluminate ferrite clinker from bauxite residue

Calcium sulfoaluminate-ferrite (CSA-F) clinkers were prepared using the alumina production waste bauxite residue (BR) as one of the raw materials, next to limestone, kaolin, and gypsum. A complementary approach was employed, aiming to maximize the desired ye’elimite phase. This consisted of thermodynamic modelling and computational statistics via response surface modelling, which enabled the identification of the best raw meal by varying the proportions of the chosen raw materials and the optimal clinkering conditions by unraveling the interaction between clinkering temperature and time. The optimal parameters were for a raw meal consisting of 38 wt% BR, 42 wt% limestone, 5 wt% kaolin, and 15 wt% of gypsum, fired for 60 min at 1300 °C. The accuracy of the model was assessed experimentally on a clinker synthesized at a temperature range of 1250–1300 °C for 30 and 60 minutes, followed by rapid cooling. Based on the characterization results, firing at 1300 °C for 60 minutes was the optimal set of conditions to maximize ye'elimite formation, which was in accordance with the response surface model. In addition, clinker 1300_60 exhibited the highest 28d compressive strength of 33 MPa due to the presence of Ternesite phase. In conclusion, the combined methodology endorsed and presented herein is a pragmatic and relatively accurate methodology for producing low-CO2 clinkers from bauxite residue and can likely be transferred to other unexploited secondary raw materials.

Prediction and optimization of combustion performance and emissions for lean methane-hydrogen non-premixed flame using RSM-PSO methodology

To reduce carbon emissions, hydrogen is a clean energy as fuel to co-combustion with natural gas. This experimental work focuses on studying the effects of hydrogen addition to methane on combustion performance and emissions in non-premixed coaxial burners, establishing performance prediction models and optimizing operation conditions involving fuel velocity, hydrogen blending ratio and equivalence ratio. The global sensitivity and analysis of variance were adopted to quantify the contributions of independent variables on the output responses of temperature, O2, NO, CO and CO2 emissions. The hydrogen blending ratio is the most influential parameter, whose total effects of NO and CO2 emissions are 64.6% and 83.5%, respectively. The equivalence ratio is not significant to NO and CO2 emissions, and the adjusted determination coefficient of the reduced prediction model is 0.9707 and 0.9360, demonstrating a better goodness of fit. Based on the response surface model and particle swarm optimization algorithm, a Pareto was solved showing NO and CO2 emissions decrease by 42.0%–73.6% and 60.8%–89.5%, and peak temperature in the chamber centerline increases by 1.02%–3.61%. The research results provide a profound guide to predict and optimize combustion performance and emissions for methane-hydrogen non-premixed flame.

Optimization of the extraction process of total steroids from Phillinus gilvus (Schwein.) Pat. by artificial neural network (ANN)-response surface methodology and identification of extract constituents

Phillinus gilvus (Schwein.) Pat has pharmacological effects such as tonifying the spleen, dispelling dampness, and strengthening the stomach, in which sterol is one of the main compounds of P. gilvus, but there has not been thought you to its extraction and detailed identification of its composition, in the present study, we used artificial neural network (ANN) and response surface methodology (RSM) to optimize the conditions of ultrasonic-assisted extraction, and the parameters of the independent and interaction effects were evaluated. Ultra performance liquid chromatography-quadrupole-time of flight mass spectrometry (UPLC-Q-TOF-MS/MS) was used to identify the major components in the purified extract. The results showed that the optimal extraction process conditions were: ultrasonic time 96 min, ultrasonic power 140 W, liquid to material ratio 1:25 g/ml, and ultrasonic temperature 30.7 °C. The compliance rates of the predicted and experimental values for the artificial neural network model and the response surface model were 98.3% and 96.12%, respectively, indicating that both models have the potential to be used for optimizing the extraction process of P. gilvus in industry. A total of 120 compounds and 30 major steroids were identified by comparison with the reference compounds. Among the major steroidal components are these findings will contribute to the isolation and utilization of active ingredients in P. gilvus.

Process Optimization for Antarctic Krill (Euphausia superba) Sauce Based on Back Propagation Neural Network Combined with Genetic Algorithm

Using Antarctic krill (Euphausia superba) as the research object, we optimized the process conditions for Antarctic krill sauce (AkS) by including three factors (salt addition, rock sugar addition, and the oil-to-material ratio) and sensory evaluation as response values. The data from the response surface were fed into the back propagation (BP) neural network training, generating a model mapping the process conditions and sensory scores, which were subsequently combined with the genetic algorithm (GA) for global optimization to determine the optimal process for AkS preparation. The results revealed that the response surface model was well suited to the BP neural network training and prediction sets, with correlation values of 0.98 and 0.95, respectively. The fitting prediction effect was obvious for the sensory scoring results of the product. The parameters obtained from the GA’s global optimization search accord with the analytical results of the response surface. The findings demonstrated that combining a BP neural network with a GA can enhance the AkS preparation technique. Under optimal processing conditions, AkS has a high sensory score and protein and carbohydrate contents, moderate fat content, minimal fat oxidation, and non-detectable pathogens, indicating that the AkS in this study was nutritious and safe to consume.

Thermal Management and Optimization of Automotive Film Capacitors based on Parallel Microchannel Cooling Plates

Abstract The automotive film capacitors (AFCs) stand as a widely employed components in electric vehicles. Yet, a notable concern arises with the potential for excessive ripple current, which can prompt self-heating in the AFC and diminish its reliability. Therefore, it becomes crucial to conduct thermal management and effective heat dissipation design for the AFC to ensure its optimal performance. In this study, considering the trend toward integrated and lightweight motor controllers, a parallel microchannel cooling plate (PMCP) is designed at the bottom of the AFC. Through optimization, the thermal performance of the AFC and the overall cooling performance of the PMCP are enhanced. The AFC thermal model is established and the calculation method for equivalent thermal properties of the film capacitor core is described. A conjugate heat transfer simulation model for the AFC and the PMCP is created by FLUENT and validated through two experimental tests. Additionally, based on an optimal Latin Hypercube Sample size, the accuracy of five fitting models is compared and the non-dominated sorting genetic algorithm II (NSGA-II) for optimization is employed. The results indicate that the error between the simulation method and the two experiments is within 5 %. The application of the PMCP effectively redistributes the hottest region of the AFC to the outer housing, reducing the maximum AFC temperature by 10.90 °C. Among the five fitting models, the RSM (Response Surface Model) proved to be the most accurate. The optimized PMCP enhances the overall cooling performance by 10.32 %, and increases the maximum withstand ripple current of the AFC by 43.83 %.