Average Error Research Articles

A theoretical model for estimating the flexural stiffness of CHC joints

The CHC joint is a steel-concrete composite insert-type joint, which has been successfully implemented at the Shenzhen Pingxi Station in China, advantages in terms of construction efficiency and environmental sustainability. However, the design methodology for CHC joints remains unreported. This study begins by clarifying the moment-rotation relationship of CHC joints through experimental methods. Subsequently, a novel dual spring-rigid arm theoretical model to estimate the joint's bending stiffness is proposes. The model parameters are derived from a series of rigorous mechanical theoretical derivations, resulting in a concise explicit expression for the bending stiffness. Results indicate that the theoretical model can accurately estimate the bending stiffness of CHC joints during service, with an average error in the calculated results of just 2.1%. This innovative approach holds promise for broader application in joint design and analysis, offering valuable insights for future research and practical implementation.

Optimizing Irrigation and Nitrogen Application for Greenhouse Tomato Using the DSSAT–CROPGRO–Tomato Model

The aim of this study was to optimize water-saving and high-efficiency irrigation and nitrogen application scheduling for greenhouse tomato cultivation in North China. Using experimental data on water and nitrogen inputs, the DSSAT-GLUE parameter adjustment tool was employed to calibrate the genetic parameters of the DSSAT–CROPGRO–Tomato model. Simulations were conducted to assess greenhouse tomato growth, water use, and yield under varying water and nitrogen conditions. After calibration, the model showed average relative errors of 3.19% for the phenological stages, 3.33% for plant height, and 4.52% for yield dry weight, meeting accuracy standards. The results from the calibrated model indicated that increasing irrigation or nitrogen levels initially enhanced yield but led to diminishing returns beyond optimal ranges. The maximum tomato yield and water–nitrogen use efficiency were achieved with irrigation quotas between 320 and 340 mm and nitrogen applications between 360 and 400 kg·ha−1. These findings provide a guideline for efficient water and nitrogen management for greenhouse tomatoes under drip irrigation conditions.

Study on Grain Growth and Grain Boundary Evolution of 740H Nickel‐Based Alloys under Different Solution Treatment Conditions

This research examines the grain growth, microstructural characteristics, and grain boundary evolution of 740H nickel‐based alloy subjected to solution treatment following hot rolling. The results demonstrate that increasing the solution temperature and extending the holding time promote grain growth. A two‐stage grain growth model is developed, achieving a correlation coefficient of 0.979 and an average absolute relative error of 9.5%. The volume fraction of Σ3 grain boundaries initially increases and then decreases, peaking at 1150 °C under constant holding time. At a solution temperature of 1150 °C and a holding time of 60 min, the proportion of special grain boundaries in the sample reaches a maximum of 61.1%, with Σ3 grain boundaries comprising the majority at 59.8%. Their morphology progressively evolves into long, straight double Σ3 grain boundaries, and the distribution of grain boundary clusters becomes increasingly concentrated. Grain boundary engineering analysis reveals that during grain growth, twins gradually form and proliferate, dislocation recovery in the alloy is enhanced, and residual stress is progressively reduced. Under the conditions of 1150 °C solution temperature and 60 min holding time, grain boundary regulation performance is optimized, establishing this as the most effective heat treatment process.

An Improved Methodology to Locate Faults in Onshore Wind Farm Collector Systems

This paper explores the growing integration of Inverter-Based Resources (IBRs) into power systems and their effects on fault diagnosis strategies. Notably, the technical literature lacks assessments of the impacts and proposals for solutions for phasor-based fault location tasks, considering faults occurring within wind power plants, i.e., in their collector systems. In this context, this study evaluates the performance of six state-of-the-art phasor-based fault location methods, which are tested through simulations in a realistic wind farm modeled using PSCAD software. These simulations cover a wide range of fault scenarios, including variations in fault types, resistances, inception angles, locations, and wind farm generation levels. The proposed methodology, which combines the various fault location methods tailored to specific fault types, results in a substantial improvement, achieving an average fault location error of 1.89%, reflecting a 92% reduction in error compared to conventional methods. Additionally, the approach consistently maintains low fault location errors across collector busbars, regardless of circuit topology, highlighting its robustness, adaptability, and potential for widespread implementation in fault diagnosis systems within wind farms.

Small-area marks realize nanoscale lithography alignment by a spatial and frequency domain fusion neural network

For traditional moiré-based lithography alignment technology, which is widely used in proximity lithography systems, complex alignment marks with larger areas are employed to achieve high-precision misalignment detection. However, every inch of space on the wafer is extremely precious in practice, leaving minimal space for alignment marks. Therefore, employing small-area alignment marks in lithography systems will be a very challenging task with considerable potential in the future. The primary challenge is that existing frequency-based analytical algorithms struggle to achieve misalignment values with high-precision from moiré fringe images generated by small-area marks. To address this challenge, a spatial and frequency information fusion neural network (SFFN) is proposed for processing the moiré fringe images. With SFFN, the area of the alignment mark can be reduced by 2/3, and the average error of SFFN is less than 1 nm on the test dataset.

Photovoltaic power one-day and multistep-hourly AI predictions using node-by-node evolved binomial tree structures to form L-transformed PDE modules

Abstract Photovoltaic (PV) power is generated by two common types of solar components that are primarily affected by fluctuations and development in cloud structures as a result of uncertain and chaotic processes. Local PV forecasting is unavoidable in supply and load planning necessary in integration of smart systems into electrical grids. Intra- or day-ahead modelling of weather patterns based on Artificial Intelligence (AI) allows one to refine available 24 h. cloudiness forecast or predict PV production at a particular plant location during the day. AI usually gets an adequate prediction quality in shorter-level horizons, using the historical meteo- and PV record series as compared to Numerical Weather Prediction (NWP) systems. NWP models are produced every 6 h to simulate grid motion of local cloudiness, which is additionally delayed and usually scaled in a rough less operational applicability. Differential Neural Network (DNN) is based on a newly developed neurocomputing strategy that allows the representation of complex weather patterns analogous to NWP. DNN parses the n-variable linear Partial Differential Equation (PDE), which describes the ground-level patterns, into sub-PDE modules of a determined order at each node. Their derivatives are substituted by the Laplace transforms and solved using adapted inverse operations of Operation Calculus (OC). DNN fuses OC mathematics with neural computing in evolution 2-input node structures to form sum modules of selected PDEs added step-by-step to the expanded composite model. The AI multi- 1…9-h and one-stage 24-h models were evolved using spatio-temporal data in the preidentified daily learning sequences according to the applied input–output data delay to predict the Clear Sky Index (CSI). The prediction results of both statistical schemes were evaluated to assess the performance of the AI models. Intraday models obtain slightly better prediction accuracy in average errors compared to those applied in the second-day-ahead evening approach. However, the first-morning strategy application includes a much more complicated and time-consuming procedure, so in conclusion, the second one-mode series processing could be recommended. Model statistics are generally more successful than NWP data processing due to limitations, delays, and unavailability of free radiation or cloudiness forecasts in the systems. An all-day model enables the prediction of complete PVP periods in sequenced computing with acceptable operational quality in the late afternoon, which is inevitable in day-ahead management and load scheduling in smart grids based on PV energy.

Research on the Method of Calculating Anchorage Flaw Length Based on the Propagation Characteristics of Bolt Excitation Stress Waves

This research aimed to detect the defects of anchoring agents’ empty slurries in anchor support. The influence of anchoring defects on the propagation law of stress waves was comprehensively investigated using laboratory tests, theoretical calculations, and other methods. The characteristic modal components with symmetry and periodicity laws were extracted by adopting a variable modal decomposition (VMD) signal decomposition method. It was found that the bottom reflection time of stress waves had an inverse function relationship with the length of the anchorage flaw. The average propagation speed of the stress wave in the free rod was obtained as 5150 m/s, and the average consolidation wave speed was 4198 m/s. The calculation method of the bolt flaw length was finally proposed. After experimental verification, the average error rate was 2.65%, which meets the requirement of testing accuracy in the engineering field, which provides a guarantee for safe production.

LEVERAGING THE UBIQUITOUS GPS SENSOR ON THE SMARTPHONES FOR ACCESSIBLE LAND SURVEYING

The Global Positioning System (GPS) is an essential tool in land surveying. GPS has become an alternative method of surveying that requires less manpower and less time. However, GPS devices are still expensive to buy, especially for students. On the other hand, almost every student has a smartphone with a built-in GPS sensor, so this GPS is certainly accessible to everyone with a smartphone. This study used a smartphone's GPS to conduct land surveying at the campus of Bengkulu University. This smartphone’s GPS was used to track various parameters such as coordinates, elevation, and distance between two or more points to calculate the area within the study area. The innovative method of using the built-in GPS sensor in smartphones will provide convenience for users and introduce simplified open-source software for the land measurement process. The measurement was calibrated using a conventional roll meter to verify the linear error by comparing the two measurements between the smartphone’s GPS and roll meter. The smartphone’s GPS reading was logged using GPS Waypoints and My Tracks, free Android applications on the Google Play store. This study's average error in measurements obtained using GPS on smartphones was 3.02%. This value is sufficient for the initial stage of low-cost land surveys and falls within ideal conditions for GPS measurements. Therefore, this article emphasizes the potential of smartphone GPS to optimize techniques in education and scientific investigations.

HPD-Depth: High performance decoding network for self-supervised monocular depth estimation

Self-supervised monocular depth estimation methods have shown promising results by leveraging geometric relationships among image sequences for network supervision. However, existing methods often face challenges such as blurry depth edges, high computational overhead, and information redundancy. This paper analyzes and investigates technologies related to deep feature encoding, decoding, and regression, and proposes a novel depth estimation network termed HPD-Depth, optimized by three strategies: utilizing the Residual Channel Attention Transition (RCAT) module to bridge the semantic gap between encoding and decoding features while highlighting important features; adopting the Sub-pixel Refinement Upsampling (SPRU) module to obtain high-resolution feature maps with detailed features; and introducing the Adaptive Hybrid Convolutional Attention (AHCA) module to address issues of local depth confusion and depth boundary blurriness. HPD-Depth excels at extracting clear scene structures and capturing detailed local information while maintaining an effective balance between accuracy and parameter count. Comprehensive experiments demonstrate that HPD-Depth performs comparably to state-of-the-art algorithms on the KITTI benchmarks and exhibits significant potential when trained with high-resolution data. Compared with the baseline model, the average relative error and squared relative error are reduced by 6.09% and 12.62% in low-resolution experiments, respectively, and by 11.3% and 18.5% in high-resolution experiments, respectively. Moreover, HPD-Depth demonstrates excellent generalization performance on the Make3D dataset.

State of Health Estimation of Lithium-Ion Batteries Based on Hybrid Neural Networks with Residual Connections

Accurately estimating the state of health (SOH) of lithium-ion batteries is essential for ensuring the stability and safety of the battery. Although the hybrid neural network model demonstrates strong performance in estimating the SOH, network degradation becomes a significant issue as the depth of the neural network increases, potentially undermining the accuracy of the estimation. This paper presents a hybrid neural network estimation model with residual connections to address the issue of network degradation. First, the model utilizes a combination of convolutional neural networks and an attention mechanism to automatically extract feature information highly correlated with SOH from the partial charging data of lithium-ion batteries. Subsequently, a multi-layer gated recurrent unit (GRU) is employed to capture temporal information within the extracted features. To address the issue of network degradation that arises from stacking multiple layers of neural networks, residual connections are incorporated into the multi-layer GRUs, mitigating the accumulation of errors within deep networks. Finally, three distinct datasets are employed to validate the proposed model. The experimental results demonstrate that the model exhibits an average mean absolute error and root mean square error of less than 1.8% on both of these datasets.

Research on virtual collection method of layer house temperature for the construction requirements of digital twin system.

At present, in the context of the highly intensive development of livestock and poultry breeding, digital management is becoming increasingly important, and digital twin systems are gradually being applied. To solve the contradiction between data acquisition and sensor network congestion, a virtual acquisition method based on historical data and real-time reference of point data is proposed when constructing a digital twin system. Firstly, computational fluid dynamics (CFD) simulation was used to analyze and determine the temperature distribution and environmental characteristics inside the layer house, and the collection area was preliminarily divided according to the CFD simulation results. Then, combined with gray correlation degree and cosine similarity analysis, it can effectively identify the reference points highly correlated with the temperature of the key unmonitored area. Finally, WOA was used to optimize the BiLSTM hyperparameters and construct a WOA-BiLSTM virtual acquisition model. It is based on the XGBoost algorithm to determine the actual data collection points, predict the current value based on the actual data of the reference point and the historical data of the test point, and complete virtual collection. Through the test in a farm, the average absolute error between the data of 10 virtual collection points and the actual data was within 0.25 °C, which ensured the reliability of the data. It analyzes the data volume requirements for digital twin modeling and theoretically verifies the supporting role of virtual collection in the construction of digital twin systems.

Evaluating interface pressure in a lower-limb prosthetic socket: Comparison of FEM and experimental measurements on a roll-over simulator.

Improper socket fitting in lower-limb prostheses can lead to significant complications, including pain, skin lesions, and pressure ulcers. Current suspension and socket design practices rely predominantly on visual inspection of the residual limb and patient feedback. Monitoring stress distribution at the residual limb/socket interface offers a more objective approach. Finite Element Analysis (FEA) enables to estimate interface pressure distribution prior to manufacture to provide the orthoprosthetist with quantitative data during socket rectification and interface prosthetic components selection. However, although numerous FEA models are available, few have undergone rigorous validation against experimental pressure data. Indeed, limitations of commercial pressure sensors typically include cumbersomeness or imprecision, thereby hindering systematic measurements within the socket. In this study, we introduce a low-cost, accurate pressure sensing system integrated into 3D-printed sockets for FEA validation. The system is implemented on a roll-over simulator that uses a mock limb to mimic the interaction between a transtibial residual limb and socket during the unipodal stance phase. A FEA of the simulator was then conducted, and predicted interface pressures were compared to experimental measurements at seven discrete locations. The model demonstrated a high degree of sensitivity to the geometry of the mock limb; however, with an accurate shape description, it was able to predict pressure with an average absolute error of 12kPa. This work advances the validation of residual limb FEA for estimating residual limb/socket interface pressures. It highlights the potential of FEA for designing data-driven sockets and ultimately improve the comfort of prosthesis users.

From concept to application: building and testing a low-cost light detection and ranging system for small mobile robots using time-of-flight sensors

Advancements in light detection and ranging (LiDAR) technology have significantly improved robotics and automated navigation. However, the high cost of traditional LiDAR sensors restricts their use in small-scale robotic projects. This paper details the development of a low-cost LiDAR prototype for small mobile robots, using time-of-flight (ToF) sensors as a cost-effective alternative. Integrated with an ESP32 microcontroller for real-time data processing and Wi-Fi connectivity, the prototype facilitates accurate distance measurement and environmental mapping, crucial for autonomous navigation. Our approach included hardware design and assembly, followed by programming the ToF sensors and ESP32 for data collection and actuation. Experiments validated the accuracy of the ToF sensors under static, dynamic, and varied lighting conditions. Results show that our low-cost system achieves accuracy and reliability comparable to more expensive options, with an average mapping error within acceptable limits for practical use. This work offers a blueprint for affordable LiDAR systems, expanding access to technology for research and education, and demonstrating the viability of ToF sensors in economical robotic navigation and mapping solutions.

Machine learning-assisted assessment of municipal solid waste thermal treatment efficacy via rapid image recognition and visual analysis.

Decentralized thermal treatment is a common method for municipal solid waste (MSW) disposal in rural areas. However, evaluating the effect of incineration has always been challenging owing to the difficult and time-consuming measurements involved. Herein, this study presented a rapid image recognition method for assessing the effects of thermal treatment on MSW using a neural network algorithm and a BAEVA 1.0 software based on the relation between the ignition loss of the incinerated bottom ash and its color properties. Through Pearson correlation analysis, the results demonstrated a strong correlation (R2>0.80) between the ignition loss and the R, G, and B color values. To enhance evaluation accuracy, we introduced the backpropagation artificial neural network (BPANN) algorithm, which exhibited an average evaluation error of only 3.21 in crossvalidation, 27.9% lower than that of the linear regression model. Building upon the BPANN, we developed BAEVA 1.0 as a software tool for thermal treatment effect evaluation. This tool exhibited advantages in functionality, convenience, and accuracy compared to existing methods. Overall, this research provides an important rapid assessment approach for evaluating the effects of MSW incineration when measurement conditions are unavailable.

Medicine Prescribing Practices and Prescription Errors Evaluations at Outpatient Department in Debre Berhan Comprehensive Specialized Hospital, Amhara Regional State, Ethiopia

Background: Currently, irrational uses of medicines becoming global problem largely in developing countries like Ethiopia. Inappropriate prescribing is a major cause for poor treatment outcome and higher costs. Hence, this study was aimed to investigate medicine prescribing practice and prescription errors using WHO medicine-utilization core indicators. Methods: A hospital based retrospective cross sectional study design was used to evaluate prescribing practices and prescription errors from September to October, 2024 at the OPD pharmacy using systematic random sampling technique while a prospective approach was employed for facility indicators. Presence of potential drug-drug interactions (DDIs) were evaluated using Medscape Online Drug Interaction Checker. Data were analyzed using SPSS version 25 and interpreted as tables and figures. Results: A total of 1019 medicines were prescribed in 524 prescriptions and 81.6% (n = 832) were actually dispensed. The percentage of antibiotic, injections and medicine prescribed from Essential Drug List was 33.9% (n = 345),3.5% (n = 36) and 92.3% (n = 941) respectively. The most frequently prescribed class of medicine were antibiotics 33.9% (n = 345). 65.1% (n = 341) were ≥2 medicines and 8.3% (n = 85) had at least one potential DDIs. Among overall DDIs, the monitor closely and serious level was 60% (n = 51) and 11.8% (n = 10) respectively. The average prescription error was 4.3. Prescription errors due to failure to mention diagnosis was 40.6% (n = 213). Conclusion: Based on findings, the prescribing practices had defects to the optimum value recommended by WHO and showed high prescription errors. Antibiotics prescribing was the major problem in practice. Remarkable DDIs were observed in prescribed medicines. Therefore, designing and implementing policy to improve medicine use practice is highly indispensable.

Design of a road marking violation detection system at railway level crossings

When a train passed through a railway-level crossing, a common phenomenon was that many vehicles attempted to overtake others by crossing into lanes designated for oncoming traffic, resulting in both roads becoming congested with motorized vehicles. At that time, no system was in place to enforce penalties for violating road markings at level crossings. Therefore, a system capable of detecting such violations when trains pass through was needed. The designed system utilized a Raspberry Pi 4, a webcam, and an ultrasonic sensor. The single shot detector (SSD) method was employed for vehicle classification. The optical character recognition (OCR) method was used for character recognition on license plates. The research involved object detection at level crossings using varied objects (cars and motorcycles) with license plates categorized into two types: white background plates with black numbers and black background plates with white numbers. Based on the research results, turning on the webcam when the bar opened and closed using an ultrasonic sensor got an average error of 0.573% and 0.582%. The system could distinguish objects with an average recognition delay of 0.554 seconds and 0.702 seconds for car and motorbike objects. Regarding number plate detection, the success rate of character recognition stood at 64.45%.

Non-linear shear characteristics and wear behaviors of joints considering the factors of roughness, size effect, and joint strength

This study investigates the non-linear shear behaviors and failure mechanisms of rock-like materials with three-dimensional joint regarding various sample diameters (d), joint roughness coefficients (JRC), and joint wall strength ratios (SR). A series of cylindrical specimens were prepared and underwent direct shear tests. The peak shear stress (τmax) decreases by 13.09%–25.98% with an increasing d due to the intensified stress concentration resultant from a diminished contact area. A higher JRC increases τmax by 13.16%–50.70% due to enhanced interlocking effects. An increase in SR improves the matrix mechanical properties, resulting in a gentle growth in τmax by 7.30%–18.27%. The normal displacement (δv)–shear displacement (u) curves and failure morphologies of the joints indicate that, as d decreases or JRC and SR increase, the curves gradually move upward and the failure modes of the joints transfer from plastic shear flow to brittle shear failure. Furthermore, the finite element method simulation was introduced to analyze the mesoscopic wear characteristics of the joint surfaces. The results reveal that the process of joint failure can be categorized into three stages including wear, shearing, and further smoothing, and the failure degree on the joints exacerbates with a smaller d or larger JRC and SR. Additionally, an improved non-linear shear failure criterion considering the influences of size effect, SR, and JRC is developed based on the Barton's JRC-JCS (joint compressive strength) model, with the average error reduced significantly from 8.12% to 3.23%.

A unit series–parallel unsaturated soil electrical conductivity model considering interconnections between pores

For the potential application of electrical resistivity measurements in hydrogeological investigations, the knowledge of soil electrical conductivity mechanism with moisture content variation is the key issue. The impact of interconnections between pores and weak connection to physical processes were the two limitations for unsaturated soil electrical conductivity research. In this work, we introduced the concept of equivalent conductive pathway to analyze tortuosity. Based on media series–parallel analysis, a unit series–parallel unsaturated soil electrical conductivity model considering interconnections between pores was established, in which parameters with its own physical meaning. To verify the accuracy of the proposed model, soil resistivity test with moisture content variation was conducted. Soil electrical conductivity was predicted, which was compared with results from test and previous models. The results indicate that our model is expected to produce better results than the previous models. Overcoming the limitations of weak connection to physical processes in the empirical model, the error of soil resistivity was significantly reduced when the impact of interconnections between pores was considered. The average error of the proposed model for clay and sand was 30.6% and 17.4%, respectively, compared to 72.6% and 48.2% for the model that ignored the interconnections between pores. The findings of this study could provide a reference for hydrogeological investigations, such as levee leakage detection.

Research on dynamic response of vertical displacement of runway pavement slab end under aircraft taxiing using novel FBG sensors

The vertical displacement dynamic response measuring at slab joints is crucial for pavement management. In order to obtain accurate dynamic vertical displacement response of slab end, this study developed a vertical displacement sensor based on Fiber Bragg Grating (FBG) technology, featuring an accuracy of 0.15 mm and a resolution of 0.015 mm, which embedded alongside other sensors into Lukou Airport taxiway. The time-history curves of vertical displacement of slab end at both joint sides were obtained, enabling a comprehensive capture of dynamic response during taxiing. In addition, finite element method (FEM) was conducted to simulate taxiing under varying aircraft types, wheel lateral deviation, and speeds. The findings revealed close alignment between FBG sensors and FEM results, with 0.18 mm average error for the sensor, clearly demonstrating its exceptional reliability. Additionally, the influence pattern of taxiing parameters on vertical displacement of slab end including wheel load and path were revealed using FEM.

Noninvasive photoacoustic synergistic measurement of vascular morphology and blood flow velocity

Photoacoustic imaging is a multimodal imaging technique combining optical and acoustic principles to observe and analyze structural and functional tissue information. This study presents a photoacoustic velocimetry and imaging experimental system for synergistic measurement of vascular morphology and blood flow velocity. The system, equipped with a 532 nm pulsed laser and optical-acoustic combiner, achieves a horizontal resolution of 10 μm. It measures flow velocities ranging from 25 to 200 mm/s with an average error of 8.2%. Through two-dimensional scanning and Gaussian filtering, the vascular morphology was reconstructed with a Dice Similarity Coefficient of 0.92. Compared to conventional imaging systems, the proposed system enhances small vessel resolution by up to 40% and achieves a signal-to-noise ratio improvement of 7 dB. These results highlight the system's potential for precise and noninvasive biomedical applications, including real-time diagnosis and monitoring. Vascular imaging plays a critical role in biomedical research, aiding in understanding vascular structure and function and exploring cardiovascular diseases, tumor angiogenesis, hemodynamics, and other related diseases. Noninvasive imaging techniques can evaluate disease progression and therapy efficacy. Accurate blood flow measurement is essential for diagnosing various diseases, including stroke, diabetes, and cancer. This work demonstrates a synergistic photoacoustic system capable of achieving high-resolution vascular imaging and flow velocity measurements.