This paper proposes a cascade deep convolutional neural network to address the loosening detection problem of bolts on axlebox covers.Firstly,an SSD network based on ResNet50 and CBAM module by improving bolt image fe...This paper proposes a cascade deep convolutional neural network to address the loosening detection problem of bolts on axlebox covers.Firstly,an SSD network based on ResNet50 and CBAM module by improving bolt image features is proposed for locating bolts on axlebox covers.And then,theA2-PFN is proposed according to the slender features of the marker lines for extracting more accurate marker lines regions of the bolts.Finally,a rectangular approximationmethod is proposed to regularize themarker line regions asaway tocalculate the angle of themarker line and plot all the angle values into an angle table,according to which the criteria of the angle table can determine whether the bolt with the marker line is in danger of loosening.Meanwhile,our improved algorithm is compared with the pre-improved algorithmin the object localization stage.The results show that our proposed method has a significant improvement in both detection accuracy and detection speed,where ourmAP(IoU=0.75)reaches 0.77 and fps reaches 16.6.And in the saliency detection stage,after qualitative comparison and quantitative comparison,our method significantly outperforms other state-of-the-art methods,where our MAE reaches 0.092,F-measure reaches 0.948 and AUC reaches 0.943.Ultimately,according to the angle table,out of 676 bolt samples,a total of 60 bolts are loose,69 bolts are at risk of loosening,and 547 bolts are tightened.展开更多
Under the conditions of straw returning operation,there are three major technical bottlenecks in the Phaeozem region of northeast China,namely low stubble breaking rate,poor tillage depth consistency,and high fuel con...Under the conditions of straw returning operation,there are three major technical bottlenecks in the Phaeozem region of northeast China,namely low stubble breaking rate,poor tillage depth consistency,and high fuel consumption.In this research,a bionic stubble-deep loosening combined tillage machine(BSD)was designed through bionic prototype analysis,coupled bionic analysis,coupled bionic design,theoretical analysis and application of intelligent control techniques.It consists of a bionic stubble breaking device and a bionic self-excited vibratory deep loosening device.Based on the unique biting pattern of locust mouthparts on maize rootstocks,the bionic stubble breaking device adopted a new multi-segment serrated bionic structure and a symmetrical rotational motion,which could significantly increase the stubble breaking rate(p<0.05)and reduce the resistance to stubble breaking operations(p<0.05).Based on the unique biology of the hare’s paws,toes and nails,the bionic self-excited vibration deep loosening device adopted a new series-parallel composite bionic elastic system and an intelligent tilling depth control system with a fuzzy algorithm,which significantly improved the tilling depth consistency(p<0.05).The operational performance of the BSD was verified at different operating speeds through comparative experiments and reveals the mechanism of its excellent performance through theoretical analysis.The final experiment results showed that,at the same operating speed,the BSD improved the stubble breaking rate by 9.62%and 10.67%,reduced the stubble breaking torque by 28 N·m and 33 N·m,reduced the tillage depth coefficient of variation by 12.73%and 13.48%,and reduced the specific fuel consumption by 36 g/km·h and 40 g/km·h compared to the two most common models.The operating performance of the three kinds of machines will decrease with the increase of operating speed,and the BSD has the least decrease.展开更多
文摘This paper proposes a cascade deep convolutional neural network to address the loosening detection problem of bolts on axlebox covers.Firstly,an SSD network based on ResNet50 and CBAM module by improving bolt image features is proposed for locating bolts on axlebox covers.And then,theA2-PFN is proposed according to the slender features of the marker lines for extracting more accurate marker lines regions of the bolts.Finally,a rectangular approximationmethod is proposed to regularize themarker line regions asaway tocalculate the angle of themarker line and plot all the angle values into an angle table,according to which the criteria of the angle table can determine whether the bolt with the marker line is in danger of loosening.Meanwhile,our improved algorithm is compared with the pre-improved algorithmin the object localization stage.The results show that our proposed method has a significant improvement in both detection accuracy and detection speed,where ourmAP(IoU=0.75)reaches 0.77 and fps reaches 16.6.And in the saliency detection stage,after qualitative comparison and quantitative comparison,our method significantly outperforms other state-of-the-art methods,where our MAE reaches 0.092,F-measure reaches 0.948 and AUC reaches 0.943.Ultimately,according to the angle table,out of 676 bolt samples,a total of 60 bolts are loose,69 bolts are at risk of loosening,and 547 bolts are tightened.
基金The authors acknowledge that the research was financially supported by the National Key Research and Development Project of China(Grant No.2017YFD0700701)the Natural Science Foundation of China(Grant No.52075215)+1 种基金the Science and Technology Development Plan Project of Jilin Province(Grant No.20190301023NY,20190701055GH and 20200404008YY)China Postdoctoral Science Foundation(Grant No.2020M67085).
文摘Under the conditions of straw returning operation,there are three major technical bottlenecks in the Phaeozem region of northeast China,namely low stubble breaking rate,poor tillage depth consistency,and high fuel consumption.In this research,a bionic stubble-deep loosening combined tillage machine(BSD)was designed through bionic prototype analysis,coupled bionic analysis,coupled bionic design,theoretical analysis and application of intelligent control techniques.It consists of a bionic stubble breaking device and a bionic self-excited vibratory deep loosening device.Based on the unique biting pattern of locust mouthparts on maize rootstocks,the bionic stubble breaking device adopted a new multi-segment serrated bionic structure and a symmetrical rotational motion,which could significantly increase the stubble breaking rate(p<0.05)and reduce the resistance to stubble breaking operations(p<0.05).Based on the unique biology of the hare’s paws,toes and nails,the bionic self-excited vibration deep loosening device adopted a new series-parallel composite bionic elastic system and an intelligent tilling depth control system with a fuzzy algorithm,which significantly improved the tilling depth consistency(p<0.05).The operational performance of the BSD was verified at different operating speeds through comparative experiments and reveals the mechanism of its excellent performance through theoretical analysis.The final experiment results showed that,at the same operating speed,the BSD improved the stubble breaking rate by 9.62%and 10.67%,reduced the stubble breaking torque by 28 N·m and 33 N·m,reduced the tillage depth coefficient of variation by 12.73%and 13.48%,and reduced the specific fuel consumption by 36 g/km·h and 40 g/km·h compared to the two most common models.The operating performance of the three kinds of machines will decrease with the increase of operating speed,and the BSD has the least decrease.
