Prediction of Ultra-High-Performance Concrete (UHPC) Properties Using Gene Expression Programming (GEP)

Buildings, Год журнала: 2024, Номер 14(9), С. 2675 - 2675

Опубликована: Авг. 28, 2024

In today’s digital age, innovative artificial intelligence (AI) methodologies, notably machine learning (ML) approaches, are increasingly favored for their superior accuracy in anticipating the characteristics of cementitious composites compared to typical regression models. The main focus current research work is improve knowledge regarding application one new ML techniques, i.e., gene expression programming (GEP), anticipate ultra-high-performance concrete (UHPC) properties, such as flowability, flexural strength (FS), compressive (CS), and porosity. addition, process training a model that predicts intended outcome values when associated inputs provided generates graphical user interface (GUI). Moreover, reported models have been created aforementioned UHPC simple limited input parameters. Therefore, purpose this study predict while taking into account wide range factors (i.e., 21) use GUI assess how these parameters affect properties. This includes diameter steel polystyrene fibers (µm mm), length (mm), maximum size aggregate particles type cement, its class, (MPa) type, contents (%), amount water (kg/m3). it fly ash, silica fume, slag, nano-silica, quartz powder, limestone sand, coarse aggregates, super-plasticizers, with all measurements kg/m3. outcomes reveal GEP technique successful accurately predicting characteristics. obtained R2, determination coefficients, from 0.94, 0.95, 0.93, 0.94 CS, FS, porosity, respectively. Thus, utilizes forecast comprehend influence factors, simplifying procedure offering valuable instruments practical model’s capabilities within domain civil engineering.

Язык: Английский

---

Expansion Characteristics and Shear Behavior of Reinforced Concrete Beams Under Non-Uniform Expansion Induced by Alkali–Silica Reaction

Materials, Год журнала: 2025, Номер 18(2), С. 312 - 312

Опубликована: Янв. 11, 2025

Alkali–silica reaction (ASR) is an important factor that seriously affects the durability of reinforced concrete (RC) structures. The current research on alkali-aggregate mainly focuses deterioration mechanism materials and mechanical properties standard specimens. However, there a gap in field effect damage level RC In this study, five beams were tested, depth location alkali solution immersion used as test variables, with aim investigating how steel reinforcement suppresses expansion caused by ASR evaluating shear behavior after non-uniform damage. results study showed increase accelerated rate development, while inhibited development. Compared undamaged beams, initially generates stresses within concrete, which cracking yield loads delay reduces ultimate load-carrying capacity ductility due to disruption microstructure. Finally, chemo-mechanical analysis method proposed based experimental results, incorporate model pore mechanics model. efficacy precision are validated through comparison results.

Язык: Английский

---

FS-DDPG: Optimal Control of a Fan Coil Unit System Based on Safe Reinforcement Learning

Buildings, Год журнала: 2025, Номер 15(2), С. 226 - 226

Опубликована: Янв. 14, 2025

To optimize the control of fan coil unit (FCU) systems under model-free conditions, researchers have integrated reinforcement learning (RL) into processes system pumps and fans. However, traditional RL methods can lead to significant fluctuations in flow fans, posing a safety risk. address this issue, we propose novel FCU method, Fluctuation Suppression–Deep Deterministic Policy Gradient (FS-DDPG). The key innovation lies applying constrained Markov decision process model problem, where penalty term for constraints is incorporated reward function, constraint tightening introduced limit action space. In addition, validate performance proposed established variable operating conditions simulation platform based on parameters actual ten years historical weather data. platform’s correctness effectiveness were verified from three aspects: heat transfer, air side water side, different dry wet conditions. experimental results show that compared with DDPG, FS-DDPG avoids 98.20% pump 95.82% fluctuations, ensuring equipment. Compared DDPG RBC, achieves 11.9% 51.76% energy saving rates, respectively, also shows better terms operational satisfaction. future, will further improve scalability apply method more complex environments.

Язык: Английский

---

Sensorless Position Estimation in Electromagnetic Launchers Using Recurrent Neural Networks with Repeated k-Fold Cross-Validation

Arabian Journal for Science and Engineering, Год журнала: 2025, Номер unknown

Опубликована: Янв. 17, 2025

Язык: Английский

---

Development of Robust Machine Learning Models for Predicting Flexural Strengths of Fiber-Reinforced Polymeric Composites

Hybrid Advances, Год журнала: 2025, Номер unknown, С. 100385 - 100385

Опубликована: Янв. 1, 2025

Язык: Английский

---