Saved Queries

The deformation mechanism during the hot compression of PM nickel-based superalloy FGH99 and its micro-structural evolution, especially the evolution of γ’ phases, are the key factors affecting the final molding quality of aero-engine hot forged turbine disks. In this study, a new constitutive model of viscoplasticity with micro-structures as physical internal parameters were developed to simulate the hot compression behavior of FGH99 by incorporating the strengthening effect of the γ’ phase. The mechanical behavior of high-temperature (>1000 K) compressive deformation of typical superalloys under a wide strain rate (0.001~1 s⁻¹) is investigated using the Gleeble thermal-force dynamic simulation tester. The micro-structure after the hot deformation was characterized using EBSD and TEM. Work hardening as well as dynamic softening were observed in the hot compression tests. Based on the mechanical responses and micro-structural features, the model considered the coupled effects of dislocation density, DRX, and γ’ phase during hot flow. The model is programmed into a user subroutine based on the Fortran language and called in the simulation of the DEFORM-3D V6.1 software, thus realizing the multiscale predictive simulation of FGH99 alloy by combining macroscopic deformation and micro-structural evolution. The established viscoplastic constitutive model shows a peak discrepancy of 10.05% between its predicted hot flow stresses and the experimental values. For the average grain size of FGH99, predictions exhibit an error below 7.20%. These results demonstrate the high accuracy of the viscoplastic constitutive model developed in this study. Full article

(This article belongs to the Special Issue Additive Manufacturing and Forming Technologies of Alloy: Mechanical Properties and Microstructure Evaluation)

►▼ Show Figures

Figure 1

18 pages, 5224 KB

Open AccessArticle

Study on the Negative Skin Friction on Bridge Pile Foundations Induced by Backfilling in Karst Areas

by Huiyun Chen, Zhongju Feng, Xiaodong Wei and Ya Li

Buildings 2025, 15(20), 3672; https://doi.org/10.3390/buildings15203672 (registering DOI) - 12 Oct 2025

Abstract

The load transfer mechanism of piles in karst cavity areas was investigated through field tests, and an orthogonal test was carried out to establish a calculation method for negative skin friction induced by backfilling. The results indicate that the negative skin friction of piles is strongly influenced by the type of cavity. When cavities were completely filled with limestone breccia mixed with silty clay and the applied load reached 3628 kN, the unit side friction ranged from 15 to 22 kPa. In contrast, when cavities remained unfilled, soil settlement occurred around the pile after backfilling, leading to the development of negative skin friction. For cavities with heights of 3–12 m, it is recommended that the bearing capacity of piles be calculated by considering negative skin friction at depths of 0H, 0.106H, 0.214H, and 0.271H (where H denotes the cavity height). Based on 21 orthogonal tests, the sensitivity ranking of factors affecting negative skin friction was determined as follows: cavity height > elastic modulus of backfill > pile diameter > cavity span > pile length > cavity position. The calculated values of negative skin friction were further validated against engineering data, showing a variation trend consistent with the test results, with a relative error of only 7.4%. Full article

(This article belongs to the Section Building Structures)

►▼ Show Figures

Figure 1

27 pages, 13950 KB

Open AccessArticle

Predicting Perceived Restorativeness of Urban Streetscapes Using Semantic Segmentation and Machine Learning: A Case Study of Liwan District, Guangzhou

by Wenjuan Kang, Ni Kang and Pohsun Wang

Buildings 2025, 15(20), 3671; https://doi.org/10.3390/buildings15203671 (registering DOI) - 12 Oct 2025

Abstract

Urban streetscapes are among the most frequently encountered spatial environments in daily life, and their restorative visual features have a significant impact on well-being. Although existing studies have revealed the relationship between streetscape environments and perceived restorativeness, there remains a lack of scalable, data-driven methods for quantifying such perception at the street level. This study proposes an interpretable and replicable framework for predicting streetscape restorativeness by integrating semantic segmentation, perceptual evaluation, and machine learning techniques. Taking Liwan District of Guangzhou as a case study, street-view images (SVIs) were collected and processed using the Mask2Former model to extract the following five key visual metrics: greenness, openness, enclosure, walkability, and imageability. Based on the Perceived Restorativeness Scale (PRS), an online questionnaire was designed from four dimensions (fascination, being away, compatibility, and extent) to score a random sample of images. A random forest model was then trained to predict the perceptual levels of the full dataset, followed by K-means clustering to identify spatial distribution patterns. The results revealed that there were significant differences in visual characteristics among high, medium, and low restorativeness street types. The proposed framework enables scalable, data-driven evaluation of perceived restorativeness across diverse urban streetscapes. By embedding perceptual metrics into large-scale urban analysis, the framework offers a replicable and efficient approach for identifying streets with low restorative potential—thus providing urban planners and policymakers with a novel tool for prioritizing street-level renewal, improving public well-being, and supporting perception-oriented urban design without the need for labor-intensive fieldwork. Full article

(This article belongs to the Special Issue Sustainable Housing and Urban Planning: Enhancing Well-Being Through Environmental Design)

36 pages, 4822 KB

Open AccessReview

Converting Wastewater Sludge into Slow-Release Fertilizers via Biochar and Encapsulation Technologies

by Babar Azeem

Appl. Sci. 2025, 15(20), 10954; https://doi.org/10.3390/app152010954 (registering DOI) - 12 Oct 2025

Abstract

The rising demand for sustainable agriculture and circular resource management has intensified interest in converting wastewater sludge into value-added products. This review explores the transformation of sewage sludge into slow- and controlled-release fertilizers (CRFs), with a focus on biochar production and encapsulation technologies. Sewage sludge is rich in essential macronutrients (N, P, K), micronutrients, and organic matter, making it a promising feedstock for agricultural applications. However, its use is constrained by challenges including compositional variability, presence of heavy metals, pathogens, and emerging contaminants such as microplastics and PFAS (Per- and Polyfluoroalkyl Substances). The manuscript discusses a range of stabilization and conversion techniques, such as composting, anaerobic digestion, pyrolysis, hydrothermal carbonization, and nutrient recovery from incinerated sludge ash. Special emphasis is placed on coating and encapsulation technologies that regulate nutrient release, improve fertilizer efficiency, and reduce environmental losses. The role of natural, synthetic, and biodegradable polymers in enhancing release mechanisms is analyzed in the context of agricultural performance and soil health. While these technologies offer environmental and agronomic benefits, large-scale adoption is hindered by technical, economic, and regulatory barriers. The review highlights key challenges and outlines future perspectives, including the need for advanced coating materials, improved contaminant mitigation strategies, harmonized regulations, and field-scale validation of CRFs. Overall, the valorisation of sewage sludge into CRFs presents a viable strategy for nutrient recovery, waste minimization, and sustainable food production. With continued innovation and policy support, sludge-based fertilizers can become a critical component of the green transition in agriculture. Full article

(This article belongs to the Section Green Sustainable Science and Technology)

►▼ Show Figures

Figure 1

24 pages, 1551 KB

Open AccessArticle

Tester-Guided Graph Learning with End-to-End Detection Certificates for Triangle-Based Anomalies

by Manuel J. C. S. Reis

Big Data Cogn. Comput. 2025, 9(10), 257; https://doi.org/10.3390/bdcc9100257 (registering DOI) - 12 Oct 2025

Abstract

We investigate anomaly detection in complex networks through a property-testing-guided graph neural model (PT-GNN) that provides an end-to-end miss-probability certificate

(δ + α)

. The method combines (i) a wedge-sampling tester that estimates triangle-closure frequency and derives a concentration bound [...] Read more.

We investigate anomaly detection in complex networks through a property-testing-guided graph neural model (PT-GNN) that provides an end-to-end miss-probability certificate

(δ + α)

. The method combines (i) a wedge-sampling tester that estimates triangle-closure frequency and derives a concentration bound

(δ)

via Bernstein’s inequality, with (ii) a lightweight classifier over structural features whose validation error contributes

(α)

. The overall certificate is given by the sum

(δ + α)

, quantifying the probability of missed anomalies under bounded sampling. On synthetic communication graphs with n = 1000, edge probability p = 0.01, and anomalous subgraph size k = 120, PT-GNN achieves perfect detection performance (AUC = 1.0, F1 = 1.0) across all tested regimes. Moreover, the miss-probability certificate tightens systematically as the tester budget m increases (e.g., for

ε

= 0.06, enlarging m from 2000 to 8000 reduces

(δ + α)

from ≈0.87 to ≈0.49). These results demonstrate that PT-GNN effectively couples graph learning with property testing, offering both strong empirical detection and formally verifiable guarantees in anomaly detection tasks. Full article

(This article belongs to the Special Issue Advances in Graph Learning and Representation Models for Complex Network Analysis)

34 pages, 6332 KB

Open AccessArticle

Optimal Sizing of an Off-Grid Hybrid Energy System with Metaheuristics and Meteorological Forecasting Based on Wavelet Transform and Long Short-Term Memory Networks

by Yamilet González Cusa, José Hidalgo Suárez, Jorge Laureano Moya Rodríguez, Tulio Hernández Ramírez, Silvio A. B. Vieira de Melo and Ednildo Andrade Torres

Energies 2025, 18(20), 5371; https://doi.org/10.3390/en18205371 (registering DOI) - 12 Oct 2025

Abstract

This study proposes an integrated framework for the optimal sizing of off-grid hybrid energy systems, combining photovoltaic panels, wind turbines, battery storage, a diesel generator, and an inverter. The methodology uniquely integrates long-term meteorological forecasting through a hybrid approach based on the Discrete Wavelet Transform and Long Short-Term Memory networks, together with metaheuristic optimization techniques (Particle Swarm Optimization and Genetic Algorithm), to minimize the system’s total annual cost. A case study was conducted in Guanambi, Brazil, using ten years (2012–2021) of hourly data on wind speed, solar irradiance, and ambient temperature. Forecasting results show that the hybrid Discrete Wavelet Transform–Long Short-Term Memory model outperforms the conventional Long Short-Term Memory approach, reducing error metrics and improving predictive accuracy. In the optimization stage, Particle Swarm Optimization consistently achieved lower costs and more stable convergence compared to the Genetic Algorithm. The optimal configuration comprised 450 photovoltaic panels, 10 wind turbines, 66 lithium iron phosphate battery, and 1 diesel generator, yielding a total annual cost of $105,381.17, a cost of energy of $0.1243/kWh, and minimal diesel dependence ($8825.89 annually). The proposed framework demonstrates robustness, economic viability, and applicability for providing sustainable and reliable electricity in isolated regions with high renewable energy potential. Full article

20 pages, 3100 KB

Open AccessArticle

The Effect of Retention Time and Seasonal Variation on the Characterization of Phyto-Remediated Aquaculture Wastewater in a Constructed Wetland

by Shadrach A. Akadiri, Pius O. O. Dada, Adekunle A. Badejo, Olayemi J. Adeosun, Akinwale T. Ogunrinde, Oluwaseun T. Faloye, Viroon Kamchoom and Oluwafemi E. Adeyeri

Biology 2025, 14(10), 1390; https://doi.org/10.3390/biology14101390 (registering DOI) - 12 Oct 2025

Abstract

The insufficient availability of safe water has emerged as a prevalent issue severely impacting public health in developing nations. Moreover, studies reporting the efficacy of treatment plants (TPs)—specifically Phragmites karka and Typha latifolia—in removing toxic elements in aquaculture wastewater are scanty. Therefore, this study is aimed at investigating the effects of hydraulic retention time (HRT), seasonal variations, and TPs on the removal efficiency of pollutants from a vertical subsurface flow constructed wetland (VSSF-CW) in Nigeria. The experiments spanned three seasons (November–December–January—NDJ; March–April–May—MAM; and July–August–September—JAS) of the year, with samples collected from the CW at 7 day intervals for analysis. The aquaculture wastewater was analyzed in the laboratory to determine its chemical and toxic compositions before and after the introduction of treatment plants. Three-way ANOVA was used to analyze the main and interactive effects between HRT, seasons, and TPs on the physicochemical properties of the CW’s effluents. The removal efficiency was determined to evaluate the performance of the constructed wetland in comparison to the treatment plants. Results showed that these constructed wetlands effectively removed contaminants, with significant differences (p < 0.05) mostly observed in the effects of treatment plant types and seasons on the chemical and heavy metal concentrations. This was further confirmed by the main effects of HRT, seasons, and treatment plant choice, which significantly (p < 0.05) influenced treatment efficiency. Removal efficiencies increased with longer HRTs, reaching peak removal efficiencies of approximately 69, 67, and 61% for Na, K, and Ca, respectively. The BOD and COD reached 85 and 90% removal efficiency, while removal efficiency of 100% was achieved for most heavy metals at 21 day retention time. In summary, the study found that TPs (Phragmites karka and Typha latifolia), HRT, and seasonal variation are important for treating integrated poultry and aquaculture wastewater in a VSSF CWs. Full article

(This article belongs to the Special Issue Heavy Metal Pollution and Bioremediation: Application and Mechanism)

►▼ Show Figures

Figure 1

51 pages, 427 KB

Open AccessArticle

Existence of Generalized Maxwell–Einstein Metrics on Completions of Certain Line Bundles

by Jing Chen and Daniel Guan

Mathematics 2025, 13(20), 3264; https://doi.org/10.3390/math13203264 (registering DOI) - 12 Oct 2025

Abstract

In Kähler geometry, Calabi extremal metrics serves as a class of more available special metrics than Kähler metrics with constant scalar curvatures, as a generalization of Kähler Einstein metrics. In recent years, Maxwell–Einstein metrics (or conformally Kähler Einstein–Maxwell metrics) appeared as another alternative choice for Calabi extremal metrics. It turns out that some similar metrics defined by Futaki and Ono have similar roles in the Kähler geometry. In this paper, we prove that for some completions of certain line bundles, there is at least one k-generalized Maxwell–Einstein metric defined by Futaki and Ono conformally related to a metric in any given Kähler class for any integer

3 \leq k \leq 13

. Full article

15 pages, 1502 KB

Open AccessArticle

Geographical Variation in the Mineral Profiles of Camel Milk from Xinjiang: Implications for Nutritional Value and Species Identification

by Qiaoye Yang, Luhan Xu, Weihua Zheng, Delinu’er Baisanbieke, Lin Zhu, Mireguli Yimamu and Fengming Li

Agriculture 2025, 15(20), 2120; https://doi.org/10.3390/agriculture15202120 (registering DOI) - 12 Oct 2025

Abstract

To investigate the geographical and species differences regarding mineral element content of camel milk, this research used camel milk from the Tacheng, Altay, and Ili regions of Xinjiang and cow milk, goat milk, and horse milk from the Tacheng region as subjects. The contents of 22 mineral elements were measured using inductively coupled plasma mass spectrometry (ICP-MS) and inductively coupled plasma optical emission spectrometry (ICP-OES). The results showed that the contents of macro elements Ca, P, K, and Na in camel milk were significantly higher than those in other milk sources (p < 0.01). The contents of trace elements such as Se, Sr, and Ni were very significantly higher than those in other milk sources (p < 0.01). The content of 12 mineral elements in camel milk was very significantly higher than in other types of milk (p < 0.01). Principal component analysis (PCA) and factor analysis emphasized the relationship between element distribution and different milk sources, and the linear discriminant analysis (LDA) model could identify the species type of milk. Geographical analysis indicated that trace elements such as Sr, Ni, and Cr were highly significantly enriched in Tacheng camel milk (p < 0.01). The established LDA model achieved traceability of the geographical origin of Xinjiang camel milk. This research reveals the mineral nutritional advantages of camel milk and its geographical differentiation patterns, providing theoretical support for exploring the functional properties of camel milk and for identifying species and regions through minerals. It is important to promote the upgrading of the specialty dairy product industry. Full article

(This article belongs to the Section Farm Animal Production)

►▼ Show Figures

Figure 1

22 pages, 356 KB

Open AccessArticle

Optimal Hölder Regularity for Discontinuous Sub-Elliptic Systems Structured on Hörmander’s Vector Fields

by Dongni Liao and Jialin Wang

Axioms 2025, 14(10), 761; https://doi.org/10.3390/axioms14100761 (registering DOI) - 12 Oct 2025

Abstract

This paper studies discontinuous quasilinear sub-elliptic systems associated with Hörmander’s vector fields under controllable and natural growth conditions. By a new

A

-harmonic approximation reformulation for bilinear forms

A \in Bil (R^{k N}, R^{k N})

, we obtain [...] Read more.

This paper studies discontinuous quasilinear sub-elliptic systems associated with Hörmander’s vector fields under controllable and natural growth conditions. By a new

A

-harmonic approximation reformulation for bilinear forms

A \in Bil (R^{k N}, R^{k N})

, we obtain optimal partial Hölder continuity with exact exponents for weak solutions with vanishing mean oscillation coefficients. Full article

20 pages, 3797 KB

Open AccessArticle

Induced Mammary Epithelial Cell-Derived Extracellular Vesicles Promote the Repair of Skin Trauma

by Siyao Pan, Dandan Zhang, Guodong Wang, Longfei Sun, Mengzhen Wei, Shan Deng, Jianwei Chen, Prasanna Kallingappa, Xiang Yuan and Ben Huang

Int. J. Mol. Sci. 2025, 26(20), 9929; https://doi.org/10.3390/ijms26209929 (registering DOI) - 12 Oct 2025

Abstract

Although extracellular vesicles (EVs) from mesenchymal stem cells have shown potential in skin wound repair, the diversity of EV sources and the optimization of delivery systems still need further exploration. This study is the first to demonstrate that extracellular vesicles from chemically induced mammary epithelial cells (CiMECs-EVs) possess distinct skin wound repair activity. To enhance the therapeutic efficacy of CiMECs-EVs and optimize their delivery efficiency, we innovatively combined them with a chitosan hydrogel to construct a composite repair system (CiMECs-EVs-chitosan hydrogel, CMECG). This system was then applied to a rat skin wound model. The results showed that CMECG significantly promoted the proliferation and migration of fibroblasts and mammary epithelial cells (MECs). In animal experiments, the relative wound closure efficiency of the control group was approximately 70% on day 14, while that of the CMECG group (loaded with 200 μg CiMECs-Exo) was enhanced to 90%, markedly accelerating the wound healing process. Histological analysis indicated that this system could effectively restore the structural continuity of various skin layers and significantly promote the synthesis and remodeling of collagen at the wound site. Mechanistically, the wound healing effect of CiMECs-EVs is closely associated with the endogenous miRNAs they encapsulate. These miRNAs can coordinately regulate cell proliferation, migration, and angiogenesis, modulate the inflammatory microenvironment, and inhibit excessive scar formation—thus regulating the entire repair process. This process involves multiple wound healing-related signaling pathways, including MAPK, PI3K-Akt, FoxO, TGF-β, and JAK-STAT. In summary, this study successfully constructed a novel EV-chitosan hydrogel repair system. This system is expected to provide an effective and innovative EV-based therapeutic strategy for the clinical treatment of skin wound repair. Full article

(This article belongs to the Section Biochemistry)

►▼ Show Figures

Figure 1

14 pages, 605 KB

Open AccessStudy Protocol

Monitoring and Follow-Up of Patients on Vitamin K Antagonist Oral Anticoagulant Therapy Using Artificial Intelligence: The AIto-Control Project

by Adolfo Romero-Arana, Nerea Romero-Sibajas, Elena Arroyo-Bello, Adolfo Romero-Ruiz and Juan Gómez-Salgado

J. Clin. Med. 2025, 14(20), 7191; https://doi.org/10.3390/jcm14207191 (registering DOI) - 12 Oct 2025

Abstract

Background: Vitamin K antagonist oral anticoagulant (VKA) therapy, using warfarin or acenocoumarol in our health system, is indicated, according to clinical guidelines, for the prophylaxis of thromboembolic events. In Málaga, the VKA patient management program currently includes a total of 856 patients. Hypothesis: The use of an AI-based application can enhance treatment adherence among VKA patients participating in self-monitoring and self-management programs. Furthermore, it can support the comprehensive implementation of the system, leading to reduced costs and fewer interventions for anticoagulated patients. Methods: The study will be conducted in several phases. The first phase involves the development of the application and the integration of Artificial Intelligence (AI) and Machine Learning (ML) algorithms. The second phase includes preliminary testing and validation of the developed application. The third phase consists of full implementation, along with an assessment of user-identified needs and potential quality improvements. Expected Results: The implementation of the AIto-Control app is expected to reduce healthcare-related costs by decreasing primary care visits and hospital admissions due to thromboembolic or bleeding events. Additionally, it aims to ease the workload on both primary care and hospital services. These outcomes will be achieved through the involvement of advanced practice nurses who will supervise app-based monitoring and patient education. Full article

(This article belongs to the Special Issue Thrombosis and Haemostasis: Clinical Advances)

37 pages, 2123 KB

Open AccessReview

Molecular Impact of Metabolic and Endocrine Disturbance on Endometrial Function in Polycystic Ovary Syndrome

by Jim Parker, Claire O’Brien, Talat Uppal and Kelton Tremellen

Int. J. Mol. Sci. 2025, 26(20), 9926; https://doi.org/10.3390/ijms26209926 (registering DOI) - 12 Oct 2025

Abstract

Polycystic ovary syndrome (PCOS) is a systemic metabolic and endocrine disorder that significantly disrupts reproductive physiology and endometrial function. In this narrative review, we examine the molecular impact of metabolic and hormonal imbalances on the endometrium of women with PCOS. We investigate the specific mechanisms that delineate how hyperinsulinemia and insulin resistance, chronic low-grade inflammation, and estrogen/progesterone/androgen imbalance contribute to altered epigenetic, transcriptomic, metabolomic, and signaling profiles in a wide array of different cell types within endometrial tissues. The synergistic interplay between upregulated inflammatory cytokines (e.g., IL-1,2,6,8,17,18, and TNF-α), along with key changes in critical molecular pathways associated with hyperinsulinemia and insulin resistance (e.g., PI3K/AKT/MAPK, and Wnt/β-catenin), in addition to aberrant sex steroid hormone signaling (e.g., CYP19A1, COX-2, PGE₂, HOXA10, 11βHSD2), promotes deleterious changes within the endometrial microenvironment. These anomalies underpin a spectrum of clinical manifestations observed in women with PCOS at each stage of the life course, including abnormal uterine bleeding in reproductive-age women, impaired decidualization in pregnancy, and altered postmenopausal endometrial physiology. Clinically, these alterations are associated with abnormal uterine bleeding, subfertility, implantation failure, miscarriage, pregnancy complications, and postmenopausal endometrial hyperplasia and cancer. Overall, our review provides novel insights into the molecular mechanisms linking systemic metabolic and endocrine dysfunction with endometrial pathology in PCOS and has broader implications that apply to all women. Full article

(This article belongs to the Special Issue Focus on Metabolic Research Priorities in PCOS)

17 pages, 6529 KB

Open AccessArticle

Temperature Field Analysis and Experimental Verification of Mining High-Power Explosion-Proof Integrated Variable-Frequency Permanent Magnet Motor

by Xiaojun Wang, Gaowei Tian, Qingqing Lü, Kun Zhao, Xuandong Wu, Liquan Yang and Guangxi Li

Energies 2025, 18(20), 5369; https://doi.org/10.3390/en18205369 (registering DOI) - 12 Oct 2025

Abstract

An efficient cooling configuration is critical for ensuring the safe operation of electrical machines and is key for optimizing the iterative design of motors. To improve the heat dissipation performance of high-power, explosion-proof, integrated variable-frequency permanent magnet motors used in mining and reduce the risk of permanent magnet demagnetization, this study considers a 1600 kW mining explosion-proof variable-frequency permanent magnet motor as its research object. Based on the zigzag-type water channel structure of the frame, a novel rotor-cooling scheme integrating axial–radial ventilation structures and axial flow fans was proposed. The temperature field of the motor was simulated and analyzed using a fluid–thermal coupling method. Under rated operating conditions, the flow characteristics of the frame water channel and the temperature distribution law inside the motor were compared when the water supply flow rates were 5.4, 4.8, 4.2, 3.6, 3, 2.4, and 1.8 m³/h, respectively, and the relationship between the motor temperature rise and the variation in water flow rate was revealed. A production prototype was developed, and temperature rise tests were conducted for verification. The test results were in good agreement with the simulation calculation results, thereby confirming the accuracy of the simulation calculation method. The results provide an important reference for enterprises in the design optimization and upgrading of high-power explosion-proof integrated variable-frequency permanent-magnet motors. Full article

(This article belongs to the Special Issue Advanced Technology in Permanent Magnet Motors)

►▼ Show Figures

Figure 1

21 pages, 5915 KB

Open AccessArticle

A Machine Learning Approach to Predicting the Turbidity from Filters in a Water Treatment Plant

by Joseph Kwarko-Kyei, Hoese Michel Tornyeviadzi and Razak Seidu

Water 2025, 17(20), 2938; https://doi.org/10.3390/w17202938 (registering DOI) - 12 Oct 2025

Abstract

Rapid sand filtration is a critical step in the water treatment process, as its effectiveness directly impacts the supply of safe drinking water. However, optimising filtration processes in water treatment plants (WTPs) presents a significant challenge due to the varying operational parameters and conditions. This study applies explainable machine learning to enhance insights into predicting direct filtration operations at the Ålesund WTP in Norway. Three baseline models (Multiple Linear Regression, Support Vector Regression, and K-Nearest Neighbour (KNN)) and three ensemble models (Random Forest (RF), Extra Trees (ET), and XGBoost) were optimised using the GridSearchCV algorithm and implemented on seven filter units to predict their filtered water turbidity. The results indicate that ML models can reliably predict filtered water turbidity in WTPs, with Extra Trees models achieving the highest predictive performance (R² = 0.92). ET, RF, and KNN ranked as the three top-performing models using Alternative Technique for Order of Preference by Similarity to Ideal Solution (A-TOPSIS) ranking for the suite of algorithms used. The feature importance analysis ranked the filter runtime, flow rate, and bed level. SHAP interpretation of the best model provided actionable insights, revealing how operational adjustments during the ripening stage can help mitigate filter breakthroughs. These findings offer valuable guidance for plant operators and highlight the benefits of explainable machine learning in water quality management. Full article

(This article belongs to the Special Issue Application of Artificial Intelligence (AI) in Water Quality Monitoring)

►▼ Show Figures

Figure 1

Show export options Show export options

Select all

Export citation of selected articles as:

Error

Oops... you haven't selected anything for export.

Displaying article 1-50 on page 1 of 1686.

Go to page 1 2 3 4 5

Search Results (84,282)

Further Information

Guidelines

MDPI Initiatives

Follow MDPI