This paper is a collection of 10 graduation theses related to parallel processing topics, including 5 journal papers and 5 degree theses, which provides a reference for those related to parallel processing topics to write graduation theses.
1. [journal paper] design of fault diagnosis system for electromechanical special equipment based on MapReduce parallel processing
Journal: computer measurement and control, No. 002, 2021
Abstract: Aiming at the problem of large error in the detection results of DC grounding fault detection system, a design of fault diagnosis system for electromechanical special equipment based on MapReduce parallel processing is proposed; According to the overall architecture of the system, the hardware structure is divided into fault detection and display unit and data processing and transmission unit; Rectifier current, using diode rectifier device to design current collection fault detection and indication circuit; The multi-layer differential circuit is used to obtain the pulse signal, and the fault detector at the low-level voltage position is used as the detection point to design the current mutation detection module; DH08 switch state detection module with 8 AC inputs is used to detect the power failure of the equipment; 6AU1410-0AB00-0AA0 Siemens alarm module is selected to alarm the fault point; Design MapReduce execution process, analyze four MapReduce job training processes, and calculate the frequency value of data attribute feature words in each fault class, so as to complete fault diagnosis; Taking the bearing fault as an example, the experimental verification and analysis are carried out; According to the experimental test results, the difference between the system and the actual waveform is small. The time point waveform change of phase a, phase B and phase C current short-circuit fault diagnosis is basically consistent with the actual curve and fluctuates near 0A, indicating that this method has accurate detection results and can provide equipment support for the wide application of electromechanical special equipment
Keywords: MapReduce; Parallel processing; Electromechanical special equipment; fault diagnosis
---------------------------------------------------------------------------------------------------
2. [journal papers] streaming data parallel processing method based on sliding window
Journal: heavy machinery, No. 001, 2021
Abstract: with the development of Internet of things and industrial Internet, a large number of manufacturing data can be collected in the manufacturing process of complex heavy equipment In order to monitor the manufacturing process in real time, it is necessary to improve the processing speed of stream data By analyzing the existing stream data processing engine, combined with the idea of partition-based and time-based sliding window, this paper constructs a sliding window model integrating the results of sub stream processing in unit time cycle, and adopts parallel real-time operation mode to realize real-time stream data processing The streaming data processing platform is designed based on Lambda architecture, and the data processing is divided into batch processing platform and incremental processing platform to meet the diversified needs of data Finally, through throughput and delay analysis, it is verified that the parallel stream data processing method based on sliding window can meet the real-time requirements of real-time monitoring of complex heavy equipment
Keywords: real time monitoring; Stream data; Sliding window; Parallelization
Link: https://www.zhangqiaokeyan.com/academic-journal-cn_heavy-machinery_thesis/0201290093260.html
---------------------------------------------------------------------------------------------------
3. [journal papers] computer image retrieval technology based on parallel processing
Journal: information technology and informatization, No. 005, 2021
Abstract: in view of the rapid development of computer, people's demand for image information increases greatly. In order to extract valuable and useful data from massive data, we must develop and study new image retrieval technology Combined with theory and practice, this paper briefly expounds the computer retrieval technology, then analyzes the application path of parallel processing technology and computer image retrieval technology, and finally puts forward the implementation method of this technology The analysis results show that using advanced parallel processing technology in image retrieval can greatly improve the quality and efficiency of image processing, and extract valuable graphic information from massive graphic database, which is worthy of wide application
Keywords: parallel processing; computer; Image retrieval; application
---------------------------------------------------------------------------------------------------
4. [journal paper] parallel processing of RPC correction of wide range remote sensing image supported by GPU
Journal: Radio Engineering, 2021, issue 008
Abstract: RPC correction is an important link in the production process of optical remote sensing images, which converts remote sensing images from image coordinates to geographical coordinates In order to shorten the time of RPC correction process and realize the near real-time processing of images, the current first-class GPU(NVIDIA TITAN RTX) is used to realize the parallel RPC correction of wide-ranging remote sensing images. GPU allocates threads for each pixel to be calculated, completes the calculation processes such as coordinate transformation and resampling, and realizes the parallel processing; The efficiency of RPC correction process is compared with that of CPU operation The experimental results show that compared with the RPC correction without block processing by CPU, the parallel speedup ratio of GPU is related to the amount of remote sensing image data. The larger the amount of data, the higher the parallel speedup ratio; The speedup ratio of parallel processing on gaogao-2 multispectral image and panchromatic image is 3.71 and 8.27 respectively, and the overall speedup ratio is 1.59 and 5.07 respectively This method can significantly speed up the RPC correction rate of remote sensing images
Keywords: GPU parallel; RPC correction; Wide range remote sensing image
Link: https://www.zhangqiaokeyan.com/academic-journal-cn_radio-engineering_thesis/0201290366443.html
---------------------------------------------------------------------------------------------------
5. [journal paper] an interference detection method of time domain frequency domain parallel processing
Journal: radio communication technology, No. 004, 2021
Abstract: Aerospace TT & C link usually faces a complex interference environment composed of multiple interference sources and interference modes. According to the interference style and interference characteristics of TT & C link and the actual demand of fast and efficient interference detection in broadband, the interference detection method of TT & C link in time domain and frequency domain is adopted to detect the interference of TT & C link For impulse interference, the energy detection method based on dynamic threshold is used in time domain; For non pulse interference, the method based on polyphase FFT technology and regional energy detection is used in frequency domain The detection algorithm has the characteristics of low consumption of programmable gate array (FPGA) resources, high real-time performance, low detection threshold and high detection probability, and is universal to a variety of interferences. It solves the problem of interference detection in the field of Aerospace TT & C link
Key words: Aerospace TT & C link; Interference detection; Time domain frequency domain parallel
---------------------------------------------------------------------------------------------------
6. [Dissertation] Research on reserve pool computing system based on dual optical feedback VCSEL and its parallel processing ability
Catalogue
cover statement catalogue Chinese abstract English abstract Chapter 1 Introduction 1.1 Research background of this paper 1.2 SL-based RC Research progress at home and abroad 1.3 The main content and structure of this paper Chapter 2 is based on double optical feedback and optical injection disturbance VCSEL Theoretical basis of reserve pool calculation 2.1 Relevant theories of reserve pool calculation 2.1.1 artificial neural network 2.1.2 Traditional reserve pool calculation 2.1.3 Delayed reserve pool calculation 2.1.4 Common benchmark tasks for reserve pool calculation 2.2 Related theories of vertical cavity surface emitting lasers 2.2.1 850 in free running state nm-VCSEL Output characteristics of 2.2.2 850 under simultaneous disturbance of dual optical feedback and optical injection nm-VCSEL Output characteristics of 2.3 Summary of this chapter Chapter 3 is based on the simultaneous disturbance of dual optical feedback and optical injection VCSEL Reserve pool calculation system 3.1 introduction 3.2 system model 3.3 Results and discussion 3.4 conclusion Chapter 4 Summary and Prospect reference thank Papers published and scientific research projects participated in during master's study
element
Discipline: Optics
Degree awarded: Master
Year: 2020
Text language: Chinese language
Link: https://www.zhangqiaokeyan.com/academic-degree-domestic_mphd_thesis/020315558769.html
---------------------------------------------------------------------------------------------------
7. [Dissertation] Research on airborne LiDAR data parallel processing algorithm for disaster emergency
catalogue
Cover
Statement
Chinese abstract
English abstract
Catalogue
1 introduction 1.1 Research background and significance 1.2 Research status at home and abroad 1.2.1 airborne LiDAR Application of scanning technology in disaster emergency mapping 1.2.2 airborne LiDAR Overview of point cloud data processing 1.2.3 Airborne for disaster emergency LiDAR Problems in point cloud data processing 1.3 Research content and paper structure 1.3.1 research contents 1.3.2 Thesis structure 2 Parallel preprocessing of airborne point cloud data 2.1 Traditional serial point cloud preprocessing 2.1.1 airborne LiDAR Coordinate conversion of data 2.1.2 airborne LiDAR Data filtering 2.1.3 airborne LiDAR Block processing of data 2.2 utilize OpenMP Point cloud parallel preprocessing based on 2.3 Comparative analysis of point cloud parallel preprocessing and traditional methods 2.3.1 Overview of experimental area and information processing platform 2.3.2 Efficiency analysis of coordinate transformation 2.3.2 Analysis of filtering effect and efficiency of point cloud 2.3.4 Parallel blocking of point cloud based on tile 3 Parallel based direct interpolation algorithm 3.1 Serial point by point elevation interpolation method 3.2 Parallel based direct interpolation algorithm 3.3 DSM Comparative analysis of accuracy and efficiency of elevation interpolation 4 DSM Tiled parallel processing 4.1 tradition DSM Tile process 4.2 DSM Parallel tile processing based on 4.3 Comparative analysis on efficiency of tile treatment 5 Conclusion and Prospect 5.1 summary 5.2 expectation reference note appended thank
element
Discipline: Cartography and GIS
Degree awarded: Master
Year: 2020
Text language: Chinese language
Link: https://www.zhangqiaokeyan.com/academic-degree-domestic_mphd_thesis/020316000224.html
---------------------------------------------------------------------------------------------------
8. [Dissertation] image oriented distributed parallel processing system
catalogue
Cover
Statement
Chinese abstract
English abstract
Catalogue
The first chapter is introduction 1.1 Research background and significance of this paper 1.2 Research status at home and abroad 1.3 The main contributions and innovations of this paper 1.4 The organizational structure of this paper Chapter II theoretical basis and related technologies 2.1 Image fusion 2.2 GPU programming 2.3 General scheduling algorithm 2.4 LSM-Tree storage structure 2.5 Data distribution strategy 2.5.1 round-robin policy 2.5.2 Consistent hash policy 2.5.3 Consistent hash policy with load cap 2.5.4 Consistent hash policy with virtual nodes 2.6 Redis Memory reclamation strategy 2.7 Summary of this chapter The third chapter is the design of image oriented distributed parallel processing system 3.1 Application scenarios and requirements analysis 3.2 System composition 3.3 System hierarchy 3.4 System physical deployment 3.5 Design of satellite image fusion module under heterogeneous platform 3.5.1 Image fusion on heterogeneous platforms 3.5.2 Heterogeneous platform image fusion processing flow 3.5.3 GPU Optimization of 3.6 Design of adaptive weighted threshold scheduling algorithm 3.6.1 Algorithm model 3.6.2 State switching 3.6.3 Algorithm pseudo code 3.7 Design of high concurrency and low redundancy memory module 3.7.1 Overall architecture 3.7.2 Functions of each node in the enclosure 3.7.3 key technology 3.7.4 Distributed design 3.8 Summary of this chapter The fourth chapter is the implementation of image oriented distributed parallel processing system 4.1.1 BMP Image access 4.1.2 utilize CPU Image fusion 4.1.3 utilize GPU Image fusion 4.2.1 Implementation of loading components 4.2.2 Implementation of data transmission process 4.2.3 Implementation of data processing process 4.2.4 Module execution process 4.3 Implementation of high concurrency and low redundancy storage module 4.3.1 Implementation of communication interaction component 4.3.2 Implementation of network communication framework 4.3.3 Implementation of timer 4.3.4 Implementation of time class 4.3.5 Implementation of work node 4.3.6 Client Concrete implementation of 4.3.7 Module key process 4.4 Summary of this chapter Chapter V system test and result analysis 5.1.1 testing environment 5.1.2 CPU And GPU Contrast test of image fusion effect 5.1.3 CPU And GPU Contrast test of image fusion speed 5.2.1 testing environment 5.2.2 functional testing 5.3.1 testing environment 5.3.2 functional testing 5.3.3 performance testing 5.4 Section of this chapter Chapter VI summary and Prospect 6.1 Full text summary 6.2 Looking ahead thank reference Research results obtained during master's study
element
Subject: computer technology
Degree awarded: Master
Year: 2020
Text language: Chinese language
Link: https://www.zhangqiaokeyan.com/academic-degree-domestic_mphd_thesis/020315968990.html
---------------------------------------------------------------------------------------------------
9. [Dissertation] Research on multi-core parallel processing method of polarimetric radar data
Catalogue
cover statement abstract English abstract catalogue The first chapter is introduction 1．1 Research background and significance 1．2 Research status of polarimetric radar data processing 1．3 Research status of parallel processing of polarimetric radar data 1．4 Innovation and key technology of this paper 1．5 Chapter arrangement of the thesis 2．1 introduction 2．2 Introduction to polarimetric radar data processing 2．3 GPU Introduction to calculation model 2．3．1 GPU Design purpose 2．3．2 GPU storage structure 2．3．3 GPU Programming method 2．3．4 GPU Thread computing The third chapter is based on GPU Research on qualitative parallel processing of polarimetric radar data based on 3．1 introduction 3．2 Refined polarization Lee Parallel research on filtering 3．2．1 Algorithm principle 3．2．2 Algorithm parallel analysis 3．2．3 Optimization results 3．3 H／A／alpha Parallel research on decomposition algorithm 3．2．1 Algorithm principle 3．2．2 Algorithm parallel analysis 3．2．3 Optimization results 3．4 H／A／alpha wishart Parallel research on classification algorithm 3．2．1 Algorithm principle 3．2．2 Algorithm parallel analysis 3．2．3 Optimization results 3．5 Summary of qualitative processing parallel research Chapter IV based on GPU Research on quantitative parallel processing of polarimetric radar data based on 4．1 introduction 4．2 Dubois Parallel research of model 4．2．1 Algorithm principle 4．2．2 Parallel analysis 4．3．1 Algorithm principle 4．3．2 Parallel analysis 4．4 Oh2004 Parallel research of model 4．4．1 Algorithm principle 4．4．2 Parallel analysis 4．5．1 Algorithm principle 4．5．2 Parallel analysis 4．6 Summary of parallel results of quantitative processing Chapter V Conclusion and Prospect 5．1 conclusion 5．2 expectation reference thank Research results and published academic papers Introduction to the author and tutor
element
Discipline: control science and Engineering
Degree awarded: Master
Year: 2020
Text language: Chinese language
Link: https://www.zhangqiaokeyan.com/academic-degree-domestic_mphd_thesis/020315366356.html
---------------------------------------------------------------------------------------------------
10. [Dissertation] Research on parallel processing strategy and application of video big data based on Spark framework
Catalogue
cover statement Chinese abstract English abstract catalogue The first chapter is introduction 1.1 Research background and significance of this paper 1.2 Research status at home and abroad 1.2.1 Research status of video parallel analysis framework 1.2.2 Research status of video parallel analysis algorithms 1.2.3 Research status of video intelligent analysis algorithm 1.3 Research content of the paper 1.4 Paper organization structure Chapter II fundamentals of video big data parallel processing technology 2.1 Spark Parallel processing framework 2.1.1 Memory computing model 2.1.2 Elastic distributed data set 2.1.3 Stream data processing model 2.1.4 Message Oriented Middleware 2.1.5 distributed file system 2.2 Key algorithms and technologies of video analysis 2.2.1 Machine learning algorithm 2.2.2 Deep learning algorithm 2.3.3 Key technologies of video analysis 2.3 Data parallel principle 2.3.1 Data level parallel mechanism 2.3.2 Pipelined parallel mechanism 2.4 Summary of this chapter The third chapter is based on Spark Video big data parallel processing framework based on 3.1 Overall design of parallel framework 3.2 Video data distribution layer 3.3 Video data transmission layer 3.4 Video data processing layer 3.4.1 Image preprocessing 3.4.2 Key frame extraction 3.4.3 Video stream analysis 3.5 Video data application layer 3.6 Summary of this chapter Chapter IV based on Spark Parallel strategy and algorithm of video inter frame independence analysis based on Framework 4.1 Data level parallel mechanism 4.2 Parallel strategy of video frame independent analysis 4.3 Parallel algorithm for elevator passenger number analysis 4.4 Parallel algorithm for elevator debris analysis 4.5 Experimental results and analysis 4.5.1 Experimental analysis of inter frame independent parallel strategy 4.5.2 Experimental analysis of key frame extraction algorithm 4.6 Summary of this chapter Chapter V based on Spark Parallel strategy and algorithm of video inter frame correlation analysis based on Framework 5.1 Pipelined parallel mechanism 5.2 Parallel strategy of video inter frame correlation analysis 5.3 Parallel algorithm for elevator door anomaly analysis 5.4 Experimental results and analysis 5.5 Summary of this chapter Chapter VI development of elevator video big data processing prototype system 6.1 System overview 6.2 development environment 6.3 Overall architecture 6.4 Main functional modules 6.4.1 Elevator passenger number real-time monitoring module 6.4.2 Elevator door abnormal real-time monitoring module 6.4.3 Elevator relic real-time monitoring module 6.4.4 Fault Elevator Distribution visualization module 6.5 Summary of this chapter Chapter VII conclusion and Prospect 7.1 Conclusion 7.2 Prospect reference thank Introduction to the author 1 Resume of the author 2 Academic papers published during the study of master's degree 3 Participating scientific research projects and Awards 4 Invention patent Dissertation dataset
element
Subject: computer technology
Degree awarded: Master
Year: 2020
Text language: Chinese language
Link: https://www.zhangqiaokeyan.com/academic-degree-domestic_mphd_thesis/020316058048.html