Search

Simple Summary: Through beef cattle behavior recognition technology, livestock farmers can promptly identify abnormal behaviors in beef cattle, providing data and technological support for intelligent and welfare-oriented cattle farming. This study is based on computer vision's convolutional neural networks. It constructs a new method for the real-time detection of beef cattle behavior by optimizing the convolutional module and adding attention mechanisms. The method is evaluated on nine behavior datasets, including standing, lying, mounting, fighting, licking, eating, drinking, walking, and searching. It significantly improves the performance of beef cattle behavior recognition, achieving an average accuracy of 96.5%, which serves as a foundation for beef cattle health assessment and information-based farming. Behavior recognition in beef cattle is a crucial component of beef cattle behavior warning and intelligent farming. Traditional beef cattle behavior recognition faces challenges in both difficulty in identification and low accuracy. In this study, the YOLOv8n_BiF_DSC (Fusion of Dynamic Snake Convolution and BiFormer Attention) algorithm was employed for the non-intrusive recognition of beef cattle behavior. The specific steps are as follows: 45 beef cattle were observed using a fixed camera (A LINE OF DEFENSE) and a mobile phone (Huawei Mate20Pro) to collect and filter posture data, yielding usable videos ranging from 1 to 30 min in length. These videos cover nine different behaviors in various scenarios, including standing, lying, mounting, fighting, licking, eating, drinking, walking, and searching. After data augmentation, the dataset comprised 34,560 samples. The convolutional layer (CONV) was improved by introducing variable convolution and dynamic snake-like convolution modules. The dynamic snake-like convolution, which yielded the best results, expanded the model's receptive field, dynamically perceived key features of beef cattle behavior, and enhanced the algorithm's feature extraction capability. Attention mechanism modules, including SE (Squeeze-and-Excitation Networks), CBAM (Convolutional Block Attention Module), CA (Coordinate Attention), and BiFormer (Vision Transformer with Bi-Level Routing Attention), were introduced. The BiFormer attention mechanism, selected for its optimal performance, improved the algorithm's ability to capture long-distance context dependencies. The model's computational efficiency was enhanced through dynamic and query-aware perception. Experimental results indicated that YOLOv8n_BiF_DSC achieved the best results among all improved algorithms in terms of accuracy, average precision at IoU 50, and average precision at IoU 50:95. The accuracy of beef cattle behavior recognition reached 93.6%, with the average precision at IoU 50 and IoU 50:95 being 96.5% and 71.5%, respectively. This represents a 5.3%, 5.2%, and 7.1% improvement over the original YOLOv8n. Notably, the average accuracy of recognizing the lying posture of beef cattle reached 98.9%. In conclusion, the YOLOv8n_BiF_DSC algorithm demonstrates excellent performance in feature extraction and high-level data fusion, displaying high robustness and adaptability. It provides theoretical and practical support for the intelligent recognition and management of beef cattle. [ABSTRACT FROM AUTHOR]

Individual identification of pigs is a critical component of intelligent pig farming. Traditional pig ear-tagging requires significant human resources and suffers from issues such as difficulty in recognition and low accuracy. This paper proposes the YOLOv5-KCB algorithm for non-invasive identification of individual pigs. Specifically, the algorithm utilizes two datasets—pig faces and pig necks—which are divided into nine categories. Following data augmentation, the total sample size was augmented to 19,680. The distance metric used for K-means clustering is changed from the original algorithm to 1-IOU, which improves the adaptability of the model's target anchor boxes. Furthermore, the algorithm introduces SE, CBAM, and CA attention mechanisms, with the CA attention mechanism being selected for its superior performance in feature extraction. Finally, CARAFE, ASFF, and BiFPN are used for feature fusion, with BiFPN selected for its superior performance in improving the detection ability of the algorithm. The experimental results indicate that the YOLOv5-KCB algorithm achieved the highest accuracy rates in pig individual recognition, surpassing all other improved algorithms in average accuracy rate (IOU = 0.5). The accuracy rate of pig head and neck recognition was 98.4%, while the accuracy rate for pig face recognition was 95.1%, representing an improvement of 4.8% and 13.8% over the original YOLOv5 algorithm. Notably, the average accuracy rate of identifying pig head and neck was consistently higher than pig face recognition across all algorithms, with YOLOv5-KCB demonstrating an impressive 2.9% improvement. These results emphasize the potential for utilizing the YOLOv5-KCB algorithm for precise individual pig identification, facilitating subsequent intelligent management practices. [ABSTRACT FROM AUTHOR]

Air pollution constitutes a substantial peril to human health, thereby catalyzing the evolution of an array of air quality prediction models. These models span from mechanistic and statistical strategies to machine learning methodologies. The burgeoning field of deep learning has given rise to a plethora of advanced models, which have demonstrated commendable performance. However, previous investigations have overlooked the salience of quantifying prediction uncertainties and potential future interconnections among air monitoring stations. Moreover, prior research typically utilized static predetermined spatial relationships, neglecting dynamic dependencies. To address these limitations, we propose a model named D ynamic S patial- T emporal D enoising D iffusion P robabilistic M odel (DST-DDPM) for air quality prediction. Our model is underpinned by the renowned denoising diffusion model, aiding us in discerning indeterminacy. In order to encapsulate dynamic patterns, we design a dynamic context encoder to generate dynamic adjacency matrices, whilst maintaining static spatial information. Furthermore, we incorporate a spatial-temporal denoising model to concurrently learn both spatial and temporal dependencies. Authenticating our model's performance using a real-world dataset collected in Beijing, the outcomes indicate that our model eclipses other baseline models in terms of both short-term and long-term predictions by 1.36% and 11.62% respectively. Finally, we conduct a case study to exhibit our model's capacity to quantify uncertainties. • Model based on diffusion model for air quality forecasting is proposed. • A dynamic context encoder to learn dynamic spatial relations is designed. • Two Transformer encoders to simultaneously capture spatial-temporal relationships among monitoring stations are introduced. • We implement experiments in a real-world dataset and prove the effectiveness of our model. [ABSTRACT FROM AUTHOR]

• A novel M-PEH with built-in outer-inner magnetic tri-stable oscillators (TOS) is proposed for energy harvesting enhancement. • Built-in outer-inner magnetic TOSs brings more compact structure but wider bandwidth and higher power generation. • Synchronous nonlinear coupling interaction between the outer and inner TOS benefits to energy harvesting enhancement. • Effective bandwidth and output power increased by 300% and 662% compared to the 1-DOF magnet coupling tri-stable PEH. • Practical applications prove the M-PEH meeting the power supplying requirements of engineering community. Tri-stable oscillators have been widely used to develop various nonlinear broadband energy harvesters. Focusing on further achieving high-performance piezoelectric energy harvester under broadband vibrations, a novel M-shaped 2-DOF piezoelectric energy harvester (M-PEH) with built-in outer-inner magnetic tri-stable oscillators is proposed, in which an inner magnetic tri-stable oscillator is built-in into a U-shaped outer magnetic tri-stable oscillator using its cut-out cantilever beam, assisted by the coupling interface to further enhance the dynamic responses. A complete physical model by considering the nonlinear coupling interaction between the outer-inner magnetic tri-stable oscillators is established to describe the dynamic performances under different system parameters. The theoretical model is well validated by experiments, which show that the proposed M-PEH has more compact size, wider operating frequency and higher power generation, offering an effective bandwidth of 20.1 Hz (5.1–25.2 Hz) at base excitation A = 5 m/s2, and achieving maximum output voltage 8.5 V and total peak power 26.7 μW. Compared to the original magnetic-coupling nonlinear tri-stable PEH, the effective bandwidth and total power have increased by 300 % and 662 % respectively, and the average power density is increased by 917 %. In addition, four main coupling interaction modes between the outer and inner built-in tri-stable oscillator are observed, in which the nonlinear synchronous coupling interaction is conducive to generating inter-well motion, leading to more mechanical energy being transferred from the outer tri-stable oscillator to the inner built-in tri-stable oscillator, thereby improving power generation capacity. Besides, the coupling interaction between these two tri-stable oscillators produces two close resonances which can be easily tuned to form a wide operating bandwidth. Practical applications prove that the proposed M-PEH has the ability to power low-powered electronic products, meeting the power supplying requirements of engineering community. [ABSTRACT FROM AUTHOR]

Taper angle is an important factor for characterizing a hyrocyclone's separation performance. Appropriate setting of the taper angle can contribute to particle grading and separation performance of the hydrocyclone. Through numerical analysis, this study investigated the flow fields and separation performances of the hydrocyclones with different taper angles in depth. Moreover, the numerical simulation results were compared with experimental data of the conventional hydrocyclone for accuracy validation. The results showed that the tangential velocity, axial velocity, and the diameter of air core increased with the increasing taper angle. At a too large or a too small taper angle, the inner flow field would become unstable, thereby leading to the increasing number of dislocated particles and reducing the particle separation precision. For the hydrocyclone with a diameter of 75 mm, the separation precision of fine particles can be remarkably enhanced by setting the taper angle of 15°, accompanied with the improvement of product quality. The present data can provide the related enterprises with favorable guidance and a theoretical foundation for further designing the taper angle of novel hydrocyclones. [ABSTRACT FROM AUTHOR]

• A thulium-doped fiber laser with bidirectional single-longitudinal-mode output was proposed. • The proposed TDFL had a simple structure to realize bidirectional output, using only two 2 × 2 optical couplers. • A novel cascaded triple-ring cavity filter was proposed and investigated. The reversibility and the capability of mode selection filtering were investigated by simulation and experiment. • The minimum relative intensity noise was less than −130 dB/Hz at 2 ≥ 2 MHz around the 2-μm band. Single-longitudinal-mode (SLM) fiber lasers play important roles in optical fiber communication. However, there has been little reported research on bidirectional SLM output, especially in the 2-μm band. In the work reported here, a bidirectional SLM output thulium-doped fiber laser (TDFL) using a cascaded triple-ring cavity (CTRC) filter was proposed and investigated. To realize bidirectional output, the proposed TDFL had a simple structure, using only two 2 × 2 optical couplers. To realize SLM output, the transmission spectrum, as well as the reversibility of the CTRC filter were discussed and investigated in detail. The direction-switchable SLM output was realized by polarization-dependent loss. By tuning the polarization controller carefully, each direction could generate SLM output with an optical signal-to-noise ratio (OSNR) > 72 dB. The maximum wavelength and peak power fluctuation over 60 min were, respectively, 0.01 nm and 0.48 dB. The variation of output power with pump power, linewidth, relaxation oscillation, and relative intensity noise were also investigated. To the best of our knowledge, this is the first time that a sub-ring cavity filter has been used to realize bidirectional SLM output. [ABSTRACT FROM AUTHOR]

A piezo-magneto-elastic-electric hybrid structure composed of a piezo-magneto-elastic multi-stable harvester (PMEH), a mass-spring-damper elastic supporting oscillator (ESO) and a nonlinear synchronous switching power extraction (SSPE) circuit is proposed to achieve high-performance energy harvesting over a wide bandwidth. Governing equations describing the dynamic and electrical responses of the hybrid structure with two SSPE circuits are derived, and their steady-state solutions are then solved with harmonic balance method and given along via theoretical simulations. Experimental results exhibit qualitative agreement with the theory. The results show that the coupling nonlinearities between the PMEH and SSPE can broad the bandwidth and enhance the power extraction. The hybrid structure with SSPE can offer a wide range of global inter-well oscillation from 2Hz − 13Hz, and generates a maximum extracted power of 20 µW. Field experiment and practical applications demonstrate that the hybrid structure combining with SSPE circuit has the potential promising in powering low-powered electronic devices. [ABSTRACT FROM AUTHOR]

• A five wavelength-switchable single-longitudinal-mode erbium-doped fiber laser was proposed. • A cascaded PS-FBG was designed and analyzed. To the best of our knowledge, this is the first time that the PS-FBG is used as a multi-channel filter to realize stable multiple SLM single-wavelength outputs. • Five single-wavelength switchable SLM outputs were generated in stable condition over 60 min. • The minimum linewidth at an integration time of 0.001 s was 1.3102 kHz. Wavelength-switchable fiber lasers play an important role in optical fiber communication systems. Although phase-shifted fiber Bragg grating (PS-FBG) has a narrow 3 dB bandwidth, it has rarely been reported as a multi-channel optical filter. Therefore, a five wavelength-switchable single-longitudinal-mode (SLM) erbium-doped fiber laser (EDFL) was implemented and analyzed, where the cascaded PS-FBG with a 3 dB bandwidth of each channel narrower than 0.024 nm was used as a five-channel filter. The design, fabrication, and characterization of cascaded PS-FBG were implemented and analyzed in detail. By tuning the micro-displacement platform, five-single wavelength switchable SLM outputs with a high optical signal-to-noise ratio (OSNR) were generated and the outputs were monitored over 60 min. The measured maximum wavelength drift and the peak power fluctuation were 0.02 nm and 1.73 dB. The relaxation oscillation peaks were lower than −100 dB/Hz and the minimum linewidth at an integration time of 0.001 s was 1.3102 kHz. To the best of our knowledge, this is the first time that the PS-FBG is used as a multi-channel filter to realize multiple SLM single-wavelength outputs. [ABSTRACT FROM AUTHOR]

Ultrasound vibrations, which are abundant in our everyday environment, have long been widely used in industrial applications and medical therapeutics. While there have massive number of studies on harvesting the low-frequency vibration energy to power the low-powered electronic devices, there are few literatures on the ultrasound vibration energy harvesting. In order to validate the feasibility of ultrasound vibration energy harvesting, this article takes a rotary-type piezoelectric ultrasound actuator (PUA) as the research object, a new PZT ring with appropriate zoning and polarization is bonded onto its elastomer top surface to harvest the vibration energy induced by the ultrasound waves. The manufacturing process and operating mechanism of the modified PUA are firstly introduced. Then, the electromechanical coupled model is derived based on the equivalent circuit method, and the dynamic and energetic characteristics of the ultrasound vibration energy harvesting under three excitation modes (i.e., two standing waves and one traveling wave excitations) are theoretically studied. The experimental results testify the theoretical values, and the feasibility of ultrasound vibration energy harvesting is also validated by experiments and an application example. When the PUA is excited by traveling ultrasound wave with amplitude 67 V and frequency 35.1 kHz, the maximum harvested power of each sector of the new PZT ring and λ/4 sector of the bottom PZT ring reach 58 mW and 82.4 mW, respectively, which fully meet the power supplying requirements of low-powered electronic products and implantable medical systems. [ABSTRACT FROM AUTHOR]

In this paper, a new semi-analytical model considering the non-isothermal effect is proposed to predict heated radius and production performance in a cyclic steam stimulated (CSS) horizontal well. Firstly, an assumption of three heated areas based non-isothermal distribution is introduced. Based on the fluid flow behavior of CSS process in a horizontal well, the spreading of heated area can be classified into three stages. And the heated radius can be derived using the energy conservation equation and area equivalent principle respectively. Then a new semi-analytical model for predicting the production performance by using the equivalent flow resistance method is established. After validation of the new model with numerical simulation and oil field data, the effect of non-isothermal on production performance is analyzed in detail. The results show that a larger injection rate, a bigger error caused by isothermal model when steam injection rate exceeds 200 m3/d. Also, the average water saturation rise has an increase of 1.85% and the average temperature has a drop of 21.12% if the non-isothermal effect is considered. This paper provides an accurate and quick method for estimating the heated radius and production performance in a horizontal well during CSS process. • Temperature distribution in reservoir during steam injection is non-isothermal. • A new semi-analytical model considering non-isothermal effect is proposed. • Non-isothermal effect on cyclic steam stimulation with horizontal well is analyzed. • Non-isothermal effect should not be neglected in a high steam injection rate. • The proposed model predicts the production performance more accurately. [ABSTRACT FROM AUTHOR]

Steam-Assisted gravity drainage (SAGD) technique has been successfully applied to unlock the oilsands resources in Canada. But to accurately estimate the steam chamber expansion and productivity of the entire lifetime of a SAGD well pair is still challenging. In this paper, a new comprehensive model is proposed encompassing the different stages of SAGD. Firstly, for the rising stage, based on the theory of volumetric displacement, an analytical model is developed for the rise speed and oil rate. Thereafter, based on Butler's theory, a 2D numerical approach is modelled for predicting the lateral spreading and confinement stage of steam chamber. Thus, combining the two different mathematical models, the steam chamber shape and recovery performance of a SAGD process can be calculated. The calculation results are also compared against 3D experimental results and field data to confirm the accuracy of this model. Then, a sensitivity analysis is performed to discuss the effects of operation parameters on steam chamber expansion and recovery performance of SAGD process. Results indicate that the calculated results can match the experiment results and field date very well. The calculated peak oil production rate causes a large error and the plateau time of SAGD process obviously is shortened if the preheating stage is neglected. A higher steam chamber pressure results in a narrower inverted triangle interface shape in the rising stage and a more concave-like interface in the sideways expansion stage. In addition, the oil rate increases with steam chamber pressure allowing for varying steam injection rate, whereas the increasing trend of ultimate recovery decreases with the steam chamber pressure increases, which indicates that a proper steam chamber pressure to be chosen can not only achieve a preferable ultimate recovery but also benefit for project economics. • A new comprehensive model encompassing different stages of SAGD was developed. • A 3D physical experiment was conducted for verifying the proposed model. • The preheating period has a little impact on the ultimate recovery. • Steam chamber pressure has a great effect on steam chamber evolution. [ABSTRACT FROM AUTHOR]