Abstract
The indirect integration method (IIM) is a numerical method for initial value problem of ordinary differential equation, which is based on the residual iteration of differential equation to optimize the parameters of the highest order derivative function layer, and then integrates to the solution function layer step by step through complex trapezoid quadrature formula. Considering the recurrence format of IIM, a single-layer recurrent neural network is first designed to simulate the first-order IIM; On this basis, a deep recurrent neural network is constructed by stacking single-layer recurrent networks to implement any order IIM algorithm; Taking the structure of deep RNN as the calculation diagram, the error back-propagation paths are graded according to the order of traversal differential state and with the help of BPTT algorithm, the recursive relationship between the path parameter gradients (equivalent to the parameter gradients of different order IIM) is obtained.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Data availability
No data, models, or code were generated or used during the study.
References
Parapari H F, Menhaj M B. Solving nonlinear ordinary differential equations using neural networks[C]. International Conference on Control, 2016.
Zhang Z, Yan Z. A Varying Parameter Recurrent Neural Network for Solving Nonrepetitive Motion Problems of Redundant Robot Manipulators[J]. IEEE Transactions on Control Systems Technology, 2018, PP(99): 1–8.
Takeda M, Goodman JW (1986) Neural networks for computation: number representations andprogramming complexity[J]. Appl Opt 25(18):3033–3046
Barnard E, Casasent D. New Optical Neural System Architectures And Applications[J]. Proceedings of SPIE - The International Society for Optical Engineering, 1989, 963.
Lu H, Jin L, Luo X et al (2019) RNN for Solving Perturbed Time-Varying Underdetermined Linear System With Double Bound Limits on Residual Errors and State Variables[J]. IEEE Trans Industr Inf 15(11):5931–5942
Li Z, Li S. An L1-norm based Optimization Method for Sparse Redundancy Resolution of Robotic Manipulators[J]. IEEE Transactions on Circuits and Systems II: Express Briefs, 2021: 1–1.
Cybenko G (1989) Approximation by superposition of sigmoidale function[J]. Mathematics of Control Signal and Systems 2:304–314
Haykin S S, Gwynn R. Neural Networks and Learning Machines[M], 2009.
Hava, T., Siegelmann, et al. Turing computability with neural nets - ScienceDirect[J]. Applied Mathematics Letters, 1991, 4(6): 77–80.
Goodfellow I, Bengio Y, Courville A. Deep Learning[M], 2017: 242–243.
Yiming Wang. Solving ordinary differential equations based on LS-SVM [D]. Tianjin University, 2014.
Qiu Xipeng. Neural network and deep learning [M]. 2019: 137–138.
Parlos A, Atiya A, Chong K, et al. Recurrent multilayer perceptron for nonlinear system identification[M]. 1991: 537–540 vol.2.
Author information
Authors and Affiliations
Corresponding authors
Ethics declarations
Conflict of interest
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Xie, Z., Ai, Y., Chen, J. et al. Deep Recurrent Neural Network Architecture of High Order Indirect Integration Method. Neural Process Lett 54, 1233–1253 (2022). https://doi.org/10.1007/s11063-021-10677-6
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11063-021-10677-6