Research on Surface Defect Detection Method of E-TPU Midsole Based on Machine Vision

In the industrial production of expanded thermoplastic polyurethane (E-TPU) midsoles, the surface defects still rely on manual inspection at present, and the eligibility criteria are uneven. Therefore, this paper proposes an E-TPU midsole surface defect detection method based on machine vision to achieve automatic detection and defect classification. The proposed method is divided into three parts: image preprocessing, block defect detection, and linear defect detection. Image preprocessing uses RGB three channel self-inspection to identify scorch and color pollution. Block defect detection uses superpixel segmentation and background prior mining to determine holes, impurities, and dirt. Linear defect detection uses Gabor filter and Hough transform to detect indentation and convex marks. After image preprocessing, block defect detection and linear defect detection are simultaneously performed by parallel computing. The false positive rate (FPR) of the proposed method in this paper is 8.3%, the false negatives rate (FNR) of the hole is 4.7%, the FNR of indentation is 2.1%, and the running time does not exceed 1.6 s. The test results show that this method can quickly and accurately detect various defects in the E-TPU midsole.

Low Complexity Multiuser Detection with Recursively Successive Zero-Forcing and SIC Based on Nullspace for Multiuser MIMO-OFDM System

To improve the spectrum efficiency, this paper considers the multiuser detection with the MU-MIMO technology for multiuser MIMO-OFDM system uplink with the same subcarrier shared by multiple users. A low complexity multiuser detection algorithm with recursively successive zero-forcing and successive interference cancellation(RSZF-SIC) based on nullspace is proposed. The RSZF process based on the block diagonalization(BD) technique eliminates the co-channel interference(CCI) by a recursive method based on the nullspace orthogonal theorem. The SIC process detects the user signals respectively with the reasonable user detection sequence based on the results of the RSZF process. The computational complexity of the proposed algorithm is effectively reduced by reducing the total number of singular value decomposition(SVD) operations and the dimension of the SVD matrix in the recursive procedure. The performance of the proposed algorithm is improved in terms of bit error rate and sum capacity of the system, especially in the highSNR regime.

Development of 2D computer program to determine geometry of rock mass blocks

Due to various geological processes such as tectonic activities fractures might be created in rock mass body which causes creation of blocks with different shapes and sizes in the rock body. Exact understand- ing of these blocks geometry is an essential issue concerned in different domains of rock engineering such as support system of underground spaces built in jointed rock masses, design of blasting pattern, optimi- zation of fragmentation, determination of cube blocks in quarry mines, blocks stability, etc. The aim of this paper is to develop a computer program to determine geometry of rock mass blocks in two dimen- sional spaces. In this article, the eometrv of iointed rock mass is programmed in MATLABTM.

Development of the entire roller inlayed block intelligence cold strip shape meter and its industrial application

The entire roller inlayed block intelligence cold strip shape meter is developed successfully in this paper,which can stably work chronically in the cold rolling workshop,the independent innovation of the key technology of cold strip shape measuring,the foreign technical monopoly and blockade are broken through.The structural form which is different from the international popular segmented detection roller is put forward,the entire detection roller inlayed block is developed,a group of elastic blocks in that the sense organs are installed are inlayed in the radial symmetry rectangular grooves of the detection roller body,the entire roller structure is composed of elastic pieces and the roller body,the scratch of the strip surface due to different thermal expansion of the traditional detection roller measuring unit is avoided effectively.The signal transmission form which is different form the international popular dry slip ring is put forward,the spray wet slip ring is developed,slip ring and detection roller can be taken down and installed as possibly as fast,so the maintain is convenient,while through the section auto-spray cleaning,lubrication and cooling, the service life is remarkable improved and the outside interference signal is reduced.The signal processing form which is different form the international popular gathering card is put forward,the embedded DSP shape signal processing hardware system is developed,the electromagnet,temperature,humidity and vibration's interference of industry practice to shape transmission signal can be effectively avoided,which can guarantee the instantaneity and measured precision of shape measured signal.The shape signal processing software system whose function contains error compensation,pattern recognition,establishment of the target shape and the closed-loop control calculation is built,the shape meter' computer human interaction is friendly,and function of shape meter is strong,at the same time the intellectualized level of the shape meter is improved.Now,this shape meter has been successfully applied on the 1 250 mm cold strip mill of Angang Steel Company Limited,the long-term operating testing is done,every guide line achieves the anticipative effect,the measuring precision is high and the measuring signal is stable and reliable,the shape meter is tested on high speed,big tension,changed wrap angle and big depressed complicated working conditions,any scratch of the detection roller surface and the strip surface is not appearing,it can completely adapt well to the industrial production environment,reaching the international advanced level.

Advanced optimization of gas turbine aero-engine transient performance using linkage-learning genetic algorithm:PartⅠ,building blocks detection and optimization in runway

This paper proposes a Linkage Learning Genetic Algorithm(LLGA)based on the messy Genetic Algorithm(mGA)to optimize the Min-Max fuel controller performance in Gas Turbine Engine(GTE).For this purpose,a GTE fuel controller Simulink model based on the Min-Max selection strategy is firstly built.Then,the objective function that considers both performance indices(response time and fuel consumption)and penalty items(fluctuation,tracking error,overspeed and acceleration/deceleration)is established to quantify the controller performance.Next,the task to optimize the fuel controller is converted to find the optimization gains combination that could minimize the objective function while satisfying constraints and limitations.In order to reduce the optimization time and to avoid trapping in the local optimums,two kinds of building block detection methods including lower fitness value method and bigger fitness value change method are proposed to determine the most important bits which have more contribution on fitness value of the chromosomes.Then the procedures to apply LLGA in controller gains tuning are specified stepwise and the optimization results in runway condition are depicted subsequently.Finally,the comparison is made between the LLGA and the simple GA in GTE controller optimization to confirm the effectiveness of the proposed approach.The results show that the LLGA method can get better solution than simple GA within the same iterations or optimization time.The extension applications of the LLGA method in other flight conditions and the complete flight mission simulation will be carried out in partⅡ.