Intelligent control system for industrial robots based on multi-source data fusion

Yang Zhang

DOI: https://doi.org/10.1515/jisys-2022-0286

2023-01-01

Journal of Intelligent Systems

Abstract:Abstract Industrialization has advanced quickly, bringing intelligent production and manufacturing into people’s daily lives, but it has also created a number of issues with the ability of intelligent control systems for industrial robots. As a result, a study has been conducted on the use of multi-source data fusion methods in the mechanical industry. First, the research analyzes and discusses the existing research at home and abroad. Then, a robot intelligent control system based on multi-source fusion method is proposed, which combines multi-source data fusion with principal component analysis to better fuse data of multiple control periods; In the process, the experimental results are dynamically evaluated, and the performance of the proposed method is compared with other fusion methods. The results of the study showed that the confidence values and recognition correctness of the intelligent control system under the proposed method were superior compared to the Yu, Murphy, and Deng methods. Applying the method to the comparison of real-time and historical data values, it is found that the predicted data under the proposed method fits better with the actual data values, and the fit can be as high as 0.9945. The dynamic evaluation analysis of single and multi-factor in the simulation stage demonstrates that the control ability in the training samples of 0–100 is often better than the actual results, and the best evaluation results may be obtained at the sample size of 50 per batch. The aforementioned findings demonstrated that the multi-data fusion method that was suggested had a high degree of viability and accuracy for the intelligent control system of industrial robots and could offer a fresh line of enquiry for the advancement and development of the mechanical industrialization field.