Zeroing Neural Network for Bound-Constrained Time-Varying Nonlinear Equation Solving and Its Application to Mobile Robot Manipulators.

Ma Zhisheng,Yu Shihang,Han Yang,Guo Dongsheng
DOI: https://doi.org/10.1007/s00521-021-06068-6
2021-01-01
Neural Computing and Applications
Abstract:A typical class of recurrent neural networks called zeroing neural network (ZNN) has been considered as a powerful alternative for time-varying problems solving. In this paper, a new ZNN model is proposed and studied to solve the bound-constrained time-varying nonlinear equation (BCTVNE). Specifically, by introducing a time-varying nonnegative vector, the BCTVNE is reformulated as a combined system of nonlinear equations. On the basis of two indefinite error functions and the exponential decay formula, the new ZNN model is thus developed, which can zero in on the combined system. Theoretical analysis and simulation results are provided to verify the effectiveness of the proposed ZNN model. The applicability is further indicated under the simulations on an omnidirectional mobile robot manipulator via the proposed ZNN model.
What problem does this paper attempt to address?