This study integrates the Backpropagation (BP) Neural Network with several optimization algorithms, namely Hippopotamus Optimization (HO), Parrot Optimization (PO), Osprey Optimization Algorithm (OOA), and Goose Optim...
详细信息
This study integrates the Backpropagation (BP) Neural Network with several optimization algorithms, namely Hippopotamus Optimization (HO), Parrot Optimization (PO), Osprey Optimization Algorithm (OOA), and Goose Optimization (GO), to develop four predictive models for the adhesive strength of heat-treated wood: HO-BP, PO-BP, OOA-BP, and GO-BP. These models were used to predict the adhesive strength of the wood that was heat-treated under multiple variables such as treatment temperature, time, feed rate, cutting speed, and abrasive particle size. The efficacy of the BP neural network models was assessed utilizing the coefficient of determination (R2), error rate, and CEC test dataset. The outcomes demonstrate that, relative to the other algorithms, the Hippopotamus Optimization (HO) method shows better search efficacy and convergence velocity. Furthermore, XGBoost was used to statistically evaluate and rank input variables, revealing that cutting speed (m/s) and treatment time (hours) had the most significant impact on model predictions. Taken together, these four predictive models demonstrated effective applicability in assessing adhesive strength under various processing conditions in practical experiments. The MAE, RMSE, MAPE, and R2 values of the HO-BP model reached 0.0822, 0.1024, 1.1317, and 0.9358, respectively, demonstrating superior predictive accuracy compared to other models. These findings support industrial process optimization for enhanced wood utilization.
Advanced sensor technology has driven the remaining useful life (RUL) prediction of aircraft engines. However, only a few studies have considered incorporating RUL prediction results into maintenance plans. To address...
详细信息
Advanced sensor technology has driven the remaining useful life (RUL) prediction of aircraft engines. However, only a few studies have considered incorporating RUL prediction results into maintenance plans. To address this problem, this paper investigates a novel predictive maintenance framework for aircraft engines. First, a hybrid deep learning model is developed to predict the aircraft engine RUL. Based on the predicted RUL, two new mixed integer linear programming models are developed to deal with the predictive maintenance problem of aircraft engines, which targets to minimize the maximum maintenance completion time for all aircraft engines. Since commercial solvers (e.g. CPLEX) solving it is time-consuming as the problem scale increases, we develop a new fast and effective hybrid metaheuristic algorithm based on the problem features, which combines a genetic algorithm and a variable neighborhood search algorithm. Finally, numerical experiments from the NASA aircraft engine dataset validate the proposed predictive maintenance framework can provide the optimal predictive maintenance plan in less than 10 s for large-scale maintenance problems, thereby reducing aircraft maintenance completion time.
Sustainable management of available water resources needs robust and reliable intelligent tools to address emerging water challenges. These days, artificial intelligence (AI) based tools are more efficient and promine...
详细信息
Sustainable management of available water resources needs robust and reliable intelligent tools to address emerging water challenges. These days, artificial intelligence (AI) based tools are more efficient and prominent in addressing issues related to water treatment plants. Therefore, in the current study, the extreme learning machine (ELM) was optimized with four different metaheuristic algorithms, namely particle swarm optimization (PSO-ELM), genetic algorithm (GA-ELM), biogeography-based optimization (BBO-ELM), and BBO-PSO-ELM for modelling treated water quality parameters, i.e., pHT, Turbidity (TurbT), total dissolved solids (TDST), and HardnessT of Tamburawa water treatment plant (TWTP) located in Nigeria. The performance of the hybrid ELM models was evaluated using mean absolute error (MAE), root mean square error (RMSE), Nash-Sutcliffe efficiency (NSE), Pearson correlation coefficient (PCC), and Willmott index (WI) as well as graphically. The obtained numerical and visualized results indicate that the BBO-PSO-ELM model performed superior in modeling pHT (MAE = 0.403, RMSE = 0.514, NSE = 0.863, PCC = 0.935, WI = 0.964), TDST (MAE = 11.818 mg/L, RMSE = 16.058 mg/L, NSE = 0.711, PCC = 0.853, WI = 0.923), and HardnessT (MAE = 2.624 mg/L, RMSE = 3.497 mg/L, NSE = 0.818, PCC = 0.909, WI = 0.947), while BBO-ELM demonstrated superior performance in TurbT (MAE = 0.385 mg/L, RMSE = 0.694 mg/L, NSE = 0.996, PCC = 0.999, WI = 0.999) modelling. Generally, the findings suggested that the proposed hybrid ELM model has the potential to predict the water quality parameters of TWTP in Nigeria effectively.
Due to the growing emphasis on food safety, peptide research is increasingly focusing on food sources. Traditional methods for determining peptide properties are expensive. While artificial intelligence (AI) models ca...
详细信息
Due to the growing emphasis on food safety, peptide research is increasingly focusing on food sources. Traditional methods for determining peptide properties are expensive. While artificial intelligence (AI) models can reduce cost, existing peptide models often lack accuracy. This study aimed to develop a regression model capable of predicting peptide properties. We integrated the ESM-2 model with the LSTM architecture and optimized the model structure using three metaheuristic algorithms, including WOA, SSA, and HHO. Using an antioxidant tripeptide dataset, our model achieved an R2 of 0.9458 and RMSE of 0.3135, outperforming the state-of-the-art (SOTA) model by 11.66 % and 50.00 %, respectively. The developed model was further applied to the bitter peptide dataset, resulting in R2 of 0.8385 and RMSE of 0.4414, respectively. These results suggest that our model has the potential to accurately predict the properties of various types of peptides.
Breast cancer, caused by uncontrolled cell growth in milk ducts or lobules, has the highest mortality rate among women worldwide, with Asia reporting the most deaths. Early detection improves survival rates and reduce...
详细信息
Breast cancer, caused by uncontrolled cell growth in milk ducts or lobules, has the highest mortality rate among women worldwide, with Asia reporting the most deaths. Early detection improves survival rates and reduces treatment costs. This study aims to develop a feature selection-based classifier to enhance breast cancer prediction, using a minimal dataset to maximize performance. Earlier works on the Breast Cancer Coimbra dataset used many features but failed to achieve high accuracy or explain misclassifications. We addressed this by reducing the features while maintaining performance. A Wrapper Model with metaheuristic algorithms: Whale Optimization, Bald Eagle Search, and Sea Lion Optimization and Extreme Gradient Boost Classifier. SHAP explained feature importance for both overall and individual predictions. Our study achieved F-scores of 97.43%, 95%, and 94.74% for SLOA_XGB, BESA_XGB, and WOA_XGB, respectively, on the Breast Cancer Coimbra dataset. Each method reduced the features from 9 to 4. SHAP analysis identified Glucose as having the highest impact on model predictions. Additionally, we found a link between the mean values of certain features and misclassification likelihood. This study analyzed data from 116 subjects, with the SLOA_XGB classifier achieving the best performance: 97.43% F-score, 97.14% accuracy, 97.14% precision, and 100% recall using only Glucose, Age, Resistin, and Adiponectin. These results highlight the potential for early breast cancer detection with fewer features while maintaining high predictive accuracy.
The globalization of economies has heightened the importance of efficient logistics networks, particularly in the post-COVID-19 era where supply chain resilience is paramount. This study focuses on designing logistics...
详细信息
The globalization of economies has heightened the importance of efficient logistics networks, particularly in the post-COVID-19 era where supply chain resilience is paramount. This study focuses on designing logistics hub networks to enhance global shipment flow management. A hub-based structure is adopted, aligned with existing global logistics frameworks. The research addresses key challenges such as optimizing warehouse and hub locations to streamline cargo delivery. An integrated model is developed for route planning, incorporating multi-modal transportation and cost-time trade-offs. The study employs metaheuristic algorithms including Strength Pareto Evolutionary Algorithm (SPEA-II), Non-Dominated Sorting Genetic Algorithm II (NSGA-II), and Multi-Objective Grey Wolf Optimization Algorithm to solve the proposed model. Detailed numerical analyses demonstrate the model's capability to generate optimal global solutions, underscoring its practical utility in real-world logistics management.
Photovoltaic-Thermal (PV/T) systems integrated with Direct Contact Membrane Distillation (DCMD) offer a promising solution for sustainable water desalination by simultaneously utilizing solar energy for electricity ge...
详细信息
Photovoltaic-Thermal (PV/T) systems integrated with Direct Contact Membrane Distillation (DCMD) offer a promising solution for sustainable water desalination by simultaneously utilizing solar energy for electricity generation and thermal applications. This study presents a dynamic optimization framework to enhance the seasonal performance of PV/T-DCMD systems under variable climatic conditions. The objective is to maximize electrical efficiency, permeate flux, and outlet temperature by considering the effects of flow rates (1-10 L/min), solar irradiation (10-1000 W/m2), and ambient temperatures (291-324 K). metaheuristic algorithms, including Particle Swarm Optimization (PSO) and Genetic Algorithm (GA), are employed to determine the optimal operating parameters. The results reveal that PSO achieved an optimal outlet temperature of 329.45 K and permeate flux of 19.89 kg/m2 & sdot;h at a flow rate of 1.0 L/min, solar irradiation of 1000 W/m2, and ambient temperature of 324 K. In comparison, GA yielded a slightly lower outlet temperature of 329.01 K and permeate flux of 19.80 kg/ m2 & sdot;h under the same conditions. Electrical efficiency stabilized at approximately 0.13 for both optimization techniques. These findings highlight the advantages of reduced flow rates and elevated solar irradiation in optimizing the system's thermal and electrical performance. This comprehensive analysis not only demonstrates the operational characteristics of PV/T-DCMD systems but also establishes a foundation for future advancements in hybrid optimization techniques and material innovations to further enhance system performance under dynamic weather conditions.
Recent studies on carbon fiber-reinforced mortars for electromagnetic interference (EMI) shielding have predominantly relied on practical experiments to investigate the correlation between shielding effectiveness (SE)...
详细信息
Recent studies on carbon fiber-reinforced mortars for electromagnetic interference (EMI) shielding have predominantly relied on practical experiments to investigate the correlation between shielding effectiveness (SE) and design attributes. However, these experiments are resource intensive. Machine learning (ML) models present a faster, cost-effective alternative for simulating outcomes and exploring various scenarios. This study adopts a novel approach by utilizing hybrid models, which offer greater accuracy than individual or ensemble ML models. Specifically, support vector regression (SVR) was combined with three optimization algorithms: firefly algorithm (FFA), particle swarm optimization (PSO), and grey wolf optimization (GWO) to create hybrid models for estimating the SE of carbon fiber-reinforced mortars. Conventional ML techniques like random forest (RF) and decision tree (DT) were also employed for comparison. A dataset of 346 experimental data sets from existing literature was used to evaluate model performance. The SVRPSO hybrid model demonstrated superior performance, achieving the highest coefficient of determination (R2) value of 0.994, compared to SVR-FFA (0.964) and SVR-GWO (0.929). Model interpretability methods identified the aspect ratio (AR) as the most influential parameter, showing that shielding effectiveness (SE) increases significantly with fiber content (FC) up to 0.7 %, after which it stabilizes, with a linear correlation between SE and AR. A user-friendly interface was developed for instant SE prediction of carbon fiber reinforced mortar, requiring only essential input parameters.
Background: Accurately predicting the specific capacity of supercapacitors (SCs) is essential for improving their energy efficiency and performance. This requires robust methods to model the complex, nonlinear relatio...
详细信息
Background: Accurately predicting the specific capacity of supercapacitors (SCs) is essential for improving their energy efficiency and performance. This requires robust methods to model the complex, nonlinear relationships among variables. Methods: In this study, the dataset was divided into three optimal clusters using k-means, based on supercapacitor capacity, each displaying distinct features. Additionally, the unclustered dataset was also analyzed. The training of Multi-Layer Perceptron (MLP) neural networks was examined using six metaheuristic algorithms. Neural network hyperparameters were optimized via grid search, and metaheuristic algorithms via random search. Performance, convergence, and adaptability were evaluated for clustered and unclustered datasets, focusing on accuracy, speed, and generalization. Significant findings: The cluster-based MLP models demonstrated exceptional predictive accuracy, outperforming unclustered models. Notably, the MLP integrated with Invasive Weed Optimization (MLP-IWO) in cluster 2, with a population size (Np) of 40, achieved the highest coefficient of determination (R2=0.9998), representing a 105.53 % improvement compared to the best unclustered model (R2 = 0.4864). Similarly, the MLP integrated with the Firefly Algorithm (MLP-FA) in clusters 1 and 3 (Np = 30) achieved R2 values of 0.9983 and 0.9927, respectively. These findings highlight the effectiveness of integrating clustering with metaheuristic optimization for enhancing prediction accuracy in SCs capacity modeling.
The Internet of Things highly depends on computing environments like cloud computing to process and store information. The use of cloud computing by smart devices leads to challenges such as delay and increased energy...
详细信息
The Internet of Things highly depends on computing environments like cloud computing to process and store information. The use of cloud computing by smart devices leads to challenges such as delay and increased energy consumption of sensors. A primary solution to the mentioned problems is fog computing. Task scheduling is the most critical issue that significantly affects improving the performance of cloud-fog systems. Task scheduling is an NP-hard problem, and applying data mining methods and metaheuristic algorithms to obtain optimal solutions in a reasonable computing time is a fundamental requirement. This paper proposes a new model based on metaheuristic algorithms using the combination of golden jackal optimization (GJO) and beluga whale optimization (BWO) algorithm called GJOBWO to solve the task scheduling problem in a cloud-fog environment. In the hybrid model, the BWO algorithm is used to solve the issues of the GJO algorithm, such as getting stuck in the local optimum and imbalance between the exploration and exploitation stages. Performing the exploration and exploitation steps is essential because correct execution may lead to efficient solutions. Also, the k-means algorithm inspired by clustering is used to prioritize tasks. The evaluation of the hybrid model has been done using continuous optimization functions and task scheduling problems. First, the hybrid model has been implemented on 68 standard functions and compared with particle swarm optimization (PSO), sine cosine algorithm (SCA), whale optimization algorithm (WOA), grey wolf optimizer (GWO), ant lion optimizer (ALO), and GJO and BWO algorithms. Then, the hybrid model has been tested on the task scheduling problem and compared with WOA, GJO, and BWO algorithms. The results show that the hybrid model has effectively minimized the makespan rate and the degree of imbalance. Also, the average improvement percentage of the combined model based on the PIR criterion compared to the four algorithms S
暂无评论