Ji F.,Shijiazhuang Mechanical Engineering College |
Sun S.,Shijiazhuang Mechanical Engineering College |
Wang C.,Shijiazhuang Mechanical Engineering College |
Zhang H.,China Institute of Technology |
Liu D.,Unit 6398.1
Insight: Non-Destructive Testing and Condition Monitoring | Year: 2010
The processing of magnetic flux leakage signals is a key element in the MFL inspection technique and guarantees for the implementation of quantitative testing of pipelines. A de-noising algorithm - adaptive fuzzy lifting wavelet transform - is presented to solve the problem of noise reduction in MFL signals. According to the theory and characteristics of the lifting wavelet transform, the improved algorithm is proposed by using an adaptive algorithm. The problem of nonlinearity caused by the adaptive algorithm is solved by using an update first lifting scheme. To verify the effectiveness of the improved lifting scheme, a fuzzy threshold filter algorithm is applied to the noise reduction of the MFL signals. The results show that the improved lifting scheme has achieved better noise reduction than that achieved by traditional wavelet transform. It is a feasible way to process MFL inspection signals.
Donggen C.,Shijiazhuang Mechanical Engineering College |
Jiangtao X.,Unit 63981 |
Hongwei L.,Unit 63981 |
Kaibo C.,Shijiazhuang Mechanical Engineering College
Proceedings of 2016 IEEE Information Technology, Networking, Electronic and Automation Control Conference, ITNEC 2016 | Year: 2016
In actual field test environments, unstable CCD measuring spot center position, the total appears slight shaking, so that measurements and calculated spot position and displacement there is a big error. This paper studies the field environments offline Array CCD measurement technology characteristics, analyzes the key sources of interference affecting the measurement accuracy linear CCD, combined with the test environment design solutions. © 2016 IEEE.
Tian X.,Ocean University of China |
Liu Y.,China University of Petroleum - East China |
Deng W.,Unit 63981 |
Liu G.,Ocean University of China
International Journal of Advanced Manufacturing Technology | Year: 2015
Electric arc cutting process parameters play a very significant role in determining the cutting effect. Sensitivity analysis can be utilized to identify the process parameters exerting the most influence on the cutting effect and to know the parameters that must be most carefully controlled. Experiment data analysis and finite element method are both introduced to carry out sensitivity analysis based on the response surface methodology. Changeable process parameters such as workpiece thickness, cutting current, and electrode diameter are used as design variables. Cutting hole geometry in experiment part is considered as the response, while the response for PDS is the simulation temperature value by finite element method. The results of two methods both show that a change in process parameters affects the cutting characteristics. The arc cutting process is most sensitive to the cutting current, less sensitive to electrode diameter, and least sensitive to workpiece thickness. It also reveals that experiment data analysis obtains the detail numerical results, and PDS gives an intuitive analysis result without much trial and error. © 2014, Springer-Verlag London.
Huang Y.,Unit 63981 |
Bai H.,Unit 63981 |
Feng J.,Unit 63981 |
Chen J.,Unit 63981
Proceedings - 2013 4th International Conference on Digital Manufacturing and Automation, ICDMA 2013 | Year: 2013
As the new intelligent method was applied constantly to the fault predication field, the technology of fault predication has already become the key direction of electronic equipment support studies. On the basis of summarizing several kinds of more common fault predication method modernly, Support Vector Regression (SVR) was introduced. To avoid the blind establishment of the parameter, this study proposes intelligent genetic algorithms for optimizing the SVR's parameters, then the SVR model which had been set up was apply to a type of electronic equipment fault prediction. Finally, we adopt the number of a set of equipment condition monitoring data to verify the SVR model. The experimental result demonstrated that SVR model can predict the radar fault effectively. © 2013 IEEE.
Lian W.-H.,Unit 63981 |
Wang Y.-J.,Unit 63981 |
Yang X.-L.,Unit 63981 |
Fan Z.-Q.,Unit 63981
Guangxue Jingmi Gongcheng/Optics and Precision Engineering | Year: 2015
How to capture the directions and distribution of refraction stars in refraction starlight navigation was researched. To obtain the information of the refraction stars, a method to analyze the distribution of refraction stars on standard orbit was proposed. By proposed method, the direction of refractive starlight from a spacecraft at a special position could be obtained by adjusting the attitude of a star sensitive sensor. On the basis of star information gotten in specific time interval, three kinds of influences on refraction stars' number and distribution were calculated according to the geometric positional relationship of the sun, the moon and the earth. Comparing the stars' number and distribution before and after the interferences, a navigation simulation was analyzed. The experimental results show that the largest source of interference is the earth, followed by the sun, the moon. Under the three interferences, refraction navigation, blank section may exist for the spacecraft. In the range of navigation blank section navigation error increases rapidly, instantaneous position error in r direction could be 20 times larger than before. It is verified that the method proposed to analyze interference on refraction starlight navigation has important practical significance, especially in design of navigation methods, adjusting star sensor's attitude, and predicting navigation blank section. © 2015, Chinese Academy of Sciences. All right reserved.