A robust cutting pattern recognition method for shearer based on Least Square Support Vector Machine equipped with Chaos Modified Particle Swarm Optimization and Online Correcting Strategy

Cheng

IEEE Trans. Evol. Computat.

et al. 2021

Section: A Discussion On the Used Optimizers In Doasmentioning

confidence: 99%

A Survey of Evolutionary Continuous Dynamic Optimization Over Two Decades—Part B

Cheng

IEEE Trans. Evol. Computat.

et al. 2021

“…SVM has fast training speed and high classification accuracy for small-scale datasets, making it widely used in fault diagnosis. However, SVM has been successfully applied in fields such as text classification [7] and pattern recognition [8][9][10]. However, in practical applications, it is susceptible to the influence of parameter selection [11,12], which includes kernel function parameters and penalty factor C. The quality of parameter selection can seriously affect the classification performance and generalization ability of SVM models.…”

Section: Introductionmentioning

confidence: 99%

SVM bearing fault diagnosis based on fast inter-class distance in the feature space and DMPSO algorithm

Song,

Yu,

Yang

et al. 2024

Meas. Sci. Technol.

Support Vector Machines (SVM) have good processing performance for small sample datasets. The giant search space for kernel parameters and the tendency of parameter optimization to fall into local optima are two essential factors that affect the generalization ability of SVM models and, thus, affect the accuracy of fault diagnosis results. Propose using Fast Inter-Class Distance (FICDF) in the feature space to reduce the search space for kernel function parameters and then use Differential Mutation Particle Swarm Optimization (DMPSO) to optimize kernel function parameters to improve the generalization ability and classification accuracy of the SVM model. Firstly, the FICDF algorithm is used to calculate the Euclidean distance between classes, and a fast segmentation idea is proposed for fast operations to obtain a smaller kernel parameter search space. Then, the global search ability of the differential mutation Particle swarm optimization algorithm is used to obtain the optimal parameter combination of the support vector machine model. Finally, the fault diagnosis model of the support vector machine is applied to the fault diagnosis of rolling bearings. The experimental results show that compared with other fault diagnosis methods, this model method has higher classification accuracy and verifies its better classification speed.

“…Dewangan et al [14,15] used scanning electron microscopes and energy dispersive X-ray detectors to deeply study the wear parts of picks, and classified the wear mechanisms of picks. Liu et al [16] proposed a new method for coal mining machine cutting pattern recognition based on least squares support vector machine (LSSVM) and chaos modified particle swarm optimization (CMPSO) optimal online correction strategy. Lei et al [17] introduced a new state diagnosis algorithm based on SVM and multi-scale fuzzy entropy.…”

Section: Introductionmentioning

confidence: 99%

An Unsupervised Intelligent Method for Cutting Pick State Recognition of Coal Mining Shearer

Zhang

Meng

Qi³

et al. 2020

IEEE Access

This paper proposes an intelligent recognition method for shearer cutting state based on deep learning theory, to solve the problems where the picks are prone to various failure forms during the cutting of coal and rock masses by the shearer. The failure will lead to the decline on the stability of the entire machine of the shearer and affect the safety production. Specially, a 1:1 simulation bench is used for simulating underground mining conditions to measure and collect the cutting loads of picks and establish a sample database. Deep learning-based intelligent recognition method is an effective tool that can break away from the dependency of prior knowledge and recognition experience, and sparse. In this paper, a promising deep learning method called sparse filtering is proposed for intelligent recognition of shearer cutting. So sparse filtering is applied to construct an automatic feature extraction model, and softmax regression is adopted as a classifier for cutting pick state recognition. Furthermore, L1/2 regularization term is added to the cost function of sparse filtering to prevent the problem of excessive model training and weights. The proposed method for identifying the cutting status of the shearer can effectively monitor the cutting status of the picker, thereby improving the safety and stability of the cutting of the shearer and promote the coal mining efficiency. INDEX TERMS Shearer pick; sparse filtering; softmax regression; state recognition. K. Zhang et al.: An Unsupervised Intelligent Method for cutting pick state recognition