Finite Frequency Vibration Suppression for Space Flexible Structures in Tip Position Control

Shidong Xu,Guanghui Sun,Zhan Li 제어·로봇·시스템학회 2018 International Journal of Control, Automation, and Vol.16 No.3

This paper presents a novel control strategy for the tip position and vibration control of a class of space flexible structures. The proposed control algorithm consists of finite frequency H∞ vibration control technique and fractional-order PDv control technique. More specially, a new finite frequency H∞ controller working in the inner feedback loop is proposed to suppress vibration modes and external disturbances, and a new fractional-order PDv controller is developed in the outer feedback loop to guarantee the desired position tracking performance. Compared with conventional methods, the proposed one could achieve better control results. Finally, an illustrative example is presented to demonstrate the robustness and effectiveness of the proposed composite control strategy

Three-dimensional Interval Identification of Permanent Magnet Spherical Motor Based on Improved Deep Neural Network

Zhao Shidong,Che Yanbo,Li Hongfeng 대한전기학회 2024 Journal of Electrical Engineering & Technology Vol.19 No.1

The accurate identifcation of rotor position is a prerequisite for the closed-loop control of spherical motors. In this paper, the spherical interval of the rotor of the permanent magnet spherical motor was partitioned by the Quaternary Triangular Mesh (QTM). According to diferent levels of division, the sphere was partitioned by triangles of diferent sizes. After the partition, a Hall sensor was placed at the vertex of the spherical triangle obtained by the level 3 partition. After the rotor starts rotating, the triangular intervals where the rotor poles were located could be identifed according to the output of the sensor array. To achieve more precise identifcation of spherical intervals, the paper proposes a strategy that combines Deep Neural Networks (DNN) with rotor spherical interval identifcation. However, due to the small change rate of the magnetic feld near the rotor pole, it was easy to get the local optimal solution by the neural network algorithm. To solve this problem, the parameters of DNN were optimized by the Krill Herd (KH) algorithm. Through 50 simulations, the classifcation accuracy of KH-DNN reached 97.7%, and the average Mean Squared Error (MSE) value was 0.0165, which was better than the original DNN algorithm and the control group algorithm. Finally, the algorithm was verifed by a rotation experiment. The R2 value of the trajectory remained above 0.99, and high classifcation accuracy was guaranteed. The algorithm proposed in the paper provided a new research idea for the position identifcation of spherical motors.

In vitro Inhibition of Fungal Root-Rot Pathogens of Panax notoginseng by Rhizobacteria

Rongjun Guo,Xingzhong Liu,Shidong Li,Zuoqing Miao 한국식물병리학회 2009 Plant Pathology Journal Vol.25 No.1

The rhizobacteria of Panax notoginseng were isolated from six sites in Yanshan, Maguan and Wenshan Counties, Yunnan Province of China, and their antagonistic activity against P. notoginseng root-rot fungal pathogens was determined. Of the 574 rhizobacteria isolated, 5.8% isolates were antagonistic in vitro to at least one of the five pathogens, Cylindrocarpon didynum, Fusarium solani, Phytophthora cactorum, Phoma herbarum, and Rhizoctonia solani. The number of rhizo bacteria and the number that inhibited fungi differed depending on sampling sites and isolation methods. Rhizobacteria isolated from the site in Yanshan and Maguan showed more antagonistic effect than them in Wenshan. Heat treatment of rhizosphere soil at 80oC for 20 min scaled the antagonists up to 14.0%. Antagonistic bacteria in the roots proportioned 3.9% of the total isolates. The most antagonistic isolates 79-9 and 81-4 are Bacillus subtilis based on their 16S rDNA sequence and biochemical and physiological characteristics. Identification and evaluation of antagonistic bacteria against P. notoginseng root-rot pathogens in the main planting areas improved our understanding of their distribution in rhizosphere soil. Furthermore these results indicated that the interactions between biocontrol agent and soil microbes should be seriously considered for the successful survival and biocontrol efficacy of the agents in soil.

In vitro Inhibition of Fungal Root-Rot Pathogens of Panax notoginseng by Rhizobacteria

Guo, Rongjun,Liu, Xingzhong,Li, Shidong,Miao, Zuoqing The Korean Society of Plant Pathology 2009 Plant Pathology Journal Vol.25 No.1

The rhizobacteria of Panax notoginseng were isolated from six sites in Yanshan, Maguan and Wenshan Counties, Yunnan Province of China, and their antagonistic activity against P. notoginseng root-rot fungal pathogens was determined. Of the 574 rhizobacteria isolated, 5.8% isolates were antagonistic in vitro to at least one of the five pathogens, Cylindrocarpon didynum, Fusarium solani, Phytophthora cactorum, Phoma herbarum, and Rhizoctonia solani. The number of rhizo bacteria and the number that inhibited fungi differed depending on sampling sites and isolation methods. Rhizobacteria isolated from the site in Yanshan and Maguan showed more antagonistic effect than them in Wenshan. Heat treatment of rhizosphere soil at $80^{\circ}C$ for 20 min scaled the antagonists up to 14.0%. Antagonistic bacteria in the roots proportioned 3.9% of the total isolates. The most antagonistic isolates 79-9 and 81-4 are Bacillus subtilis based on their 168 rDNA sequence and biochemical and physiological characteristics. Identification and evaluation of antagonistic bacteria against P. notoginseng root-rot pathogens in the main planting areas improved our understanding of their distribution in rhizosphere soil. Furthermore these results indicated that the interactions between biocontrol agent and soil microbes should be seriously considered for the successful survival and biocontrol efficacy of the agents in soil.

Effect of Low-Temperature Heat Treatment on PM2.5 Adsorption Properties of GO Films

Weiwu Zou,Baoshan Gu,Shiqing Sun,Shidong Wang,Xin Li,Haoqi Zhao,Peiyan Yang 성균관대학교(자연과학캠퍼스) 성균나노과학기술원 2019 NANO Vol.15 No.01

To explore the mechanism of GO acting on PM 2.5, a graphene oxide (GO) film was prepared via a spraying method for air purification. The effects of different media, temperature and heat treatment times on the adsorption of PM 2.5 on GO film were investigated. The morphology, composition and structure of GO materials were characterized by scanning electron microscopy (SEM), electron spectroscopy (EDS), X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), infrared spectroscopy (FT-IR) and Raman spectroscopy. When the vacuum heat-treatment temperature is below 80 ℃ and the atmospheric heat-treatment temperature is below 100 ℃, the air purification performance of the film does not change significantly. With the increase in the vacuum heat-treatment temperature, the removal efficiency of PM 2.5 by GO film decreases gradually from 95% to 83%. At different times, the vacuum heat treatment increases with time, and the film removal rate shows a downward trend. As the heat-treatment temperature and time increase, a certain redox reaction occurs in the GO, and the air purification performance decreases. At a temperature of 120 ℃ and a time of 8 h, the removal rate drops to 81.68%. The adsorption of PM 2.5 by GO film mainly relies on the action of oxygen-containing functional groups.

Microblog Sentiment Analysis Method Based on Spectral Clustering

Shi Dong,Xingang Zhang,Ya Li 한국정보처리학회 2018 Journal of information processing systems Vol.14 No.3

This study evaluates the viewpoints of user focus incidents using microblog sentiment analysis, which hasbeen actively researched in academia. Most existing works have adopted traditional supervised machinelearning methods to analyze emotions in microblogs; however, these approaches may not be suitable inChinese due to linguistic differences. This paper proposes a new microblog sentiment analysis method thatmines associated microblog emotions based on a popular microblog through user-building combined withspectral clustering to analyze microblog content. Experimental results for a public microblog benchmarkcorpus show that the proposed method can improve identification accuracy and save manually labeled timecompared to existing methods.