| Title | Modeling of Aggregation Process Based on Feature Selection Extreme Learning Machine of Atomic Search Algorithm |
| Publication Type | Conference Paper |
| Year of Publication | 2021 |
| Authors | Wu, Chao, Ren, Lihong, Hao, Kuangrong |
| Conference Name | 2021 IEEE 10th Data Driven Control and Learning Systems Conference (DDCLS) |
| Date Published | may |
| Keywords | atom search optimization, Atomic measurements, binary coding, encoding, extreme learning machine, Extreme learning machines, feature extraction, K-NearestNeighbor algorithm, Measurement, Metrics, nearest neighbor search, Prediction algorithms, Predictive models, Production, pubcrawl |
| Abstract | Polymerization process is a process in the production of polyester fiber, and its reaction parameter intrinsic viscosity has an important influence on the properties of the final polyester fiber. In this paper, a feature selection extreme learning machine model based on binary encoding Atom Search Optimization algorithm is proposed and applied to the polymerization process of polyester fiber production. Firstly, the distance measure of K-NearestNeighbor algorithm, combined with binary coding, and Atom Search Optimization algorithm are used to select features of industrial data to obtain the optimal data set. According to the data set, atom search optimization algorithm is used to optimize the weight and threshold of extreme learning machine and the activation function of the improved extreme learning machine. A prediction model with root mean square error as fitness function was established and applied to polyester production process. The simulation results show that the model has good prediction accuracy, which can be used for reference in the follow-up industrial production. |
| DOI | 10.1109/DDCLS52934.2021.9455712 |
| Citation Key | wu_modeling_2021 |
Modeling of Aggregation Process Based on Feature Selection Extreme Learning Machine of Atomic Search Algorithm