Abstract Recently, intelligent fault diagnosis methods have been employed in the condition monitoring of rotating machinery. Among them, graph neural networks are emerging as a new feature extraction tool that can mine the relationship characteristics between samples. However, many existing graph construction methods suffer from structural redundancy or missing node relationships, thus limiting the diagnosis accuracy of the models in practice. In this paper, an adaptive adjustment k-nearest neighbor graph-driven dynamic-weighted graph attention network (AAKNN-DWGAT) is proposed to address this problem. First, time-domain signals are transformed into frequency-domain features by using fast Fourier transformation. Subsequently, a frequency similarity evaluation method based on dynamic frequency warping is proposed, which enables the conversion of distance measurements into a frequency similarity matrix (FSM). Then, an adaptive edge construction operation is conducted on the basis of FSM, whereby the effective domain is captured for each node using an adaptive edge adjustment method, generating an AAKNN graph (AAKNNG). Next, the constructed AAKNNG is fed into a dynamic-weighted graph attention network (DWGAT) to extract the fault features of nodes layer by layer. In particular, the proposed DWGAT employs a dynamic-weighted strategy that can update the edge weight periodically using high-level output features, thereby eliminating the adverse impacts caused by noisy signals. Finally, the model outputs fault diagnosis results through a softmax classifier. Two case studies verified the effectiveness and the superiority of the proposed method compared with other graph neural networks and graph construction methods.
Farm mechanization in Bangladesh is exponentially growing which has led to timeliness of farm operation and increased crop productivity and farm income. For shifting of farm labour to non-farm sectors, sustainable crop production, efficient management of farm inputs, doubling the productivity and enhancing 4AR (Four Agricultural Revolution) dictating the use of appropriate and quality farm machinery. Different types of local and imported farm machinery are used in crop production. There was no legal authority functioning for testing and certification of these farm machinery in Bangladesh since 1989 to 2024. So, the farmers are not always getting quality and standard farm machinery although they have to pay good price to purchase machines. The farm machinery manufacturers are also suffering from proper marketing of machines due to unavailable of legal testing and certification facilities. Testing and certification of farm machinery are crucially essential to assess their functional suitability and performance reliability so that it will help farmers and other users in determining the comparative performance of machines available in market. Earlier, there was a Farm Machinery Testing and Certification System in Bangladesh but it was abolished in 1989 for rapid expansion of farm mechanization. Some experts opined that testing and certification system may restrict the abandoned uses of different farm machinery which may slow down the rapid speed of farm mechanization. So, the policy makers are confused about introducing the testing and standardization of farm machinery in Bangladesh, although there is a provision for Farm Machinery Testing and Certification System in National Agricultural Mechanization Policy 2020 and National Agricultural Mechanization Action Plan 2023. Recently, (February, 2025) National Standardization Committee (NSC) and Technical Sub-committee (TSC) have been formed. This paper explores the past and current states of farm machinery testing and standardization in Bangladesh, discussing existing challenges and potential solutions.
Fernando Martinez-Gil, Christopher Sansom, Aránzazu Fernández-García
et al.
This review explores advanced maintenance techniques aimed at improving solar energy production efficiency. The study analyzes the rapid growth of solar energy and the challenges posed by environmental factors such as soiling, harsh climate conditions and hotspots, which reduce photovoltaic (PV) and concentrated solar power (CSP) system performance. Predictive models for solar energy generation and soiling detection, including artificial intelligence (AI) and machine learning (ML) algorithms and Internet of Things (IoT), are discussed as means for optimizing energy production and reducing maintenance costs. It is also emphasized the role of Unmanned Aerial Vehicles (UAVs) to capture images for fault detection and failure prediction, enhancing maintenance accuracy and minimizing downtime. The study concludes by analyzing the role of these techniques to reduce water consumption in cleaning tasks, as well as solutions to increase the operational lifespan and performance of solar plants such as anti-soiling coatings, robotic cleaning systems and accurate predictive models.
AbstractIn order to study the cutting force and variation of anisotropic composite materials such as plastic vertical drainage board left in soft soil foundation by the cutter on the cutterhead of shield machine. Finite element model of orthogonal cutting theory based on macroscopic anisotropic composite materials was established by numerical simulation. Based on the above model, the finite element analysis software LS-DYNA was used to numerically simulate the cutting of plastic vertical drainage board by shield machine cutter in a project in Singapore. Meanwhile, the single cutter test device and triaxial force sensor were used to build a test platform to focus on the cutting ability test of the cutting tool on the plastic vertical drainage board under the cutting conditions of 0° and 90°. The validity and rationality of the simulation model are verified by comparing the simulated cutting force value with the experimental value. The results show that the cutting process of the plastic vertical drainage board is from the notch generation to the notch expansion and then to the final breaking process. So it can be concluded that the sharp-edge cutter is more suitable for the cutting of the plastic vertical drainage board. In the actual construction, ensuring the wear resistance and impact resistance of the sharp-edge cutter is the key to improve the service life. Sponsored by Shanghai Rising-Star Program (22QB1401900).
DC microgrid systems based on hybrid photovoltaic-electrical energy storage systems for power generation are widely used in new energy cruise ships. For the problem of power supply voltage unbalance arising when a daily load power supply converter is connected to an unbalanced load or single-phase load, this paper proposes a dead-beat predictive voltage control strategy based on positive and negative sequence separation of output voltage. This strategy can realize control of positive and negative sequence components of the power supply converter voltage under the positive and negative sequence coordinate systems respectively, to ensure the three-phase voltage balance of the power supply converter, enhance the anti-interference ability of the system against unbalanced load and avoid additional power quality control devices. The simulation results show that, after the proposed control strategy is adopted, the voltage unbalance of the daily load power supply converter is reduced from 28% to less than 0.5%, which verifies the effectiveness of the proposed control algorithm .
Control engineering systems. Automatic machinery (General), Technology
Viscosity is one of the most important fundamental properties of fluids. However, accurate acquisition of viscosity for ionic liquids (ILs) remains a critical challenge. In this study, an approach integrating prior physical knowledge into the machine learning (ML) model was proposed to predict the viscosity reliably. The method was based on 16 quantum chemical descriptors determined from the first principles calculations and used as the input of the ML models to represent the size, structure, and interactions of the ILs. Three strategies based on the residuals of the COSMO-RS model were created as the output of ML, where the strategy directly using experimental data was also studied for comparison. The performance of six ML algorithms was compared in all strategies, and the CatBoost model was identified as the optimal one. The strategies employing the relative deviations were superior to that using the absolute deviation, and the relative ratio revealed the systematic prediction error of the COSMO-RS model. The CatBoost model based on the relative ratio achieved the highest prediction accuracy on the test set (R2 = 0.9999, MAE = 0.0325), reducing the average absolute relative deviation (AARD) in modeling from 52.45% to 1.54%. Features importance analysis indicated the average energy correction, solvation-free energy, and polarity moment were the key influencing the systematic deviation.
To address the problem of problematic spray design inside mining anchor-digging equipment, a switching seal using a permanent magnet eddy current drive is initially presented here. The layer model of the permanent magnet eddy current structure is established, the subdomain analysis model is introduced, the permanent magnet eddy current structure is divided into six regions along the axial direction, and the boundary equations are established at the interfaces of each region. The vector magnetic potential equations in each region are deduced, along with the electromagnetic torque and axial force equations. The computational results are compared and analyzed with the results of finite element simulation, verifying the accuracy of the theoretical model. The design of experiments is used to verify the feasibility of the switching seal using the permanent magnet eddy current structure.
When faced with challenges such as adapting to dynamic environments and handling ambiguous identification, indoor service robots encounter manifold difficulties. This paper aims to address this issue by proposing the design of a service robot equipped with precise small-object recognition, autonomous path planning, and obstacle-avoidance capabilities. We conducted in-depth research on the suitability of three SLAM algorithms (GMapping, Hector-SLAM, and Cartographer) in indoor environments and explored their performance disparities. Upon this foundation, we have elected to utilize the STM32F407VET6 and Nvidia Jetson Nano B01 as our processing controllers. For the program design on the STM32 side, we are employing the FreeRTOS operating system, while for the Jetson Nano side, we are employing ROS (Robot Operating System) for program design. The robot employs a differential drive chassis, enabling successful autonomous path planning and obstacle-avoidance maneuvers. Within indoor environments, we utilized the YOLOv3 algorithm for target detection, achieving precise target identification. Through a series of simulations and real-world experiments, we validated the performance and feasibility of the robot, including mapping, navigation, and target detection functionalities. Experimental results demonstrate the robot’s outstanding performance and accuracy in indoor environments, offering users efficient service and presenting new avenues and methodologies for the development of indoor service robots.
AbstractDroplet impingement on a wall is a fundamental scientific problem with wide engineering applications. When a droplet impacts the surface of an aircraft, it generates shock waves, airflow disturbances, and splashing phenomena. This not only has a negative impact on the aerodynamic performance and stability of the aircraft but also obstructs the field of view of optical sensors or causes distortion in optical devices. It can also damage the aircraft's structure, thus it’s vital to assess the droplet impact force for flight safety. However, droplets are often treated as rigid spheres for simplicity, but this does not reflect the real physical situation. In this paper, we utilized high-precision force sensors and high-speed imaging technology to experimentally investigate the impact dynamic of droplet impingement on a dry wall. The temporal evolution of force, the associated morphology changes and their relationship during collisions were analyzed systematically, we also elucidated the physical mechanisms underlying flow phenomenon. An unified and accurate mechanical model were established for droplet impingement, providing guidance for related engineering designs.
AbstractTo address the problem of high-pressure low-frequency hydraulic impact on the hydraulic buffer system mounted on self-propelled hydraulic trailer, a parametric simulation and design framework is provided for the widespread use of integrated accumulator systems in the low-speed heavy-load vehicle. A mathematical model of the accumulator was established for theoretical analysis, and numerical simulation were conducted on corresponding parameters using AMESim. Finally, a routine method was proposed to achieve feasible arrangement of accumulators. In this study, for the problem of high-pressure impact conditions caused by severe fluctuations, the system performance could be increased by arranging different combinations of accumulators with different parameter configurations. Taking a commercial model of heavy trailer as an actual case, a reasonable parameterized design was carried out for its comprehensive accumulator buffer system, and verification was conducted.
AbstractThe suspension cable structure has the advantages of saving steel, light weight, beautiful in shape, and has been widely used in long-span bridges. In recent years, the suspension cable structure has been used in the flexible photovoltaic supports of small and medium-sized spans. The existing piecewise catenary method and finite element method have the problems of initial value sensitivity and easy divergence due to the highly geometric nonlinearity of suspension cable structure, and a practical calculation method of piecewise linearization of the suspension cable is proposed in this paper. The suspension cable is divided into n (n ≥ 40) segments. First, the initial values of inclination angle and tension of the left end of the suspension cable are given roughly, and the position of the right end of the suspension cable can be obtained by piecewise recursive method according to the external forces. Based on the deviation between the calculated position and the target position of the right end of the cable, the values of the inclination angle and tension of the left endpoint of the suspension cable are modified, and the second iteration is carried out. In this way about 3–6 iterations, the calculated position of the right end of the cable can converge to the target position, and the exact geometry and internal force of the suspension cable can be obtained. The method can be used to calculate the internal force and deformation of suspension bridges under static loads, and the results accuracy fully meets the engineering design requirements.
Bearings are crucial components in mechanical equipment, and bearing failure is one of the reasons for mechanical equipment safety accidents. In this work, an optical fiber vibration sensor based on Sagnac interferometer (SI) is proposed and applied to bearing fault detection. The polarization-maintaining fiber (PMF) is spliced between two single-mode fibers (SMFs) to form a SMF-PMF-SMF (SPS) fiber structure, which is connected with a 3 dB coupler to form a SI based on SPS fiber structure. The mass block is fixed in the middle of the PMF. When the monitored surface or structure vibrates, the stress of PMF will change and the Sagnac interference spectrum will be shifted, so that the vibration can be measured. The fabrication technology of sensor based on 3D printing is studied, and the structural parameters of the sensor are optimized through experiments and theoretical analysis. A vibration detection system and a rolling bearing vibration test platform based on SI and fiber ring laser are built. The experimental results show that the relative error between the experimental and theoretical results of healthy and faulty bearings is less than 1.60%, and the fault detection of bearings is realized. Temperature cross-sensitivity and stability of the sensing system are analyzed. The sensor has the advantages of simple structure, easy fabrication, anti-interference, etc, and has wide application prospects in engineering, machinery, security and other fields.
Pattern transformation in a periodic porous structure has inspired multifarious mechanical metamaterials/metastructures due to the induced unusual negative Poisson's ratio behavior of macroscopic materials. Recently, it has been leveraged to architect a variety of designable and multifunctional structural members. Inspired by this design methodology, a novel porous cylindrical shell, which is perforated by a large number of staggered openings, is constructed and investigated meticulously. A stable, anti‐disturbed, and controllable waisted deformation of the architected cylindrical shell will be triggered under an axial compression. A stoma‐shaped biomimetic hole and graded distribution of initial openings are proposed to ensure that the holes distributed throughout the shell can be closed up concurrently while the closed states of holes can be flexibly programmed. To explore the applications of such shells, a handy cylindrical vessel is elaborately designed and its multiple functions including reagent release, underwater sampling, and flow control are exhibited by experiments. The results reflect that the designed vessel can be facilitated with many advantages such as uniform release, quick action, easy actuation, and repeated usage. Moreover, it also may open a new avenue for metamaterials in the fields of biomedical engineering, underwater detection, fluid machinery, etc.
In this work, a miniaturized, low-cost, low-power and high-sensitivity AlN-based micro-electro-mechanical system (MEMS) hydrophone is proposed for monitoring water pipeline leaks. The proposed MEMS Hydrophone consists of a piezoelectric micromachined ultrasonic transducer (PMUT) array, an acoustic matching layer and a pre-amplifier amplifier circuit. The array has 4 (2 × 2) PMUT elements with a first-order resonant frequency of 41.58 kHz. Due to impedance matching of the acoustic matching layer and the 40 dB gain of the pre-amplifier amplifier circuit, the packaged MEMS Hydrophone has a high sound pressure sensitivity of −170 ± 2 dB (re: 1 V/μPa). The performance with respect to detecting pipeline leaks and locating leak points is demonstrated on a 31 m stainless leaking pipeline platform. The standard deviation (STD) of the hydroacoustic signal and Monitoring Index Efficiency (MIE) are extracted as features of the pipeline leak. A random forest model is trained for accurately classifying the leak and no-leak cases using the above features, and the accuracy of the model is about 97.69%. The cross-correlation method is used to locate the leak point, and the localization relative error is about 10.84% for a small leak of 12 L/min.
Abstract Calcium ion batteries, though more sustainable than lithium‐ion batteries, still face significant challenges, including the lack of highly rechargeable electrodes. Prepared Prussian blue nanodisk electrodes can significantly extend the longevity of Ca‐based cells. The method involves precipitating 20‐nm‐thick nanodisks of Prussian green from Fe(NO3)3 and K3Fe(CN)6, followed by sodiation with NaI. The Prussian‐blue‐based cathode with polyacrylic acid/polyaniline binder delivers an initial discharging capacity of 77.6 mAh g−1 at 0.1 A g−1 and retains 91.0% capacity after 700 cycles. Polyvinyl fluoride is detrimental to the Ca‐based cell, as its coulombic efficiency decreases from 94.8% (120th cycle) to 86.4% (400th cycle). With the same cathode, the Ca‐based cell is much less sensitive to high current densities than the Na‐based cell. This can be because only half the amount of cations is required to move in Ca‐based systems compared to Na‐based systems; thus, the charge‐transfer resistance is noticeably reduced in Ca‐based systems.