Engineering Science and Technology-An International Journal-Jestech最新文献_第10页

In-band spurious suppression in first-order antisymmetric-mode resonators using air holes and half electrode 利用空穴和半电极抑制一阶反对称模谐振器带内杂散

IF 5.4 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Engineering Science and Technology-An International Journal-Jestech

Pub Date : 2025-11-01 Epub Date: 2025-10-14 DOI: 10.1016/j.jestch.2025.102207

Shitao Lv , Dongchen Sui , Qiner Xu , Yi Cao , Qiang Xu , Haiyan Sun , Tengfei Xu , Jicong Zhao

Laterally excited bulk acoustic resonator (XBAR) exhibits a high electromechanical coupling coefficient (k_t²) due to their intrinsically superior piezoelectric properties, which is highly advantageous for the subsequent design and implementation of acoustic filters. However,it is not yet suitable for commercial application due to its numerous spurious modes and low quality factor (Q value). This paper presents an exhaustive analysis of the origins of in-band spurious modes and the method to suppress based on the first-order antisymmetric (A1) mode XBAR. Simulation results indicate that these spurious modes primarily originate from acoustic wave scattering along unintended propagation paths and standing waves caused by crosstalk between electrodes. These spurious modes lead to energy dissipation, thereby reducing the resonator’s Q value. To effectively suppress them without increasing fabrication complexity while significantly enhancing the Q value, an etching-hole strategy is proposed, introducing the transverse holes that create acoustic impedance mismatch based on velocity differences. Additionally, the adoption of a half electrode configuration optimizes the electric field and stress redistribution within the resonator, thereby effectively reducing interfacial losses at the electrode-piezoelectric layer interface and enhancing the Q value by approximately 3.5 times. The fabricated XBAR, integrating holes and half electrode configuration, exhibits an in-band spurious-free response, a Bode-Q value of 1954, a k_t² of 27.5 % and an exceptionally high figure of merit (FOM = k_t² × Bode-Q) of 537.

横向激发体声谐振器（XBAR）由于其固有的优异压电特性而具有较高的机电耦合系数（kt2），这对后续声滤波器的设计和实现非常有利。但由于其杂散模式多，质量因子（Q值）低，尚不适合商业应用。本文详细分析了带内杂散模的来源，并提出了一阶反对称（A1）模XBAR的抑制方法。仿真结果表明，这些杂散模式主要来源于非预期传播路径上的声波散射和电极间串扰引起的驻波。这些伪模导致能量耗散，从而降低了谐振器的Q值。为了在不增加制造复杂性的同时有效地抑制它们，同时显著提高Q值，提出了一种蚀刻孔策略，该策略引入了基于速度差产生声阻抗失配的横向孔。此外，采用半电极结构优化了谐振腔内的电场和应力分布，从而有效降低了电极-压电层界面处的界面损耗，使Q值提高了约3.5倍。所制备的XBAR集成了孔和半电极结构，具有带内无杂散响应，波德q值为1954，kt2为27.5%，优异值（FOM = kt2 ×波德q）为537。

{"title":"In-band spurious suppression in first-order antisymmetric-mode resonators using air holes and half electrode","authors":"Shitao Lv , Dongchen Sui , Qiner Xu , Yi Cao , Qiang Xu , Haiyan Sun , Tengfei Xu , Jicong Zhao","doi":"10.1016/j.jestch.2025.102207","DOIUrl":"10.1016/j.jestch.2025.102207","url":null,"abstract":"<div><div>Laterally excited bulk acoustic resonator (XBAR) exhibits a high electromechanical coupling coefficient (<em>k<sub>t</sub></em><sup>2</sup>) due to their intrinsically superior piezoelectric properties, which is highly advantageous for the subsequent design and implementation of acoustic filters. However,it is not yet suitable for commercial application due to its numerous spurious modes and low quality factor (<em>Q</em> value). This paper presents an exhaustive analysis of the origins of in-band spurious modes and the method to suppress based on the first-order antisymmetric (A1) mode XBAR. Simulation results indicate that these spurious modes primarily originate from acoustic wave scattering along unintended propagation paths and standing waves caused by crosstalk between electrodes. These spurious modes lead to energy dissipation, thereby reducing the resonator’s <em>Q</em> value. To effectively suppress them without increasing fabrication complexity while significantly enhancing the <em>Q</em> value, an etching-hole strategy is proposed, introducing the transverse holes that create acoustic impedance mismatch based on velocity differences. Additionally, the adoption of a half electrode configuration optimizes the electric field and stress redistribution within the resonator, thereby effectively reducing interfacial losses at the electrode-piezoelectric layer interface and enhancing the <em>Q</em> value by approximately 3.5 times. The fabricated XBAR, integrating holes and half electrode configuration, exhibits an in-band spurious-free response, a Bode-<em>Q</em> value of 1954, a <em>k<sub>t</sub></em><sup>2</sup> of 27.5 % and an exceptionally high figure of merit (<em>FOM</em> = <em>k<sub>t</sub></em><sup>2</sup> × Bode-<em>Q</em>) of 537.</div></div>","PeriodicalId":48609,"journal":{"name":"Engineering Science and Technology-An International Journal-Jestech","volume":"71 ","pages":"Article 102207"},"PeriodicalIF":5.4,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145321448","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Design and analysis of a continuously linear polarization agile antenna using microwave liquid crystal technology 基于微波液晶技术的连续线极化捷变天线设计与分析

IF 5.4 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Engineering Science and Technology-An International Journal-Jestech

Pub Date : 2025-11-01 Epub Date: 2025-10-01 DOI: 10.1016/j.jestch.2025.102201

Onur Hamza Karabey

This paper presents the design and analysis of a continuously linear polarization-agile antenna based on Microwave Liquid Crystal (LC) technology. The proposed concept is specifically developed for Ku-band (e.g. 12 GHz) satellite communication applications, where systems typically operate with linear polarization and maintaining polarization alignment, referred to as skew angle adjustment, is critical for reliable on-the-move operation and for avoiding link loss. The antenna system incorporates a tunable feed network comprising two directional couplers and LC-based continuously tunable phase shifters. These feed a dual-fed microstrip patch antenna, enabling full-range linear polarization control over

\pm 18 0^{\circ}

by electronically adjusting the differential phase. The proposed tunable feed network is implemented using microwave LC technology to enable a fully passive architecture without requiring low-noise amplifiers. LC materials offer electrically tunable dielectric properties with inherently low dielectric loss, typically below 0.06 at frequencies beyond 10 GHz, and their loss tangent decreases with increasing frequency, making them particularly well-suited for Ku-band applications and beyond. In addition, LC materials benefit from a mature fabrication infrastructure originating from LC display manufacturing, allowing low-cost, large-area, and scalable production. Full-wave simulations at 12 GHz show that the system achieves a cross-polarization discrimination (XPD) better than 25 dB across the entire skew angle range. The LC-based phase shifters operate with only two bias voltages and provide 180° differential phase shift with an insertion loss below 2.3 dB, yielding a figure-of-merit of 80°/dB at 12 GHz. These results demonstrate the feasibility of the proposed design for low-loss, scalable, and reconfigurable antenna arrays in Ku-band and beyond satellite systems.

介绍了一种基于微波液晶技术的连续线性极化捷变天线的设计与分析。所提出的概念是专门为ku波段（例如12 GHz）卫星通信应用开发的，其中系统通常以线性极化运行，并且保持极化对准（称为倾斜角度调整）对于可靠的移动操作和避免链路丢失至关重要。天线系统包含一个可调谐馈电网络，包括两个定向耦合器和基于lc的连续可调谐移相器。这些天线馈送一个双馈微带贴片天线，通过电子调节差动相位，在±180°范围内实现线性极化控制。所提出的可调谐馈源网络采用微波LC技术实现，无需低噪声放大器即可实现全无源架构。LC材料具有电可调谐的介电特性，具有固有的低介电损耗，在10 GHz以上的频率下通常低于0.06，并且其损耗正切随频率的增加而降低，使其特别适合ku波段及更高波段的应用。此外，LC材料受益于源于LC显示器制造的成熟制造基础设施，允许低成本，大面积和可扩展的生产。12 GHz全波仿真结果表明，该系统在整个斜角范围内的交叉极化判别（XPD）优于25 dB。基于lc的移相器仅在两个偏置电压下工作，并提供180°差分移相，插入损耗低于2.3 dB，在12 GHz时产生80°/dB的性能因数。这些结果证明了所提出的低损耗、可扩展和可重构天线阵列在ku波段及以上卫星系统中的可行性。

{"title":"Design and analysis of a continuously linear polarization agile antenna using microwave liquid crystal technology","authors":"Onur Hamza Karabey","doi":"10.1016/j.jestch.2025.102201","DOIUrl":"10.1016/j.jestch.2025.102201","url":null,"abstract":"<div><div>This paper presents the design and analysis of a continuously linear polarization-agile antenna based on Microwave Liquid Crystal (LC) technology. The proposed concept is specifically developed for Ku-band (e.g. 12 GHz) satellite communication applications, where systems typically operate with linear polarization and maintaining polarization alignment, referred to as skew angle adjustment, is critical for reliable on-the-move operation and for avoiding link loss. The antenna system incorporates a tunable feed network comprising two directional couplers and LC-based continuously tunable phase shifters. These feed a dual-fed microstrip patch antenna, enabling full-range linear polarization control over <span><math><mrow><mo>±</mo><mn>18</mn><msup><mrow><mn>0</mn></mrow><mrow><mo>∘</mo></mrow></msup></mrow></math></span> by electronically adjusting the differential phase. The proposed tunable feed network is implemented using microwave LC technology to enable a fully passive architecture without requiring low-noise amplifiers. LC materials offer electrically tunable dielectric properties with inherently low dielectric loss, typically below 0.06 at frequencies beyond 10 GHz, and their loss tangent decreases with increasing frequency, making them particularly well-suited for Ku-band applications and beyond. In addition, LC materials benefit from a mature fabrication infrastructure originating from LC display manufacturing, allowing low-cost, large-area, and scalable production. Full-wave simulations at 12 GHz show that the system achieves a cross-polarization discrimination (XPD) better than 25 dB across the entire skew angle range. The LC-based phase shifters operate with only two bias voltages and provide 180° differential phase shift with an insertion loss below 2.3 dB, yielding a figure-of-merit of 80°/dB at 12 GHz. These results demonstrate the feasibility of the proposed design for low-loss, scalable, and reconfigurable antenna arrays in Ku-band and beyond satellite systems.</div></div>","PeriodicalId":48609,"journal":{"name":"Engineering Science and Technology-An International Journal-Jestech","volume":"71 ","pages":"Article 102201"},"PeriodicalIF":5.4,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145221095","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Preference-based deep reinforcement learning with automatic curriculum learning for map-free UGV navigation in factory-like environments 基于偏好的深度强化学习与自动课程学习在工厂环境下无地图UGV导航

IF 5.4 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Engineering Science and Technology-An International Journal-Jestech

Pub Date : 2025-10-01 Epub Date: 2025-08-12 DOI: 10.1016/j.jestch.2025.102147

Shunyu Tian , Changyun Wei , Shaojie Jian , Ze Ji

Autonomous navigation for unmanned ground vehicles in smart factory environments involves satisfying multiple, often conflicting objectives such as safety, efficiency, and motion smoothness. Traditional reinforcement learning approaches typically rely on fixed, manually weighted reward functions to encode these objectives. However, such static formulations struggle to generalize across varying user preferences and dynamic operational contexts common in real-world factory scenarios. Consequently, they require retraining for every new preference configuration, leading to inefficiency and limited practical deployment. To address this challenge, we propose a novel preference-based reinforcement learning framework that enables a single policy to dynamically adapt its behavior according to a user-defined preference vector that encodes trade-offs among multiple objectives. This allows the agent to modify its navigation strategy on-the-fly without additional retraining. To further improve training efficiency and learning stability, we incorporate automatic curriculum learning, which gradually increases the complexity of training tasks based on the agent’s performance, accelerating convergence and robustness. We validate our method in a simulated smart factory environment that reflects realistic navigation constraints. Experimental results demonstrate that our proposed approach ensures faster convergence during training and achieves up to a 93% navigation success rate in challenging factory-like environments compared to recent advances.

智能工厂环境中无人驾驶地面车辆的自主导航涉及满足多个经常相互冲突的目标，如安全、效率和运动平顺性。传统的强化学习方法通常依赖于固定的、手动加权的奖励函数来编码这些目标。然而，这种静态公式很难泛化到不同的用户偏好和真实工厂场景中常见的动态操作上下文。因此，它们需要对每个新的首选项配置进行再培训，从而导致效率低下和实际部署受限。为了解决这一挑战，我们提出了一种新的基于偏好的强化学习框架，该框架使单个策略能够根据用户定义的偏好向量动态调整其行为，该偏好向量编码多个目标之间的权衡。这使得智能体无需额外的再训练即可实时修改其导航策略。为了进一步提高训练效率和学习稳定性，我们结合了自动课程学习，根据智能体的性能逐渐增加训练任务的复杂性，加速收敛和鲁棒性。我们在模拟的智能工厂环境中验证了我们的方法，该环境反映了现实的导航约束。实验结果表明，与最近的进展相比，我们提出的方法确保了训练过程中的更快收敛，并且在具有挑战性的工厂环境中实现了高达93%的导航成功率。

{"title":"Preference-based deep reinforcement learning with automatic curriculum learning for map-free UGV navigation in factory-like environments","authors":"Shunyu Tian , Changyun Wei , Shaojie Jian , Ze Ji","doi":"10.1016/j.jestch.2025.102147","DOIUrl":"10.1016/j.jestch.2025.102147","url":null,"abstract":"<div><div>Autonomous navigation for unmanned ground vehicles in smart factory environments involves satisfying multiple, often conflicting objectives such as safety, efficiency, and motion smoothness. Traditional reinforcement learning approaches typically rely on fixed, manually weighted reward functions to encode these objectives. However, such static formulations struggle to generalize across varying user preferences and dynamic operational contexts common in real-world factory scenarios. Consequently, they require retraining for every new preference configuration, leading to inefficiency and limited practical deployment. To address this challenge, we propose a novel preference-based reinforcement learning framework that enables a single policy to dynamically adapt its behavior according to a user-defined preference vector that encodes trade-offs among multiple objectives. This allows the agent to modify its navigation strategy on-the-fly without additional retraining. To further improve training efficiency and learning stability, we incorporate automatic curriculum learning, which gradually increases the complexity of training tasks based on the agent’s performance, accelerating convergence and robustness. We validate our method in a simulated smart factory environment that reflects realistic navigation constraints. Experimental results demonstrate that our proposed approach ensures faster convergence during training and achieves up to a 93% navigation success rate in challenging factory-like environments compared to recent advances.</div></div>","PeriodicalId":48609,"journal":{"name":"Engineering Science and Technology-An International Journal-Jestech","volume":"70 ","pages":"Article 102147"},"PeriodicalIF":5.4,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144831332","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

QSAC: Quantum-assisted Secure Audio Communication using quantum entanglement, audio steganography, and classical encryption 使用量子纠缠、音频隐写术和经典加密的量子辅助安全音频通信

IF 5.4 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Engineering Science and Technology-An International Journal-Jestech

Pub Date : 2025-10-01 Epub Date: 2025-08-22 DOI: 10.1016/j.jestch.2025.102167

Md. Raisul Islam Rifat , Md. Mizanur Rahman , Md. Abdul Kader Nayon , Md Shawmoon Azad , M.R.C. Mahdy

The emergence of quantum computing poses a significant security threat to the current classical cryptography system since it has the potential to outperform the current classical computer in some specific tasks due to its unique principle of operation. This necessitates finding a method that is resistant to quantum computers to securely transfer information. This research addresses these challenges by proposing a novel method combining quantum key distribution (QKD), specifically the E91 protocol, with the classical encryption-authentication protocol ChaCha20-Poly1305, and concealing information within another message through steganography to securely transfer audio messages. A shared secret key is created between the communicating parties using E91 QKD, which exploits the stellar protection of quantum entanglement against eavesdropping. The shared key is hashed through the SHA–3 hash function to generate a fixed-length, high-entropy symmetric key. The audio message is hidden inside another audio signal through steganography. The steganographic signal is encrypted using ChaCha20-Poly1305 AEAD in order to provide another layer of obfuscation as well as a means to verify integrity. Through rigorous experiments, we validated the robustness of the proposed methodology in both classical and quantum attacks. The processing of secret audio signals of varying duration (00:01:32 to 00:01:36) exhibits consistent encryption results. The encrypted stego audios show high randomness, with an average entropy of 15.9984, an average correlation of

1.4627 \times 1 0^{- 5}

, an average UACI of 49.9977%, and an average NSCR of 99.9985%. We demonstrated the safety of the shared key using the CHSH inequality test, where in the presence of an eavesdropper, the CHSH value is much less than

2 \sqrt{2}

. In addition, the integrity of the secret audio is also validated through the verification of the authenticator tag generated during the encryption process. Our research offers a novel framework for secure audio transmission, combining classical encryption and authentication methods with QKD to enhance confidentiality, integrity, and resilience against eavesdropping and tampering, ensuring robust end-to-end security.

量子计算的出现对当前的经典密码系统构成了重大的安全威胁，因为量子计算由于其独特的工作原理，在某些特定任务中有可能超越当前的经典计算机。这就需要找到一种能够抵抗量子计算机的方法来安全地传输信息。本研究通过提出一种将量子密钥分发（QKD）（特别是E91协议）与经典加密认证协议ChaCha20-Poly1305相结合的新方法来解决这些挑战，并通过隐写术将信息隐藏在另一条消息中以安全地传输音频消息。使用E91 QKD在通信方之间创建共享密钥，它利用量子纠缠的恒星保护来防止窃听。通过SHA-3散列函数对共享密钥进行散列，以生成固定长度的高熵对称密钥。音频信息通过隐写术隐藏在另一个音频信号中。隐写信号使用ChaCha20-Poly1305 AEAD进行加密，以便提供另一层混淆以及验证完整性的手段。通过严格的实验，我们验证了所提出的方法在经典和量子攻击中的鲁棒性。对不同持续时间（00:01:32至00:01:36）的秘密音频信号的处理显示出一致的加密结果。加密后的stego音频具有较高的随机性，平均熵为15.9984，平均相关系数为1.4627×10−5，平均UACI为49.9977%，平均NSCR为99.9985%。我们使用CHSH不等式测试证明了共享密钥的安全性，在存在窃听者的情况下，CHSH值远小于22。此外，还通过对加密过程中生成的验证者标签的验证来验证秘密音频的完整性。我们的研究为安全音频传输提供了一个新的框架，将经典的加密和认证方法与QKD相结合，以增强机密性、完整性和抗窃听和篡改的弹性，确保强大的端到端安全性。

{"title":"QSAC: Quantum-assisted Secure Audio Communication using quantum entanglement, audio steganography, and classical encryption","authors":"Md. Raisul Islam Rifat , Md. Mizanur Rahman , Md. Abdul Kader Nayon , Md Shawmoon Azad , M.R.C. Mahdy","doi":"10.1016/j.jestch.2025.102167","DOIUrl":"10.1016/j.jestch.2025.102167","url":null,"abstract":"<div><div>The emergence of quantum computing poses a significant security threat to the current classical cryptography system since it has the potential to outperform the current classical computer in some specific tasks due to its unique principle of operation. This necessitates finding a method that is resistant to quantum computers to securely transfer information. This research addresses these challenges by proposing a novel method combining quantum key distribution (QKD), specifically the E91 protocol, with the classical encryption-authentication protocol ChaCha20-Poly1305, and concealing information within another message through steganography to securely transfer audio messages. A shared secret key is created between the communicating parties using E91 QKD, which exploits the stellar protection of quantum entanglement against eavesdropping. The shared key is hashed through the SHA–3 hash function to generate a fixed-length, high-entropy symmetric key. The audio message is hidden inside another audio signal through steganography. The steganographic signal is encrypted using ChaCha20-Poly1305 AEAD in order to provide another layer of obfuscation as well as a means to verify integrity. Through rigorous experiments, we validated the robustness of the proposed methodology in both classical and quantum attacks. The processing of secret audio signals of varying duration (00:01:32 to 00:01:36) exhibits consistent encryption results. The encrypted stego audios show high randomness, with an average entropy of 15.9984, an average correlation of <span><math><mrow><mn>1</mn><mo>.</mo><mn>4627</mn><mo>×</mo><mn>1</mn><msup><mrow><mn>0</mn></mrow><mrow><mo>−</mo><mn>5</mn></mrow></msup></mrow></math></span>, an average UACI of 49.9977%, and an average NSCR of 99.9985%. We demonstrated the safety of the shared key using the CHSH inequality test, where in the presence of an eavesdropper, the CHSH value is much less than <span><math><mrow><mn>2</mn><msqrt><mrow><mn>2</mn></mrow></msqrt></mrow></math></span>. In addition, the integrity of the secret audio is also validated through the verification of the authenticator tag generated during the encryption process. Our research offers a novel framework for secure audio transmission, combining classical encryption and authentication methods with QKD to enhance confidentiality, integrity, and resilience against eavesdropping and tampering, ensuring robust end-to-end security.</div></div>","PeriodicalId":48609,"journal":{"name":"Engineering Science and Technology-An International Journal-Jestech","volume":"70 ","pages":"Article 102167"},"PeriodicalIF":5.4,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144886129","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Neural network-based adaptive super-twisting sliding mode control for hybrid magnetic levitation system with external disturbance 基于神经网络的混合磁悬浮系统自适应超扭滑模控制

IF 5.4 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Engineering Science and Technology-An International Journal-Jestech

Pub Date : 2025-10-01 Epub Date: 2025-08-30 DOI: 10.1016/j.jestch.2025.102177

Menglei Zhang , Liwei Zhang , Lu Shen

Hybrid magnetic levitation system compared with pure electromagnetic levitation system, because of the introduction of permanent magnets, the hybrid levitation system is more sensitive to parameter uncertainty and external disturbance. The traditional control strategy is unable to meet the operational requirements of the hybrid levitation system. To solve this problem, an adaptive super-twisting sliding mode controller based on a neural network is designed to address unknown parameters and external disturbance. Firstly, this study analyses the impact of the introduction of permanent magnets on the controllability and safety of the system. An adaptive sliding mode controller is designed. To address the issue of model parameter uncertainty, a neural network is employed to fit the unknown quantities. Then, to further solve the chattering issue on the platform in the face of disturbances, an adaptive super-twisting controller was developed and designed based on the neural network controller. Finally, related experimental verification was carried out on a hybrid levitation experimental platform. The experimental results indicate that the proposed control strategy is able to maintain stable levitation of the platform even under external disturbance and ensure the airgap tracking and levitation safety of the system.

混合磁悬浮系统与纯电磁悬浮系统相比，由于永磁体的引入，混合磁悬浮系统对参数的不确定性和外界干扰更加敏感。传统的控制策略已不能满足混合悬浮系统的运行要求。为了解决这一问题，设计了一种基于神经网络的自适应超扭滑模控制器来处理未知参数和外部干扰。首先，本文分析了永磁体的引入对系统可控性和安全性的影响。设计了一种自适应滑模控制器。为了解决模型参数的不确定性问题，采用神经网络对未知量进行拟合。然后，为了进一步解决平台在面对扰动时的抖振问题，在神经网络控制器的基础上开发设计了自适应超扭控制器。最后，在混合悬浮实验平台上进行了相关实验验证。实验结果表明，所提出的控制策略能够在外界干扰下保持平台稳定悬浮，保证系统的气隙跟踪和悬浮安全。

{"title":"Neural network-based adaptive super-twisting sliding mode control for hybrid magnetic levitation system with external disturbance","authors":"Menglei Zhang , Liwei Zhang , Lu Shen","doi":"10.1016/j.jestch.2025.102177","DOIUrl":"10.1016/j.jestch.2025.102177","url":null,"abstract":"<div><div>Hybrid magnetic levitation system compared with pure electromagnetic levitation system, because of the introduction of permanent magnets, the hybrid levitation system is more sensitive to parameter uncertainty and external disturbance. The traditional control strategy is unable to meet the operational requirements of the hybrid levitation system. To solve this problem, an adaptive super-twisting sliding mode controller based on a neural network is designed to address unknown parameters and external disturbance. Firstly, this study analyses the impact of the introduction of permanent magnets on the controllability and safety of the system. An adaptive sliding mode controller is designed. To address the issue of model parameter uncertainty, a neural network is employed to fit the unknown quantities. Then, to further solve the chattering issue on the platform in the face of disturbances, an adaptive super-twisting controller was developed and designed based on the neural network controller. Finally, related experimental verification was carried out on a hybrid levitation experimental platform. The experimental results indicate that the proposed control strategy is able to maintain stable levitation of the platform even under external disturbance and ensure the airgap tracking and levitation safety of the system.</div></div>","PeriodicalId":48609,"journal":{"name":"Engineering Science and Technology-An International Journal-Jestech","volume":"70 ","pages":"Article 102177"},"PeriodicalIF":5.4,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144916434","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Transformer differential protection with wavelet transform and difference function 采用小波变换和差分功能的变压器差动保护

IF 5.1 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Engineering Science and Technology-An International Journal-Jestech

Pub Date : 2025-10-01 Epub Date: 2025-07-27 DOI: 10.1016/j.jestch.2025.102144

Merve Oztekin , Serap Karagol , Okan Ozgonenel

The Transformer Differential Protection (TDP) algorithm instantly compares the target transformer’s terminal currents for each phase. The differential current is not expected to appear during regular operation. However, nonlinear characteristics of the core material, such as the hysteresis curve, result in significant variation in the differential current, known as magnetizing inrush current. This inrush current lasts for a while before disappearing, causing significant variation in the differential current. TDP algorithm is supposed to remain silent during this transient time (selectivity). In addition, one of the most difficult tasks for protection systems is detecting inter-turn faults in their early stages. This fault type typically begins at low levels due to moisture, high temperature, and so on, and gradually spreads to other turns. It is vital to detect inter-turn faults before they expand more than 10% of total windings [RW-2-1]. This paper presents a transformer differential protection algorithm that distinguishes between inter-turn, low-level internal faults, and inrush current. Maximum Overlapped Discrete Wavelet Transform (MODWT) energy and difference function are used for feature extraction, and the traditional 87T TDP method has been updated. Performance is evaluated using data collected from a laboratory-based experimental rig. The results demonstrate that the suggested approach performs very well in a range of low-level, inter-turn fault, and transient scenarios, including internal fault, inrush current, and sympathetic inrush current. These results are confirmed by the identified indices for Accuracy (AC), Dependability (DP), and Selectivity (SE).

变压器差动保护（TDP）算法可即时比较目标变压器各相的终端电流。在正常运行期间，不希望出现差动电流。然而，磁芯材料的非线性特性，如磁滞曲线，会导致差动电流（即磁化涌流）的显著变化。这种涌流在消失之前持续一段时间，导致差分电流的显著变化。TDP算法应该在这个短暂的时间内保持沉默（选择性）。此外，保护系统最困难的任务之一是在早期阶段检测匝间故障。这种故障类型通常由于潮湿、高温等原因从低电平开始，并逐渐扩散到其他匝数。在匝间故障扩展到总绕组的10%以上之前进行检测是至关重要的[RW-2-1]。提出了一种变压器差动保护算法，该算法能够区分匝间故障、低电平内部故障和涌流。利用最大重叠离散小波变换（MODWT）能量和差分函数进行特征提取，对传统的87T TDP方法进行了更新。使用从基于实验室的实验装置收集的数据来评估性能。结果表明，所提出的方法在低电平、匝间故障和瞬态场景（包括内部故障、涌流和共感涌流）中表现良好。这些结果通过确定的准确性（AC），可靠性（DP）和选择性（SE）指标得到证实。

{"title":"Transformer differential protection with wavelet transform and difference function","authors":"Merve Oztekin , Serap Karagol , Okan Ozgonenel","doi":"10.1016/j.jestch.2025.102144","DOIUrl":"10.1016/j.jestch.2025.102144","url":null,"abstract":"<div><div>The Transformer Differential Protection (TDP) algorithm instantly compares the target transformer’s terminal currents for each phase. The differential current is not expected to appear during regular operation. However, nonlinear characteristics of the core material, such as the hysteresis curve, result in significant variation in the differential current, known as magnetizing inrush current. This inrush current lasts for a while before disappearing, causing significant variation in the differential current. TDP algorithm is supposed to remain silent during this transient time (selectivity). In addition, one of the most difficult tasks for protection systems is detecting inter-turn faults in their early stages. This fault type typically begins at low levels due to moisture, high temperature, and so on, and gradually spreads to other turns. It is vital to detect inter-turn faults before they expand more than 10% of total windings [RW-2-1]. This paper presents a transformer differential protection algorithm that distinguishes between inter-turn, low-level internal faults, and inrush current. Maximum Overlapped Discrete Wavelet Transform (MODWT) energy and difference function are used for feature extraction, and the traditional 87T TDP method has been updated. Performance is evaluated using data collected from a laboratory-based experimental rig. The results demonstrate that the suggested approach performs very well in a range of low-level, inter-turn fault, and transient scenarios, including internal fault, inrush current, and sympathetic inrush current. These results are confirmed by the identified indices for Accuracy (AC), Dependability (DP), and Selectivity (SE).</div></div>","PeriodicalId":48609,"journal":{"name":"Engineering Science and Technology-An International Journal-Jestech","volume":"70 ","pages":"Article 102144"},"PeriodicalIF":5.1,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144712896","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A comprehensive review on YOLO versions for object detection 全面审查的YOLO版本的目标检测

IF 5.4 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Engineering Science and Technology-An International Journal-Jestech

Pub Date : 2025-10-01 Epub Date: 2025-08-05 DOI: 10.1016/j.jestch.2025.102161

Ayşe Aybilge Murat , Mustafa Servet Kiran

The need for methods used for object detection has gained increasing momentum in recent years. Starting with traditional image processing techniques, this process has been facilitated by the addition of deep learning. Object detection is currently used in areas such as autonomous vehicles, disease diagnosis, robotic vision and industry. The types of systems that are predicted to be needed more and more in the age of developing technology are also increasing. In particular, YOLO (You Only Look Once), which is mostly preferred in real-time object detection, is preferred because it achieves high accuracy in a short time. This paper analyses the main versions of the YOLO algorithm since its first release. The paper systematically analyses the architectural differences between the versions of the YOLO algorithm, the strengths and weaknesses of the models and their contribution to performance. At the same time, in most of the previous studies on YOLO, a comprehensive comparison of the YOLOv9-v11 models is not presented and new architectural features are not evaluated. This review provides an in-depth analysis of the main versions from YOLOv1 to YOLOv11, including recent innovations such as NMS-free, Oriented Bounding Boxes (OBB), GELAN and PGI. This work is intended to be a useful guide for researchers and developers interested in the field.

近年来，对用于目标检测的方法的需求日益增长。从传统的图像处理技术开始，深度学习的加入促进了这一过程。物体检测目前用于自动驾驶汽车、疾病诊断、机器人视觉和工业等领域。在技术发展的时代，预计需要越来越多的系统类型也在增加。特别是在实时目标检测中最受青睐的YOLO (You Only Look Once)，因为它可以在短时间内实现较高的精度。本文分析了YOLO算法自首次发布以来的主要版本。本文系统地分析了不同版本的YOLO算法在体系结构上的差异、模型的优缺点及其对性能的贡献。同时，在以往的大多数关于YOLO的研究中，并没有对YOLOv9-v11模型进行全面的比较，也没有对新的架构特性进行评估。本文对从YOLOv1到YOLOv11的主要版本进行了深入的分析，包括最近的创新，如NMS-free，定向边界盒（OBB）， GELAN和PGI。这项工作旨在为对该领域感兴趣的研究人员和开发人员提供有用的指导。

{"title":"A comprehensive review on YOLO versions for object detection","authors":"Ayşe Aybilge Murat , Mustafa Servet Kiran","doi":"10.1016/j.jestch.2025.102161","DOIUrl":"10.1016/j.jestch.2025.102161","url":null,"abstract":"<div><div>The need for methods used for object detection has gained increasing momentum in recent years. Starting with traditional image processing techniques, this process has been facilitated by the addition of deep learning. Object detection is currently used in areas such as autonomous vehicles, disease diagnosis, robotic vision and industry. The types of systems that are predicted to be needed more and more in the age of developing technology are also increasing. In particular, YOLO (You Only Look Once), which is mostly preferred in real-time object detection, is preferred because it achieves high accuracy in a short time. This paper analyses the main versions of the YOLO algorithm since its first release. The paper systematically analyses the architectural differences between the versions of the YOLO algorithm, the strengths and weaknesses of the models and their contribution to performance. At the same time, in most of the previous studies on YOLO, a comprehensive comparison of the YOLOv9-v11 models is not presented and new architectural features are not evaluated. This review provides an in-depth analysis of the main versions from YOLOv1 to YOLOv11, including recent innovations such as NMS-free, Oriented Bounding Boxes (OBB), GELAN and PGI. This work is intended to be a useful guide for researchers and developers interested in the field.</div></div>","PeriodicalId":48609,"journal":{"name":"Engineering Science and Technology-An International Journal-Jestech","volume":"70 ","pages":"Article 102161"},"PeriodicalIF":5.4,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144771218","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

PV-integrated coordinated control for enhanced grid performance in next-gen EV charging systems 下一代电动汽车充电系统中增强电网性能的pv集成协调控制

IF 5.4 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Engineering Science and Technology-An International Journal-Jestech

Pub Date : 2025-10-01 Epub Date: 2025-08-26 DOI: 10.1016/j.jestch.2025.102176

Umashankar Subramaniam , S Saravanan , K.R.M Vijayachandrakala , Sivakumar Selvam

This study aims to contribute to the United Nations’ Sustainable Development Goal (SDG) 7, which promotes affordable and clean energy by investigating the potential of solar photovoltaic systems (PVS) integration into EV charging infrastructure in evolving power networks for mutual benefits. The objective is to develop a coordinated control strategy for PVS-integrated EV charging stations that ensures seamless grid interaction with enhanced power quality. It focuses on modelling and coordinated control to achieve stable grid integration. Two distinct EV charging stations are considered in this work. Station 1 employs a conventional unidirectional power flow model, drawing power from the three-phase grid’s point of common coupling (PCC). The second station utilizes a PVS-based charging infrastructure connected to the PCC via a three-phase inverter. It facilitates power exchange with the distribution grid and charging stations, which addresses the reliability concerns of the PVS-based charging infrastructure. Coordinated control of the overall system is achieved through the dual Second-Order Generalized Integral (SOGI) based voltage and load current processing loops. This research ensures that the proposed dual SOGI-based controller maintains a unity power factor, reduces total harmonic distortion to below 3%, and eliminates the need for external filters meeting high grid power quality by ensuring power transfer between the grid and any charging stations. The PVS system mitigates harmonics and fulfills the reactive power demands of station 1 and local loads, obviating the necessity for separate filters and compensators. The developed control algorithm was tested on a hardware prototype under various loads and PV side conditions, demonstrating effective harmonics mitigation, reactive power compensation, and grid current balancing. The extensive hardware analysis conducted in steady state and dynamic operating modes confirms that the presented system improves voltage stability by over 20% and cuts network losses by more than 25%, establishing its effectiveness for next-generation sustainable EV infrastructure.

本研究旨在为联合国可持续发展目标（SDG） 7做出贡献，该目标通过调查太阳能光伏系统（pv）与电动汽车充电基础设施在不断发展的电力网络中的潜力，促进可负担和清洁能源的发展，以实现互利共赢。目标是为pvs集成的电动汽车充电站制定协调控制策略，以确保与电网的无缝交互并提高电力质量。它着重于建模和协调控制，以实现稳定的网格集成。本文考虑了两种不同的电动汽车充电站。电站1采用传统的单向潮流模型，从三相电网的共耦合点（PCC）获取电力。第二个充电站利用基于pv的充电基础设施，通过三相逆变器连接到PCC。它促进了与配电网和充电站的电力交换，解决了基于pv的充电基础设施的可靠性问题。通过基于二阶广义积分（SOGI）的电压和负载电流处理回路实现对整个系统的协调控制。本研究确保所提出的基于双sogi的控制器保持统一的功率因数，将总谐波失真降低到3%以下，并通过确保电网与任何充电站之间的电力传输，消除了满足高电网电能质量的外部滤波器的需要。PVS系统减轻了谐波，满足了站1和本地负载的无功需求，避免了单独的滤波器和补偿器的需要。开发的控制算法在各种负载和光伏侧条件下的硬件样机上进行了测试，证明了有效的谐波缓解、无功补偿和电网电流平衡。在稳态和动态运行模式下进行的大量硬件分析证实，该系统将电压稳定性提高了20%以上，将网络损耗降低了25%以上，为下一代可持续电动汽车基础设施奠定了基础。

{"title":"PV-integrated coordinated control for enhanced grid performance in next-gen EV charging systems","authors":"Umashankar Subramaniam , S Saravanan , K.R.M Vijayachandrakala , Sivakumar Selvam","doi":"10.1016/j.jestch.2025.102176","DOIUrl":"10.1016/j.jestch.2025.102176","url":null,"abstract":"<div><div>This study aims to contribute to the United Nations’ Sustainable Development Goal (SDG) 7, which promotes affordable and clean energy by investigating the potential of solar photovoltaic systems (PVS) integration into EV charging infrastructure in evolving power networks for mutual benefits. The objective is to develop a coordinated control strategy for PVS-integrated EV charging stations that ensures seamless grid interaction with enhanced power quality. It focuses on modelling and coordinated control to achieve stable grid integration. Two distinct EV charging stations are considered in this work. Station 1 employs a conventional unidirectional power flow model, drawing power from the three-phase grid’s point of common coupling (PCC). The second station utilizes a PVS-based charging infrastructure connected to the PCC via a three-phase inverter. It facilitates power exchange with the distribution grid and charging stations, which addresses the reliability concerns of the PVS-based charging infrastructure. Coordinated control of the overall system is achieved through the dual Second-Order Generalized Integral (SOGI) based voltage and load current processing loops. This research ensures that the proposed dual SOGI-based controller maintains a unity power factor, reduces total harmonic distortion to below 3%, and eliminates the need for external filters meeting high grid power quality by ensuring power transfer between the grid and any charging stations. The PVS system mitigates harmonics and fulfills the reactive power demands of station 1 and local loads, obviating the necessity for separate filters and compensators. The developed control algorithm was tested on a hardware prototype under various loads and PV side conditions, demonstrating effective harmonics mitigation, reactive power compensation, and grid current balancing. The extensive hardware analysis conducted in steady state and dynamic operating modes confirms that the presented system improves voltage stability by over 20% and cuts network losses by more than 25%, establishing its effectiveness for next-generation sustainable EV infrastructure.</div></div>","PeriodicalId":48609,"journal":{"name":"Engineering Science and Technology-An International Journal-Jestech","volume":"70 ","pages":"Article 102176"},"PeriodicalIF":5.4,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144896356","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Retrofitting of seismically deficient RC bridge pier using HSP concrete jacket and FRP rebars 用HSP混凝土夹套和FRP筋加固地震缺陷RC桥墩

IF 5.4 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Engineering Science and Technology-An International Journal-Jestech

Pub Date : 2025-10-01 Epub Date: 2025-07-30 DOI: 10.1016/j.jestch.2025.102146

Muhammad Waleed Ahmad, Muhammad Rizwan, Muhammad Fiaz Tahir

Bridges in high seismic zones are prone to failure due to inadequate seismic design or cumulative damage from repeated earthquakes, compromising structural integrity and functionality This study investigates the effectiveness of retrofitting seismically deficient reinforced concrete (RC) bridge piers using high-performance strength concrete (HPSC) jackets and Fiber-reinforced polymer (FRP) rebars applied at the plastic hinge zone. Two 1:4 scaled RC pier specimens with different transverse reinforcement arrangements were tested under constant axial load and quasi-static cyclic lateral loading to induce damage. The damaged piers were then retrofitted and retested to assess seismic performance improvements. Key parameters such as load-carrying capacity, hysteresis behaviour, energy dissipation, and stiffness degradation were evaluated. The retrofitted specimens showed substantial restoration of strength and stiffness and exhibited enhanced energy dissipation compared to both the original and CFRP-wrapped specimens. Additionally, a finite element model was developed in SeismoStruct (2024), showing good agreement with experimental results and validating the proposed retrofitting approach. The study concludes that the HPSC-FRP retrofit method significantly enhances the seismic resilience of damaged bridge piers and presents a viable solution for extending the service life of aging bridge infrastructure.

高震区的桥梁容易因抗震设计不足或反复地震造成的累积损伤而失效，从而影响结构完整性和功能。本研究调查了在塑性铰区使用高性能强度混凝土（HPSC）护套和纤维增强聚合物（FRP）钢筋对地震缺陷钢筋混凝土（RC）桥墩进行改造的有效性。在恒定轴向荷载和准静力循环横向荷载作用下，对2个不同横向配筋布置的1:4比例钢筋混凝土桥墩试件进行了损伤试验。然后对受损的桥墩进行改造和重新测试，以评估抗震性能的改善。关键参数，如承载能力，迟滞行为，能量耗散和刚度退化进行了评估。与原始和cfrp包裹的样本相比，改造后的样本显示出大量的强度和刚度恢复，并表现出增强的能量耗散。此外，在SeismoStruct（2024）中开发了一个有限元模型，与实验结果吻合良好，验证了所提出的改造方法。研究表明，HPSC-FRP加固方法显著提高了受损桥墩的抗震能力，为延长老化桥梁基础设施的使用寿命提供了可行的解决方案。

{"title":"Retrofitting of seismically deficient RC bridge pier using HSP concrete jacket and FRP rebars","authors":"Muhammad Waleed Ahmad, Muhammad Rizwan, Muhammad Fiaz Tahir","doi":"10.1016/j.jestch.2025.102146","DOIUrl":"10.1016/j.jestch.2025.102146","url":null,"abstract":"<div><div>Bridges in high seismic zones are prone to failure due to inadequate seismic design or cumulative damage from repeated earthquakes, compromising structural integrity and functionality This study investigates the effectiveness of retrofitting seismically deficient reinforced concrete (RC) bridge piers using high-performance strength concrete (HPSC) jackets and Fiber-reinforced polymer (FRP) rebars applied at the plastic hinge zone. Two 1:4 scaled RC pier specimens with different transverse reinforcement arrangements were tested under constant axial load and quasi-static cyclic lateral loading to induce damage. The damaged piers were then retrofitted and retested to assess seismic performance improvements. Key parameters such as load-carrying capacity, hysteresis behaviour, energy dissipation, and stiffness degradation were evaluated. The retrofitted specimens showed substantial restoration of strength and stiffness and exhibited enhanced energy dissipation compared to both the original and CFRP-wrapped specimens. Additionally, a finite element model was developed in SeismoStruct (2024), showing good agreement with experimental results and validating the proposed retrofitting approach. The study concludes that the HPSC-FRP retrofit method significantly enhances the seismic resilience of damaged bridge piers and presents a viable solution for extending the service life of aging bridge infrastructure.</div></div>","PeriodicalId":48609,"journal":{"name":"Engineering Science and Technology-An International Journal-Jestech","volume":"70 ","pages":"Article 102146"},"PeriodicalIF":5.4,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144723025","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Transparent and reliable construction cost prediction using advanced machine learning and explainable AI 利用先进的机器学习和可解释的人工智能进行透明可靠的建筑成本预测

IF 5.4 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

Engineering Science and Technology-An International Journal-Jestech

Pub Date : 2025-10-01 Epub Date: 2025-07-30 DOI: 10.1016/j.jestch.2025.102159

Lifei Chen , Changyong Xu , Wei Hong Lim , Abhishek Sharma , Sew Sun Tiang , Kim Soon Chong , El-Sayed M. El-kenawy , Amel Ali Alhussan , Marwa M. Eid , Doaa Sami Khafaga

Accurate construction cost prediction is vital for project management, influencing budgeting, resource allocation, and overall success. This study proposes a comprehensive framework that combines machine learning models, uncertainty quantification through Confidence Intervals, and explainable AI techniques using SHAP (SHapley Additive exPlanations) to enhance transparency and decision-making. Ten machine learning models, including Ridge Regression, Lasso Regression, Elastic Net, K-Nearest Neighbor Regression, and advanced ensemble methods such as XGBoost, CatBoost, and HistGradient Boosting, were evaluated on the RSMeans dataset. Among these, HistGradient Boosting achieved the best performance on the testing dataset. Beyond traditional metrics, Confidence Intervals quantified prediction reliability, and SHAP identified critical cost drivers like “Formwork” and “Tributary Area,” enabling interpretable and robust prediction. This study highlights the potential of machine learning models to revolutionize construction cost estimation by integrating predictive accuracy, uncertainty analysis, and explainability. The proposed framework supports resource efficiency and enables process innovation in cost management. It also contributes to the advancement of sustainable building practices, offering a strong foundation for future research and promoting the adoption of machine learning-based solutions with enhanced transparency and confidence.

准确的工程造价预测对项目管理至关重要，影响预算、资源分配和整体成功。本研究提出了一个综合框架，该框架结合了机器学习模型、通过置信区间进行的不确定性量化，以及使用SHapley加性解释（SHapley Additive exPlanations）的可解释人工智能技术，以提高透明度和决策。在RSMeans数据集上评估了10种机器学习模型，包括Ridge Regression、Lasso Regression、Elastic Net、K-Nearest Neighbor Regression和高级集成方法，如XGBoost、CatBoost和HistGradient Boosting。其中，HistGradient Boosting在测试数据集上的性能最好。除了传统的指标之外，置信区间量化了预测的可靠性，而SHAP确定了关键的成本驱动因素，如“模板”和“支流区域”，从而实现了可解释和可靠的预测。这项研究强调了机器学习模型的潜力，通过整合预测准确性、不确定性分析和可解释性来彻底改变建筑成本估算。建议的框架支持资源效率，并使成本管理中的流程创新成为可能。它还有助于推进可持续建筑实践，为未来的研究提供坚实的基础，并促进采用基于机器学习的解决方案，提高透明度和信心。

{"title":"Transparent and reliable construction cost prediction using advanced machine learning and explainable AI","authors":"Lifei Chen , Changyong Xu , Wei Hong Lim , Abhishek Sharma , Sew Sun Tiang , Kim Soon Chong , El-Sayed M. El-kenawy , Amel Ali Alhussan , Marwa M. Eid , Doaa Sami Khafaga","doi":"10.1016/j.jestch.2025.102159","DOIUrl":"10.1016/j.jestch.2025.102159","url":null,"abstract":"<div><div>Accurate construction cost prediction is vital for project management, influencing budgeting, resource allocation, and overall success. This study proposes a comprehensive framework that combines machine learning models, uncertainty quantification through Confidence Intervals, and explainable AI techniques using SHAP (SHapley Additive exPlanations) to enhance transparency and decision-making. Ten machine learning models, including Ridge Regression, Lasso Regression, Elastic Net, K-Nearest Neighbor Regression, and advanced ensemble methods such as XGBoost, CatBoost, and HistGradient Boosting, were evaluated on the RSMeans dataset. Among these, HistGradient Boosting achieved the best performance on the testing dataset. Beyond traditional metrics, Confidence Intervals quantified prediction reliability, and SHAP identified critical cost drivers like “Formwork” and “Tributary Area,” enabling interpretable and robust prediction. This study highlights the potential of machine learning models to revolutionize construction cost estimation by integrating predictive accuracy, uncertainty analysis, and explainability. The proposed framework supports resource efficiency and enables process innovation in cost management. It also contributes to the advancement of sustainable building practices, offering a strong foundation for future research and promoting the adoption of machine learning-based solutions with enhanced transparency and confidence.</div></div>","PeriodicalId":48609,"journal":{"name":"Engineering Science and Technology-An International Journal-Jestech","volume":"70 ","pages":"Article 102159"},"PeriodicalIF":5.4,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144738636","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0