Chemical cleaning has been reported as the best method to restore ceramic filters flow rate by removing fouling agents. Even though there are several chemicals that can be used as cleaning agents, sodium hydroxide (Na...Chemical cleaning has been reported as the best method to restore ceramic filters flow rate by removing fouling agents. Even though there are several chemicals that can be used as cleaning agents, sodium hydroxide (NaOH) has been widely used as a cleaning agent. Literature reports that main factors of this cleaning are sodium hydroxide concentration, treatment temperature and contact time. However, the most significant factors have not been determined nor the interactions between them. The aim of this study was to determine the most significant parameter and the interactions between sodium hydroxide concentration, treatment temperature and contact time. This was done using Box-Behnken experimental design. The results, based on ANOVA analyses, showed that temperature is the most significant factor and that interaction between sodium hydroxide concentration and treatment temperature is the most significant interaction.展开更多
Based on the basic working principle of the chamber filter press, this paper analyzes some problems that often occur during its operation, and at the same time, combining with the causes of the failures, puts forward ...Based on the basic working principle of the chamber filter press, this paper analyzes some problems that often occur during its operation, and at the same time, combining with the causes of the failures, puts forward some corresponding methods. While improving the use effect of the box filter press, it can also play a role in promoting the growth of economic benefits of enterprises.展开更多
多目标跟踪技术对猪只精细化养殖具有重要意义。针对饲养环境差异、猪只的快速移动以及群猪之间的频繁遮挡带来的多目标跟踪挑战,该研究提出了一种基于Byte的生猪多目标跟踪算法UKFTrack。首先,构建了一个采用定向边界框(oriented bound...多目标跟踪技术对猪只精细化养殖具有重要意义。针对饲养环境差异、猪只的快速移动以及群猪之间的频繁遮挡带来的多目标跟踪挑战,该研究提出了一种基于Byte的生猪多目标跟踪算法UKFTrack。首先,构建了一个采用定向边界框(oriented bounding box,OBB)标注的多样化数据集,涵盖了猪只多种运动模式以及不同饲养场景和猪群密度;其次,引入了无迹卡尔曼滤波以更好地适配OBB标注,并对传统的状态向量进行扩展,新增了角度和角速度参数,设计了残差函数处理角度变量以避免直接相减所造成的误差。最后,提出了一种多阶段匹配策略,通过多次轨迹关联和补充匹配机制,确保在遮挡严重或剧烈运动的情况下,仍能保持对目标的持续跟踪。试验结果表明,在白天重度密集、白天极度密集、夜间重度密集和夜间极度密集4种复杂场景下,UKFTrack的高阶跟踪精度(higher order tracking accuracy,HOTA)分别为96.10%、83.10%、76.50%和84.00%,IDF1得分(identification F1 score)分别为95.70%、78.20%、70.10%和77.60%。相较于StrongSORT,UKFTrack的HOTA分别提高了1.2、13.3、5.9和6.3个百分点,IDF1分别提高了0.1、10.9、5.4和7.4个百分点。因此,该研究提出的跟踪算法能实现复杂环境下群体生猪的准确跟踪,且展现出较强的鲁棒性,能为实际应用中猪只行为与健康监测提供可靠的技术支持。展开更多
为了解决工厂烟草滤嘴纸盒码垛过程中易发生错误的问题,并提高码垛效率,提出了一种基于改进SGBM(Semi-Global Block Matching)算法与轻量化YOLOv5s网络结构相结合的识别检测与定位一体化方法。首先,采用MobileNetv3作为YOLOv5s的主干网...为了解决工厂烟草滤嘴纸盒码垛过程中易发生错误的问题,并提高码垛效率,提出了一种基于改进SGBM(Semi-Global Block Matching)算法与轻量化YOLOv5s网络结构相结合的识别检测与定位一体化方法。首先,采用MobileNetv3作为YOLOv5s的主干网络,以降低模型的复杂度、提高检测速率;其次,将AD-Census(Absolute Differences-Census)代价计算方法应用于SGBM算法中,以提高测距精度。实验结果显示,在自建数据集上,该方法获得的mAP(mean Average Precision)达到91.2%,帧率FPS(Frames Per Second)达到30.8,可以准确识别出纸盒的类别与位置,可为烟草滤嘴盒自动码垛机器人提供视觉系统技术支持。展开更多
文摘Chemical cleaning has been reported as the best method to restore ceramic filters flow rate by removing fouling agents. Even though there are several chemicals that can be used as cleaning agents, sodium hydroxide (NaOH) has been widely used as a cleaning agent. Literature reports that main factors of this cleaning are sodium hydroxide concentration, treatment temperature and contact time. However, the most significant factors have not been determined nor the interactions between them. The aim of this study was to determine the most significant parameter and the interactions between sodium hydroxide concentration, treatment temperature and contact time. This was done using Box-Behnken experimental design. The results, based on ANOVA analyses, showed that temperature is the most significant factor and that interaction between sodium hydroxide concentration and treatment temperature is the most significant interaction.
文摘Based on the basic working principle of the chamber filter press, this paper analyzes some problems that often occur during its operation, and at the same time, combining with the causes of the failures, puts forward some corresponding methods. While improving the use effect of the box filter press, it can also play a role in promoting the growth of economic benefits of enterprises.
文摘多目标跟踪技术对猪只精细化养殖具有重要意义。针对饲养环境差异、猪只的快速移动以及群猪之间的频繁遮挡带来的多目标跟踪挑战,该研究提出了一种基于Byte的生猪多目标跟踪算法UKFTrack。首先,构建了一个采用定向边界框(oriented bounding box,OBB)标注的多样化数据集,涵盖了猪只多种运动模式以及不同饲养场景和猪群密度;其次,引入了无迹卡尔曼滤波以更好地适配OBB标注,并对传统的状态向量进行扩展,新增了角度和角速度参数,设计了残差函数处理角度变量以避免直接相减所造成的误差。最后,提出了一种多阶段匹配策略,通过多次轨迹关联和补充匹配机制,确保在遮挡严重或剧烈运动的情况下,仍能保持对目标的持续跟踪。试验结果表明,在白天重度密集、白天极度密集、夜间重度密集和夜间极度密集4种复杂场景下,UKFTrack的高阶跟踪精度(higher order tracking accuracy,HOTA)分别为96.10%、83.10%、76.50%和84.00%,IDF1得分(identification F1 score)分别为95.70%、78.20%、70.10%和77.60%。相较于StrongSORT,UKFTrack的HOTA分别提高了1.2、13.3、5.9和6.3个百分点,IDF1分别提高了0.1、10.9、5.4和7.4个百分点。因此,该研究提出的跟踪算法能实现复杂环境下群体生猪的准确跟踪,且展现出较强的鲁棒性,能为实际应用中猪只行为与健康监测提供可靠的技术支持。
文摘为了解决工厂烟草滤嘴纸盒码垛过程中易发生错误的问题,并提高码垛效率,提出了一种基于改进SGBM(Semi-Global Block Matching)算法与轻量化YOLOv5s网络结构相结合的识别检测与定位一体化方法。首先,采用MobileNetv3作为YOLOv5s的主干网络,以降低模型的复杂度、提高检测速率;其次,将AD-Census(Absolute Differences-Census)代价计算方法应用于SGBM算法中,以提高测距精度。实验结果显示,在自建数据集上,该方法获得的mAP(mean Average Precision)达到91.2%,帧率FPS(Frames Per Second)达到30.8,可以准确识别出纸盒的类别与位置,可为烟草滤嘴盒自动码垛机器人提供视觉系统技术支持。
