Biosystems Engineering最新文献_第10页

Monitoring and blockage diagnosis in axial flow threshing and separation device under variable feed conditions 变进料条件下轴流脱粒分离装置的监测与堵塞诊断

IF 5.3 1区农林科学 Q1 AGRICULTURAL ENGINEERING

Biosystems Engineering

Pub Date : 2025-08-27 DOI: 10.1016/j.biosystemseng.2025.104262

Yongle Zhu , Zheng Ma , Zhiping Wu , Zelin Zhang , Yaoming Li , Liang Wang , Yu Pan

To prevent blockage in axial flow threshing and separation devices caused by varying material moisture and feeding rates while simplifying monitoring and diagnostic system, a test bench was used to collect vibration signals from four monitoring points of devices, analyse blockage tendencies under different conditions. Signals were denoised and reconstructed with the Slime Mould Algorithm and Variational Mode Decomposition, and segmented with overlapping moving time windows. Time, frequency, and time-frequency domain features were extracted to assess device operating status and sensitivity of signal changes at different monitoring points. Findings revealed that the duration of a slight blockage tendency was long under normal moisture content and small increments of feeding rate. With high moisture content and large increments of feeding rate, the duration of slight blockage tendency will decrease and quickly enter a severe blockage tendency state, with continued feeding resulting in immediate blockage. The monitoring point directly below the concave grate exhibited the most sensitive signal changes, with the largest waveform variations and standard deviation deviations. Feature dimensionality reduction was performed using Relief-F algorithm, and Bayesian-optimised machine learning models were trained for state identification. The diagnostic model of a monitoring point directly below the concave grate demonstrated high diagnostic accuracy, recall, and reliability, indicative of an effective monitoring point. The Bayesian-optimised Support Vector Machine model achieved the best performance, with 85.1 % and 93.6 % accuracy under different conditions and rapid prediction speeds (53000 and 40000 obs s⁻¹). This met the requirements for a simplified, accurate, and fast online monitoring system.

为了防止物料含水率和进料率变化引起轴流脱粒分离装置堵塞，同时简化监测诊断系统，利用试验台采集了轴流脱粒分离装置四个监测点的振动信号，分析了不同工况下的堵塞趋势。利用黏菌算法和变分模态分解对信号进行去噪和重构，并用重叠运动时间窗对信号进行分割。提取时间、频率和时频域特征，评估设备在不同监测点的运行状态和信号变化的灵敏度。结果表明，在正常含水率和小进料速率下，轻度堵塞趋势持续时间较长。含水率高，进料速率增量大，轻度堵塞倾向持续时间缩短，迅速进入严重堵塞倾向状态，继续进料，立即堵塞。凹栅正下方监测点的信号变化最为敏感，波形变化和标准差偏差最大。使用Relief-F算法进行特征降维，并训练贝叶斯优化的机器学习模型进行状态识别。凹栅正下方监测点的诊断模型具有较高的诊断准确率、召回率和可靠性，表明监测点是有效的。贝叶斯优化的支持向量机模型取得了最好的性能，在不同条件下的准确率分别为85.1%和93.6%，预测速度快（53000和40000 obs - 1）。这满足了简化、准确、快速的在线监测系统的要求。

{"title":"Monitoring and blockage diagnosis in axial flow threshing and separation device under variable feed conditions","authors":"Yongle Zhu , Zheng Ma , Zhiping Wu , Zelin Zhang , Yaoming Li , Liang Wang , Yu Pan","doi":"10.1016/j.biosystemseng.2025.104262","DOIUrl":"10.1016/j.biosystemseng.2025.104262","url":null,"abstract":"<div><div>To prevent blockage in axial flow threshing and separation devices caused by varying material moisture and feeding rates while simplifying monitoring and diagnostic system, a test bench was used to collect vibration signals from four monitoring points of devices, analyse blockage tendencies under different conditions. Signals were denoised and reconstructed with the Slime Mould Algorithm and Variational Mode Decomposition, and segmented with overlapping moving time windows. Time, frequency, and time-frequency domain features were extracted to assess device operating status and sensitivity of signal changes at different monitoring points. Findings revealed that the duration of a slight blockage tendency was long under normal moisture content and small increments of feeding rate. With high moisture content and large increments of feeding rate, the duration of slight blockage tendency will decrease and quickly enter a severe blockage tendency state, with continued feeding resulting in immediate blockage. The monitoring point directly below the concave grate exhibited the most sensitive signal changes, with the largest waveform variations and standard deviation deviations. Feature dimensionality reduction was performed using Relief-F algorithm, and Bayesian-optimised machine learning models were trained for state identification. The diagnostic model of a monitoring point directly below the concave grate demonstrated high diagnostic accuracy, recall, and reliability, indicative of an effective monitoring point. The Bayesian-optimised Support Vector Machine model achieved the best performance, with 85.1 % and 93.6 % accuracy under different conditions and rapid prediction speeds (53000 and 40000 obs s<sup>−1</sup>). This met the requirements for a simplified, accurate, and fast online monitoring system.</div></div>","PeriodicalId":9173,"journal":{"name":"Biosystems Engineering","volume":"258 ","pages":"Article 104262"},"PeriodicalIF":5.3,"publicationDate":"2025-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144903120","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Study on anti-blocking law of variable amplitude screening under multipoint feeding of materials 多点给料条件下变幅筛防堵规律研究

IF 5.3 1区农林科学 Q1 AGRICULTURAL ENGINEERING

Biosystems Engineering

Pub Date : 2025-08-26 DOI: 10.1016/j.biosystemseng.2025.104272

Zheng Ma , Yongle Zhu , Yanbin Liu , Zhiping Wu

The cleaning process in crop harvesting is crucial, with sieve vibration parameters significantly impacting cleaning quality. However, low load thresholds and high feed rates of sieve can cause material buildup and blockages. To address this challenge, the screening mechanism can be enhanced using the principle of variable amplitude. The anti-blocking screening process with variable amplitude is simulated under multi-point material feeding by Multi-Body Dynamics and Discrete Element Method, where the feed rate of the grain mixture, and the angle of the guide chute, are treated as variables. The study then analysed the average velocity of the particle groups on the screen surface, the material mass ratio, screening efficiency, and impurity content. The results indicated that as the guide chute angle increased, the screening time, decreased progressively, which suggested that variable amplitude can effectively enhance screening efficiency and mitigate material accumulation. More critically, variations in the feed rate, significantly impact screening performance, thereby changing both screening efficiency and anti-blocking capabilities. Based on these optimised parameters, a mathematical model was developed and validated experimentally. High-speed photography and image processing were used to analyse the material distribution on the screen surface, confirming the reliability of the amplitude adjustment mechanism. The findings demonstrate that amplitude screening offers substantial benefits in improving efficiency and reducing blockages. Ultimately, the model's R² value of 0.884 confirms its high reliability, and the consistency between the simulation and experimental results underscores the effectiveness and feasibility of the proposed amplitude screen method.

农作物收获的清洗过程至关重要，筛振参数对清洗质量有重要影响。然而，低负荷阈值和筛的高进料速率会导致物料堆积和堵塞。为了解决这一挑战，可以使用可变振幅原理增强筛选机制。采用多体动力学和离散元法模拟了多点进料条件下的变幅抗堵筛分过程，其中以混合颗粒进料速率和导溜槽角度为变量。然后分析了颗粒群在筛表面的平均速度、物料质量比、筛分效率和杂质含量。结果表明，随着滑槽角度的增大，筛分时间逐渐缩短，表明变振幅可以有效提高筛分效率，减轻物料堆积。更关键的是，进给量的变化会显著影响筛分性能，从而改变筛分效率和抗堵塞能力。基于这些优化参数，建立了数学模型并进行了实验验证。利用高速摄影和图像处理技术分析了屏幕表面的物质分布，证实了振幅调节机构的可靠性。研究结果表明，振幅筛分在提高效率和减少堵塞方面具有实质性的好处。最终，该模型的R2值为0.884，证实了该模型具有较高的可靠性，仿真结果与实验结果的一致性表明了所提出的幅度筛方法的有效性和可行性。

{"title":"Study on anti-blocking law of variable amplitude screening under multipoint feeding of materials","authors":"Zheng Ma , Yongle Zhu , Yanbin Liu , Zhiping Wu","doi":"10.1016/j.biosystemseng.2025.104272","DOIUrl":"10.1016/j.biosystemseng.2025.104272","url":null,"abstract":"<div><div>The cleaning process in crop harvesting is crucial, with sieve vibration parameters significantly impacting cleaning quality. However, low load thresholds and high feed rates of sieve can cause material buildup and blockages. To address this challenge, the screening mechanism can be enhanced using the principle of variable amplitude. The anti-blocking screening process with variable amplitude is simulated under multi-point material feeding by Multi-Body Dynamics and Discrete Element Method, where the feed rate of the grain mixture, and the angle of the guide chute, are treated as variables. The study then analysed the average velocity of the particle groups on the screen surface, the material mass ratio, screening efficiency, and impurity content. The results indicated that as the guide chute angle increased, the screening time, decreased progressively, which suggested that variable amplitude can effectively enhance screening efficiency and mitigate material accumulation. More critically, variations in the feed rate, significantly impact screening performance, thereby changing both screening efficiency and anti-blocking capabilities. Based on these optimised parameters, a mathematical model was developed and validated experimentally. High-speed photography and image processing were used to analyse the material distribution on the screen surface, confirming the reliability of the amplitude adjustment mechanism. The findings demonstrate that amplitude screening offers substantial benefits in improving efficiency and reducing blockages. Ultimately, the model's <em>R</em><sup>2</sup> value of 0.884 confirms its high reliability, and the consistency between the simulation and experimental results underscores the effectiveness and feasibility of the proposed amplitude screen method.</div></div>","PeriodicalId":9173,"journal":{"name":"Biosystems Engineering","volume":"258 ","pages":"Article 104272"},"PeriodicalIF":5.3,"publicationDate":"2025-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144895550","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Hydrodynamic responses of an aquaculture cube net cage suspended inside a jacket foundation for various net solidities and bottom weights 悬挂在护套基础内的水产养殖立方体网箱对不同网固化度和底重的水动力响应

IF 5.3 1区农林科学 Q1 AGRICULTURAL ENGINEERING

Biosystems Engineering

Pub Date : 2025-08-19 DOI: 10.1016/j.biosystemseng.2025.104271

Yuan-Mao Zhang , Chun-Wei Bi , Xue-Wen Zhou , Sen-Qi Cui , Liu-Yi Huang

The wind-aquaculture integration structure offers significant potential for marine resource use and is gaining increasing attention from academia and engineering community. This study introduces an innovative integrated structure called JOWT-AC, which suspends a net cage within a jacket structure. Based on the lumped-mass method, a cube cage is established within the internal space of the jacket. The contact interaction between the net and the jacket is considered by introducing a penetration model. Based on previous research findings and 1/25 scale flume experiments, the accuracy of the cage motion and tension was validated. The results show that as the current velocity increases, the cage state can be categorised into three stages: the safe aquaculture regime, the square bottom ring contact regime, and the cage unbalance regime. When the net solidity is between 0.12 and 0.20, the cage becomes unbalanced at a current velocity of 1.2 m s⁻¹. When the bottom weight ranges from 1000 to 1600 kg, the current velocity at which the cage becomes unbalanced varies. The increase of the bottom weight can effectively enhance the cage's ability to resist unbalance. The research results can provide valuable support for integrated systems design of jacket structure and cage.

风养殖一体化结构具有巨大的海洋资源利用潜力，正日益受到学术界和工程界的重视。本研究介绍了一种名为JOWT-AC的创新集成结构，该结构将网笼悬挂在夹套结构中。基于集中质量法，在夹套内部空间内建立了一个立方体保持架。通过引入侵彻模型，考虑了网与夹套之间的接触相互作用。基于前人的研究成果和1/25尺度水槽试验，验证了笼体运动和张力的准确性。结果表明：随着流速的增大，网箱状态可分为安全养殖状态、方底环接触状态和网箱不平衡状态三个阶段；当净固体度在0.12 ~ 0.20之间时，保持架在1.2 m s−1的流速下变得不平衡。当底部重量在1000到1600公斤之间时，使笼变得不平衡的电流速度是不同的。增加底部重量可以有效增强笼的抗不平衡能力。研究结果可为夹套结构与保持架的集成系统设计提供有价值的支持。

{"title":"Hydrodynamic responses of an aquaculture cube net cage suspended inside a jacket foundation for various net solidities and bottom weights","authors":"Yuan-Mao Zhang , Chun-Wei Bi , Xue-Wen Zhou , Sen-Qi Cui , Liu-Yi Huang","doi":"10.1016/j.biosystemseng.2025.104271","DOIUrl":"10.1016/j.biosystemseng.2025.104271","url":null,"abstract":"<div><div>The wind-aquaculture integration structure offers significant potential for marine resource use and is gaining increasing attention from academia and engineering community. This study introduces an innovative integrated structure called JOWT-AC, which suspends a net cage within a jacket structure. Based on the lumped-mass method, a cube cage is established within the internal space of the jacket. The contact interaction between the net and the jacket is considered by introducing a penetration model. Based on previous research findings and 1/25 scale flume experiments, the accuracy of the cage motion and tension was validated. The results show that as the current velocity increases, the cage state can be categorised into three stages: the safe aquaculture regime, the square bottom ring contact regime, and the cage unbalance regime. When the net solidity is between 0.12 and 0.20, the cage becomes unbalanced at a current velocity of 1.2 m s<sup>−1</sup>. When the bottom weight ranges from 1000 to 1600 kg, the current velocity at which the cage becomes unbalanced varies. The increase of the bottom weight can effectively enhance the cage's ability to resist unbalance. The research results can provide valuable support for integrated systems design of jacket structure and cage.</div></div>","PeriodicalId":9173,"journal":{"name":"Biosystems Engineering","volume":"258 ","pages":"Article 104271"},"PeriodicalIF":5.3,"publicationDate":"2025-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144866406","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Simulating light quantity and quality over plant organs using a ray-tracing method to investigate plant responses in growth chambers 利用光线追踪方法模拟植物器官上的光量和质量，研究植物在生长室内的反应

IF 5.3 1区农林科学 Q1 AGRICULTURAL ENGINEERING

Biosystems Engineering

Pub Date : 2025-08-15 DOI: 10.1016/j.biosystemseng.2025.104256

Sabine Demotes-Mainard , Hervé Autret , Christophe Pradal , Julien Le Gall , Vincent Guérin , Nathalie Leduc , Didier Combes , Christophe Renaud , Michaël Chelle , Jessica Bertheloot

Ray-tracing models enable the assessment of light quantity and quality intercepted by plant organs, supporting biological studies in growth chambers with varying light conditions. However, their validation within canopies and clear usage methods remain limited. This work establishes a reliable method for using these models. The method includes i) accounting for the intensity and spectrum of light sources in the calibration procedure; ii) a generic calibration strategy using a few well-placed light measurement points based on chamber geometry. It evaluates the method to simulate light phylloclimate at the organ scale across biologically relevant wavebands of contrasted widths and properties. Using the SEC2 light simulation framework, three virtual experiments were conducted in a growth chamber, with and without rose plants. Inputs included chamber geometry, material optical properties, lamp emissions, and digitised plant mock-ups. Simulations were compared with spectral measurements at various chamber positions and sensor orientations, both without plants and inside a canopy. Results showed high accuracy in replicating spatial light variability, with RMSE ranging 0.011 to 0.021 and 0.014–0.038 μmol m⁻²s⁻¹nm⁻¹ across different wavebands and sensor orientations, for vertical and horizontal transects, respectively. Applying this approach to a case study demonstrated its effectiveness in formulating new biological hypotheses regarding the role of local light in regulating bud outgrowth. This was achieved by highlighting differences in phylloclimate induced by variations in plant architecture. This work thus provides a comprehensive framework for facilitating the application of ray-tracing models in growth chamber studies.

光线追踪模型能够评估植物器官截获的光的数量和质量，支持在不同光照条件下的生长室内进行生物学研究。然而，它们在冠层内的验证和明确的使用方法仍然有限。这项工作为使用这些模型建立了一个可靠的方法。该方法包括i)在校准过程中计算光源的强度和光谱；Ii)一种通用的校准策略，使用基于腔室几何形状的几个位置良好的光测量点。它评估了在器官尺度上模拟光层状气候的方法，该方法跨越生物相关波段的对比宽度和特性。利用SEC2光模拟框架，在有和没有玫瑰植物的生长室内进行了三次虚拟实验。输入包括室的几何形状、材料的光学特性、灯的辐射和数字化的植物模型。模拟比较了在没有植物和在冠层内的不同室位和传感器方向下的光谱测量结果。结果表明，在不同波段和传感器方向上，垂直和水平样条的RMSE分别为0.011 ~ 0.021和0.014 ~ 0.038 μmol m−2s ~ 1nm−1，具有较高的空间光变异性复制精度。将这种方法应用到一个案例研究中，证明了它在制定关于局部光在调节芽生长中的作用的新的生物学假设方面的有效性。这是通过强调由植物结构变化引起的层状气候差异来实现的。因此，这项工作为促进射线追踪模型在生长室研究中的应用提供了一个全面的框架。

{"title":"Simulating light quantity and quality over plant organs using a ray-tracing method to investigate plant responses in growth chambers","authors":"Sabine Demotes-Mainard , Hervé Autret , Christophe Pradal , Julien Le Gall , Vincent Guérin , Nathalie Leduc , Didier Combes , Christophe Renaud , Michaël Chelle , Jessica Bertheloot","doi":"10.1016/j.biosystemseng.2025.104256","DOIUrl":"10.1016/j.biosystemseng.2025.104256","url":null,"abstract":"<div><div>Ray-tracing models enable the assessment of light quantity and quality intercepted by plant organs, supporting biological studies in growth chambers with varying light conditions. However, their validation within canopies and clear usage methods remain limited. This work establishes a reliable method for using these models. The method includes i) accounting for the intensity and spectrum of light sources in the calibration procedure; ii) a generic calibration strategy using a few well-placed light measurement points based on chamber geometry. It evaluates the method to simulate light phylloclimate at the organ scale across biologically relevant wavebands of contrasted widths and properties. Using the SEC2 light simulation framework, three virtual experiments were conducted in a growth chamber, with and without rose plants. Inputs included chamber geometry, material optical properties, lamp emissions, and digitised plant mock-ups. Simulations were compared with spectral measurements at various chamber positions and sensor orientations, both without plants and inside a canopy. Results showed high accuracy in replicating spatial light variability, with RMSE ranging 0.011 to 0.021 and 0.014–0.038 μmol m<sup>−2</sup>s<sup>−1</sup>nm<sup>−1</sup> across different wavebands and sensor orientations, for vertical and horizontal transects, respectively. Applying this approach to a case study demonstrated its effectiveness in formulating new biological hypotheses regarding the role of local light in regulating bud outgrowth. This was achieved by highlighting differences in phylloclimate induced by variations in plant architecture. This work thus provides a comprehensive framework for facilitating the application of ray-tracing models in growth chamber studies.</div></div>","PeriodicalId":9173,"journal":{"name":"Biosystems Engineering","volume":"258 ","pages":"Article 104256"},"PeriodicalIF":5.3,"publicationDate":"2025-08-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144842404","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Characterisation of airflow resistance of different moisture content wheat bulks mixed with different percentages and sizes of dockage 不同含水率、不同填料比例和尺寸的小麦散装料的气流阻力特性

IF 5.3 1区农林科学 Q1 AGRICULTURAL ENGINEERING

Biosystems Engineering

Pub Date : 2025-08-15 DOI: 10.1016/j.biosystemseng.2025.104261

Amin Sadeghizadeh , Fuji Jian , Digvir S. Jayas

Airflow resistance is a key factor related to physical properties of grain bulks and is influenced by dockage size and its percentage. A cubic chamber with inner dimensions of 500 × 500 × 500 mm was employed to investigate the airflow resistance of wheat bulks mixed with varying sizes (≤1.1 mm, 1.1–2.0 mm, and ≥3.3 mm) and percentages of dockage (0, 1.0, 2.5, and 5.0 %), along with different moisture contents (14.5, 16.5, and 18.5 %), airflow velocities (1.1–6.6 m s⁻¹), and in both vertical and horizontal directions. The airflow resistance decreased with the reduced airflow velocity supplied by the fan, and the supplied airflow rate decreased with the increased airflow resistance. Fine materials (≤1.1 mm) mixed with the clean wheat led to 30–50 % increase of airflow resistance in both directions. Adding 1 % of dockage >3.3 mm (chaff) decreased the airflow resistance by ∼10 %, while adding dockage of 1.1–2 mm size (slightly smaller than wheat) did not result in airflow resistance change. Adding dockage with mixed size had an average 45.6 and 41.2 % increase of airflow resistance in horizontal and vertical directions, respectively. Airflow resistance in both clean and wheat mixture in the vertical direction was ∼50 % higher than that in the horizontal direction in any moisture content of wheat mixtures. This increased airflow resistance in the vertical direction was in the range of 10–117 %. The best regression equation was the Modified Haque model for predicting airflow resistances tested in this study.

气流阻力是影响颗粒物性的关键因素，受堆积尺寸和堆积比例的影响。采用内部尺寸为500 × 500 × 500 mm的立方室，研究了不同粒径（≤1.1 mm、1.1 ~ 2.0 mm和≥3.3 mm）、不同含水量（14.5、16.5和18.5%）、不同气流速度（1.1 ~ 6.6 m s−1）、不同垂直和水平方向混合的小麦块体的气流阻力。气流阻力随风机供给气流速度的减小而减小，供给气流速率随气流阻力的增大而减小。细料（≤1.1 mm）与清洁小麦混合，导致气流阻力在两个方向上都增加30 - 50%。添加1%的3.3 mm（箔条）填料可使气流阻力降低~ 10%，而添加1.1-2 mm（略小于小麦）的填料则不会导致气流阻力变化。添加混合尺寸船坞在水平和垂直方向上的气流阻力分别平均增加45.6%和41.2%。在任何含水量的小麦混合料中，清洁和小麦混合料在垂直方向上的气流阻力都比在水平方向上的气流阻力高约50%。在垂直方向上增加的气流阻力在10 - 117%之间。本研究中预测气流阻力的最佳回归方程为修正Haque模型。

{"title":"Characterisation of airflow resistance of different moisture content wheat bulks mixed with different percentages and sizes of dockage","authors":"Amin Sadeghizadeh , Fuji Jian , Digvir S. Jayas","doi":"10.1016/j.biosystemseng.2025.104261","DOIUrl":"10.1016/j.biosystemseng.2025.104261","url":null,"abstract":"<div><div>Airflow resistance is a key factor related to physical properties of grain bulks and is influenced by dockage size and its percentage. A cubic chamber with inner dimensions of 500 × 500 × 500 mm was employed to investigate the airflow resistance of wheat bulks mixed with varying sizes (≤1.1 mm, 1.1–2.0 mm, and ≥3.3 mm) and percentages of dockage (0, 1.0, 2.5, and 5.0 %), along with different moisture contents (14.5, 16.5, and 18.5 %), airflow velocities (1.1–6.6 m s<sup>−1</sup>), and in both vertical and horizontal directions. The airflow resistance decreased with the reduced airflow velocity supplied by the fan, and the supplied airflow rate decreased with the increased airflow resistance. Fine materials (≤1.1 mm) mixed with the clean wheat led to 30–50 % increase of airflow resistance in both directions. Adding 1 % of dockage >3.3 mm (chaff) decreased the airflow resistance by ∼10 %, while adding dockage of 1.1–2 mm size (slightly smaller than wheat) did not result in airflow resistance change. Adding dockage with mixed size had an average 45.6 and 41.2 % increase of airflow resistance in horizontal and vertical directions, respectively. Airflow resistance in both clean and wheat mixture in the vertical direction was ∼50 % higher than that in the horizontal direction in any moisture content of wheat mixtures. This increased airflow resistance in the vertical direction was in the range of 10–117 %. The best regression equation was the Modified Haque model for predicting airflow resistances tested in this study.</div></div>","PeriodicalId":9173,"journal":{"name":"Biosystems Engineering","volume":"258 ","pages":"Article 104261"},"PeriodicalIF":5.3,"publicationDate":"2025-08-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144842405","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Slurry funnels with frequent discharge and partial pit ventilation reduce ammonia, methane, and odorants emissions in sow housing 频繁排放和部分坑通风的泥浆漏斗减少了母猪房内氨，甲烷和气味的排放

IF 5.3 1区农林科学 Q1 AGRICULTURAL ENGINEERING

Biosystems Engineering

Pub Date : 2025-08-14 DOI: 10.1016/j.biosystemseng.2025.104259

Pablo García, Anders Feilberg, Frederik Rask Dalby, Lise Bonne Guldberg, Michael Jørgen Hansen

Livestock slurry is an important anthropogenic source of ammonia, methane, and odour. Novel techniques need to replace standard practices to mitigate emissions. This study explores a slurry system based on funnels with frequent slurry discharge and partial pit ventilation in an experimental pig house with farrowing pens for loose sows. In the control section, a flat bottom slurry system was installed and emptied only at the end of the batch. Ammonia and methane concentrations were quantified with cavity ring-down spectroscopy and odorants with proton-transfer-reaction mass spectrometry. Applying 30 % pit ventilation in both sections reduced ammonia emissions by approximately 50 % in the funnel section compared to the control. Without pit ventilation in the control, the ammonia emissions were 25 % lower in the funnel section with 30 % pit ventilation. Methane was 20–80 % lower in the funnel section and the odour activity value was reduced by more than 50 % in all batches compared to the control. Alternative pit ventilation percentages of 20 and 40 % revealed that the pit ventilation should not be higher than 30 % to avoid increased emissions The funnel system with frequent slurry discharge significantly reduces ammonia, methane, and odorants emissions relative to the control system. Furthermore, the inclusion of partial pit ventilation reduces NH₃ concentration above the slats providing a healthier environment for the sow and piglets.

畜禽粪便是氨、甲烷和气味的重要人为来源。新技术需要取代标准做法来减少排放。本研究探索了一种基于频繁排浆和部分坑通风的漏斗料浆系统，该系统在一个带有散养母猪产仔栏的实验猪舍中进行。在控制段，安装了一个平底泥浆系统，仅在批次结束时排空。用空腔衰荡法测定氨和甲烷浓度，用质子转移反应质谱法测定气味剂浓度。与控制相比，在两个部分都采用30%的坑内通风，漏斗部分的氨排放量减少了约50%。在没有坑道通风的情况下，当坑道通风为30%时，漏斗段的氨排放量降低了25%。与对照相比，漏斗段的甲烷含量降低了20 - 80%，所有批次的气味活性值降低了50%以上。选择20%和40%的坑道通风百分比表明，坑道通风不应高于30%，以避免增加排放。相对于控制系统，频繁排放浆液的漏斗系统显著减少了氨、甲烷和气味剂的排放。此外，部分坑式通风降低了板条上方的NH3浓度，为母猪和仔猪提供了更健康的环境。

{"title":"Slurry funnels with frequent discharge and partial pit ventilation reduce ammonia, methane, and odorants emissions in sow housing","authors":"Pablo García, Anders Feilberg, Frederik Rask Dalby, Lise Bonne Guldberg, Michael Jørgen Hansen","doi":"10.1016/j.biosystemseng.2025.104259","DOIUrl":"10.1016/j.biosystemseng.2025.104259","url":null,"abstract":"<div><div>Livestock slurry is an important anthropogenic source of ammonia, methane, and odour. Novel techniques need to replace standard practices to mitigate emissions. This study explores a slurry system based on funnels with frequent slurry discharge and partial pit ventilation in an experimental pig house with farrowing pens for loose sows. In the control section, a flat bottom slurry system was installed and emptied only at the end of the batch. Ammonia and methane concentrations were quantified with cavity ring-down spectroscopy and odorants with proton-transfer-reaction mass spectrometry. Applying 30 % pit ventilation in both sections reduced ammonia emissions by approximately 50 % in the funnel section compared to the control. Without pit ventilation in the control, the ammonia emissions were 25 % lower in the funnel section with 30 % pit ventilation. Methane was 20–80 % lower in the funnel section and the odour activity value was reduced by more than 50 % in all batches compared to the control. Alternative pit ventilation percentages of 20 and 40 % revealed that the pit ventilation should not be higher than 30 % to avoid increased emissions The funnel system with frequent slurry discharge significantly reduces ammonia, methane, and odorants emissions relative to the control system. Furthermore, the inclusion of partial pit ventilation reduces NH<sub>3</sub> concentration above the slats providing a healthier environment for the sow and piglets.</div></div>","PeriodicalId":9173,"journal":{"name":"Biosystems Engineering","volume":"258 ","pages":"Article 104259"},"PeriodicalIF":5.3,"publicationDate":"2025-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144828017","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Mitigation of greenhouse gas and ammonia emissions due to livestock housing management practices: Analysis of the DATAMAN database 减少牲畜舍管理做法造成的温室气体和氨排放：对DATAMAN数据库的分析

IF 5.3 1区农林科学 Q1 AGRICULTURAL ENGINEERING

Biosystems Engineering

Pub Date : 2025-08-13 DOI: 10.1016/j.biosystemseng.2025.104260

Vasileios Anestis , Wajid Umar , Federico Dragoni , Tony J. van der Weerden , Mélynda Hassouna , Alasdair Noble , Thomas Bartzanas , Barbara Amon

This study analysed management effects on methane (CH₄), nitrous oxide (N₂O) and ammonia (NH₃) emissions from pig and cattle houses, using the global DATAMAN-Housing database. Emission factor (EF) data was grouped by animal sub-category and manure type, and single-variable linear models of EF with housing practice variables (e.g. air scrubber, bedding material, with identified control values) were developed. These variables were examined for statistically significant effects on mean EFs. In closed growing pig housing systems, replacing straw with sawdust-based bedding reduced (p = 0.012) the mean EF_CH4 by 46 %. In deep pit growing pig houses, installing biological air scrubbers at the air outlet decreased the mean EF_NH3 (p = 0.031) by 24 %. Limited N₂O data showed an increased EF_N2O by 57 % (p < 0.0001), but careful interpretation is required due to the low number of observations. Frequent slurry removal from the pit below slatted floors using a scraper lowered (p = 0.011) the mean EF_NH3 by 53 %. In cubicle and loose cattle houses the mean EF_NH3 decreased when using woodchips instead of straw bedding and when changing from manure removal via scraper to flushing. The latter effects need cautious interpretation due to small sample sizes and differences in potential NH₃ emission driving factors. While these findings are promising, many practices had limited observations. Further studies are required to better quantify these practices’ mitigation potential for pig and cattle houses, ensuring co-benefits and trade-offs for all gases are captured, with data collated in the DATAMAN database for improved quantification.

本研究利用全球DATAMAN-Housing数据库分析了管理对猪舍和牛舍甲烷（CH4）、氧化亚氮（N2O）和氨（NH3）排放的影响。排放因子（EF）数据按动物亚类和粪便类型分组，并建立了带有住房实践变量（如空气洗涤器、垫层材料）的单变量线性模型，并确定了控制值。对这些变量进行检验，看是否对平均EFs有统计学上的显著影响。在封闭式猪舍系统中，用木屑为基础的垫层代替稻草可使EFCH4平均降低46% （p = 0.012）。在深坑生长猪舍中，在出风口安装生物空气洗涤器可使EFNH3平均降低24% （p = 0.031）。有限的N2O数据显示，EFN2O增加了57% (p <；0.0001)，但由于观测数量少，需要仔细解释。使用刮刀从板条地板下方的坑中频繁清除浆液，使平均EFNH3降低了53% （p = 0.011）。在小隔间和松散的牛舍中，当使用木片代替稻草垫层，以及从通过刮刀清除粪便改为通过冲洗时，平均EFNH3下降。由于样本量小，且潜在NH3排放驱动因子存在差异，后一种效应需要谨慎解释。虽然这些发现很有希望，但许多实践的观察结果有限。需要进一步的研究来更好地量化这些做法对猪舍和牛舍的减排潜力，确保捕获所有气体的共同效益和权衡，并将数据整理在DATAMAN数据库中，以改进量化。

{"title":"Mitigation of greenhouse gas and ammonia emissions due to livestock housing management practices: Analysis of the DATAMAN database","authors":"Vasileios Anestis , Wajid Umar , Federico Dragoni , Tony J. van der Weerden , Mélynda Hassouna , Alasdair Noble , Thomas Bartzanas , Barbara Amon","doi":"10.1016/j.biosystemseng.2025.104260","DOIUrl":"10.1016/j.biosystemseng.2025.104260","url":null,"abstract":"<div><div>This study analysed management effects on methane (CH<sub>4</sub>), nitrous oxide (N<sub>2</sub>O) and ammonia (NH<sub>3</sub>) emissions from pig and cattle houses, using the global DATAMAN-Housing database. Emission factor (EF) data was grouped by animal sub-category and manure type, and single-variable linear models of EF with housing practice variables (e.g. air scrubber, bedding material, with identified control values) were developed. These variables were examined for statistically significant effects on mean EFs. In closed growing pig housing systems, replacing straw with sawdust-based bedding reduced (p = 0.012) the mean EF<sub>CH4</sub> by 46 %. In deep pit growing pig houses, installing biological air scrubbers at the air outlet decreased the mean EF<sub>NH3</sub> (p = 0.031) by 24 %. Limited N<sub>2</sub>O data showed an increased EF<sub>N2O</sub> by 57 % (p < 0.0001), but careful interpretation is required due to the low number of observations. Frequent slurry removal from the pit below slatted floors using a scraper lowered (p = 0.011) the mean EF<sub>NH3</sub> by 53 %. In cubicle and loose cattle houses the mean EF<sub>NH3</sub> decreased when using woodchips instead of straw bedding and when changing from manure removal via scraper to flushing. The latter effects need cautious interpretation due to small sample sizes and differences in potential NH<sub>3</sub> emission driving factors. While these findings are promising, many practices had limited observations. Further studies are required to better quantify these practices’ mitigation potential for pig and cattle houses, ensuring co-benefits and trade-offs for all gases are captured, with data collated in the DATAMAN database for improved quantification.</div></div>","PeriodicalId":9173,"journal":{"name":"Biosystems Engineering","volume":"258 ","pages":"Article 104260"},"PeriodicalIF":5.3,"publicationDate":"2025-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144828016","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Optimal temperature setpoint control for a complete greenhouse crop cycle under growth uncertainty 生长不确定性下温室作物全周期最优温度设定值控制

IF 5.3 1区农林科学 Q1 AGRICULTURAL ENGINEERING

Biosystems Engineering

Pub Date : 2025-08-13 DOI: 10.1016/j.biosystemseng.2025.104250

Simon van Mourik , Bert van ’t Ooster , Michel Vellekoop

This paper proposes to plan crop production over a complete growing season, by solving a control problem that optimises expected net revenue under stochastic disturbances. Production specifications with payouts only at harvest time (i.e. a reward only at harvest time) and under precise weight constraints on the harvested crop were considered. A case study was conducted for lettuce production in a greenhouse under Dutch weather conditions. Optimal control policies were calculated for weather measured on three different days, three different values for energy costs, and an uncertainty analysis was carried out under varying harvest weight requirements, state dynamics uncertainty levels, initial crop weight and starting time of the production round. The optimal controller balances daily energy costs and the expected maximum harvest revenues and uses state- and time-dependent feedback to adapt its actions under uncertainty. A control policy that is not based on uncertainty, is shown to perform substantially worse, with 15% less net revenues. A control policy without dynamic feedback even lead to a loss of 19% in net revenues. The sensitivity analysis showed that these performance differences persist over large ranges in uncertainty level, harvest weight constraints, deviations from the optimal starting day, and deviations in initial crop weight. Altogether, the results indicate that dynamic feedback, and uncertainty modelling can substantially improve economic outcomes in greenhouse climate control design.

本文提出通过解决随机干扰下期望净收入最优的控制问题来规划整个生长季节的作物生产。考虑了仅在收获时支付的生产规格（即仅在收获时获得奖励）和对收获作物的精确重量限制。在荷兰气候条件下，对温室莴苣生产进行了案例研究。计算了3天不同天气、3种不同能源成本值的最优控制策略，并在不同收获重量要求、状态动态不确定性水平、初始作物重量和生产周期开始时间下进行了不确定性分析。最优控制器平衡每日能源成本和预期最大收获收益，并使用状态和时间相关的反馈来适应不确定性下的行为。不以不确定性为基础的控制政策的效果要差得多，净收入会减少15%。没有动态反馈的控制政策甚至导致净收入损失19%。敏感性分析表明，这些性能差异在不确定性水平、收获重量约束、与最佳开始日期的偏差以及初始作物重量的偏差方面持续存在较大范围。综上所述，动态反馈和不确定性建模可以显著提高温室气候控制设计的经济效益。

{"title":"Optimal temperature setpoint control for a complete greenhouse crop cycle under growth uncertainty","authors":"Simon van Mourik , Bert van ’t Ooster , Michel Vellekoop","doi":"10.1016/j.biosystemseng.2025.104250","DOIUrl":"10.1016/j.biosystemseng.2025.104250","url":null,"abstract":"<div><div>This paper proposes to plan crop production over a complete growing season, by solving a control problem that optimises expected net revenue under stochastic disturbances. Production specifications with payouts only at harvest time (i.e. a reward only at harvest time) and under precise weight constraints on the harvested crop were considered. A case study was conducted for lettuce production in a greenhouse under Dutch weather conditions. Optimal control policies were calculated for weather measured on three different days, three different values for energy costs, and an uncertainty analysis was carried out under varying harvest weight requirements, state dynamics uncertainty levels, initial crop weight and starting time of the production round. The optimal controller balances daily energy costs and the expected maximum harvest revenues and uses state- and time-dependent feedback to adapt its actions under uncertainty. A control policy that is not based on uncertainty, is shown to perform substantially worse, with 15% less net revenues. A control policy without dynamic feedback even lead to a loss of 19% in net revenues. The sensitivity analysis showed that these performance differences persist over large ranges in uncertainty level, harvest weight constraints, deviations from the optimal starting day, and deviations in initial crop weight. Altogether, the results indicate that dynamic feedback, and uncertainty modelling can substantially improve economic outcomes in greenhouse climate control design.</div></div>","PeriodicalId":9173,"journal":{"name":"Biosystems Engineering","volume":"258 ","pages":"Article 104250"},"PeriodicalIF":5.3,"publicationDate":"2025-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144828013","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Improving simulation accuracy of a drip irrigation tooth emitter: Simulation and verification of tooth tip fillet 提高滴灌齿射器仿真精度：齿尖圆角的仿真与验证

IF 5.3 1区农林科学 Q1 AGRICULTURAL ENGINEERING

Biosystems Engineering

Pub Date : 2025-08-12 DOI: 10.1016/j.biosystemseng.2025.104257

Xuefeng Qiu , Haitao Wang , Qiankun Leng , Suojun Xu , Jiandong Wang , Hongyi Yang , Xurong Mei

Numerical simulations are a reliable and commonly used research method for optimizing the structure of toothed labyrinth channels. However, simulations often involve a physical model of the sharp angle (SA) that is inconsistent with the actual emitter, which distorts the results to a certain extent. This study investigated two tooth-tip filleting methods (constant and variable angles) with five radii (r0.02–r0.10 mm) and compared them with SA, analysing impacts on hydraulic characteristics. Results showed sharp angles widened flow channels by 22.77–22.85 % versus r0.10 mm fillets, reduced emitter flow rates by 23.66–27.94 % under each pressure gradients, and overestimated energy dissipation by 40.36–43.41 %. Internally, SA increased wake vortex areas at tooth tips, Q-criterion vortex intensity was overestimated by 27.75–45.04 %, while mean flow velocity, turbulent kinetic energy (TKE), and TKE dissipation rates were overestimated by 12.66–14.10 %, 18.84–20.59 %, and 20.52–22.42 %, respectively. TKE and dissipation rate zones were overestimated by 50.51–57.96 % and 37.55–47.84 %. Hence, when the SA physical model was used to optimise the structure of the toothed labyrinth channel, in addition to the deviation from the physical structure of the emitter, there were large deviations in the simulation calculations of the flow rate, energy dissipation, and internal hydraulic characteristics. However, the filleted-tooth tip, which was consistent with the machining accuracy of the emitter, further improved the simulation accuracy. The results provide important reference values for the optimal design of drip irrigation emitters and reducing their processing costs.

数值模拟是齿状迷宫通道结构优化的一种可靠且常用的研究方法。然而，在模拟过程中往往涉及到与实际发射体不一致的锐角物理模型，这在一定程度上扭曲了模拟结果。研究了5个半径（r0.02 ~ r0.10 mm）的两种齿尖圆角方法（定角和变角），并与SA进行了比较，分析了对水力特性的影响。结果表明：与r0.10 mm角相比，陡角使流道拓宽22.77 ~ 22.85%，使发射器流量降低23.66% ~ 27.94%，使能量耗散高估40.36 ~ 43.41%。内部，SA增加了齿尖尾流面积，q准则涡强度高估了27.75 ~ 45.04%，平均流速、湍流动能（TKE）和TKE耗散率分别高估了12.66 ~ 14.10%、18.84 ~ 20.59%和20.52 ~ 22.42%。TKE和耗散率区分别高估了50.51 ~ 57.96%和37.55 ~ 47.84%。因此，当采用SA物理模型对齿状迷宫通道进行结构优化时，除了与发射器的物理结构存在偏差外，在流量、能量耗散、内部水力特性的仿真计算中也存在较大偏差。然而，与射极加工精度一致的圆角齿尖进一步提高了仿真精度。研究结果对滴灌灌头的优化设计和降低滴灌灌头的加工成本具有重要的参考价值。

{"title":"Improving simulation accuracy of a drip irrigation tooth emitter: Simulation and verification of tooth tip fillet","authors":"Xuefeng Qiu , Haitao Wang , Qiankun Leng , Suojun Xu , Jiandong Wang , Hongyi Yang , Xurong Mei","doi":"10.1016/j.biosystemseng.2025.104257","DOIUrl":"10.1016/j.biosystemseng.2025.104257","url":null,"abstract":"<div><div>Numerical simulations are a reliable and commonly used research method for optimizing the structure of toothed labyrinth channels. However, simulations often involve a physical model of the sharp angle (SA) that is inconsistent with the actual emitter, which distorts the results to a certain extent. This study investigated two tooth-tip filleting methods (constant and variable angles) with five radii (r0.02–r0.10 mm) and compared them with SA, analysing impacts on hydraulic characteristics. Results showed sharp angles widened flow channels by 22.77–22.85 % versus r0.10 mm fillets, reduced emitter flow rates by 23.66–27.94 % under each pressure gradients, and overestimated energy dissipation by 40.36–43.41 %. Internally, SA increased wake vortex areas at tooth tips, Q-criterion vortex intensity was overestimated by 27.75–45.04 %, while mean flow velocity, turbulent kinetic energy (TKE), and TKE dissipation rates were overestimated by 12.66–14.10 %, 18.84–20.59 %, and 20.52–22.42 %, respectively. TKE and dissipation rate zones were overestimated by 50.51–57.96 % and 37.55–47.84 %. Hence, when the SA physical model was used to optimise the structure of the toothed labyrinth channel, in addition to the deviation from the physical structure of the emitter, there were large deviations in the simulation calculations of the flow rate, energy dissipation, and internal hydraulic characteristics. However, the filleted-tooth tip, which was consistent with the machining accuracy of the emitter, further improved the simulation accuracy. The results provide important reference values for the optimal design of drip irrigation emitters and reducing their processing costs.</div></div>","PeriodicalId":9173,"journal":{"name":"Biosystems Engineering","volume":"258 ","pages":"Article 104257"},"PeriodicalIF":5.3,"publicationDate":"2025-08-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144827500","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Industry-scale prediction of video-derived pig body weight using efficient convolutional neural networks and vision transformers 利用高效卷积神经网络和视觉变压器对视频衍生猪体重进行工业规模预测

IF 5.3 1区农林科学 Q1 AGRICULTURAL ENGINEERING

Biosystems Engineering

Pub Date : 2025-08-05 DOI: 10.1016/j.biosystemseng.2025.104243

Ye Bi , Yijian Huang , Jianhua Xuan , Gota Morota

Accurate pig body weight measurement is critical for pig growth, health, and marketing. Although there is a growing trend towards the use of computer vision approaches for pig body weight prediction, their validation with large-scale data collected in commercial environments is still limited. Therefore, the main objective of this study was to predict pig body weight collected at multiple timepoints from a commercial environment using efficient convolutional neural networks and efficient vision transformers. Top-view videos were collected from over 600 pigs at six time points over three months. Scale-based body weight records were simultaneously recorded by a digital weighing system. An automated video conversion pipeline and fine-tuned YOLOv8 were applied to preprocess the raw depth videos. Two families of lightweight deep neural networks, MobileNet and MobileViT, were initialised with the pre-trained weights from ImageNet and customised to predict pig body weight directly from depth images. Two cross-validation strategies were used: single time point random subsampling and time series forecasting with a sparse design considering limited budget scenarios. In single time point random subsampling, the best prediction mean absolute percentage error for each time point was 4.71%, 3.80%, 3.08%, 5.60%, 3.42%, and 3.77%, respectively. On average, the MobileViT-S model produced the best prediction mean absolute percentage error. In time series forecasting, although a sparse design resulted in some performance loss compared to the full design, the use of ViT models mitigated this degradation. These results suggest that efficient deep learning-based supervised learning models are a promising approach for predicting pig body weight from industry-scale depth video data.

准确的猪体重测量对猪的生长、健康和销售至关重要。尽管使用计算机视觉方法预测猪体重的趋势越来越多，但它们在商业环境中收集的大规模数据的验证仍然有限。因此，本研究的主要目的是利用高效卷积神经网络和高效视觉变压器，预测在商业环境中多个时间点采集的猪体重。在三个月内的六个时间点收集了600多头猪的俯视图视频。数字称重系统同时记录基于体重计的体重记录。采用自动视频转换流水线和微调YOLOv8对原始深度视频进行预处理。MobileNet和MobileViT这两个轻量级深度神经网络家族，使用ImageNet预训练的权重进行初始化，并进行定制，直接从深度图像中预测猪的体重。交叉验证策略采用单时间点随机子抽样和考虑有限预算情景的稀疏设计时间序列预测。在单时间点随机子抽样中，各时间点的最佳预测平均绝对百分比误差分别为4.71%、3.80%、3.08%、5.60%、3.42%和3.77%。平均而言，MobileViT-S模型产生了最好的预测平均绝对百分比误差。在时间序列预测中，尽管与完整设计相比，稀疏设计会导致一些性能损失，但ViT模型的使用减轻了这种退化。这些结果表明，高效的基于深度学习的监督学习模型是一种很有前途的方法，可以从工业规模的深度视频数据中预测猪的体重。

{"title":"Industry-scale prediction of video-derived pig body weight using efficient convolutional neural networks and vision transformers","authors":"Ye Bi , Yijian Huang , Jianhua Xuan , Gota Morota","doi":"10.1016/j.biosystemseng.2025.104243","DOIUrl":"10.1016/j.biosystemseng.2025.104243","url":null,"abstract":"<div><div>Accurate pig body weight measurement is critical for pig growth, health, and marketing. Although there is a growing trend towards the use of computer vision approaches for pig body weight prediction, their validation with large-scale data collected in commercial environments is still limited. Therefore, the main objective of this study was to predict pig body weight collected at multiple timepoints from a commercial environment using efficient convolutional neural networks and efficient vision transformers. Top-view videos were collected from over 600 pigs at six time points over three months. Scale-based body weight records were simultaneously recorded by a digital weighing system. An automated video conversion pipeline and fine-tuned YOLOv8 were applied to preprocess the raw depth videos. Two families of lightweight deep neural networks, MobileNet and MobileViT, were initialised with the pre-trained weights from ImageNet and customised to predict pig body weight directly from depth images. Two cross-validation strategies were used: single time point random subsampling and time series forecasting with a sparse design considering limited budget scenarios. In single time point random subsampling, the best prediction mean absolute percentage error for each time point was 4.71%, 3.80%, 3.08%, 5.60%, 3.42%, and 3.77%, respectively. On average, the MobileViT-S model produced the best prediction mean absolute percentage error. In time series forecasting, although a sparse design resulted in some performance loss compared to the full design, the use of ViT models mitigated this degradation. These results suggest that efficient deep learning-based supervised learning models are a promising approach for predicting pig body weight from industry-scale depth video data.</div></div>","PeriodicalId":9173,"journal":{"name":"Biosystems Engineering","volume":"257 ","pages":"Article 104243"},"PeriodicalIF":5.3,"publicationDate":"2025-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144771589","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0