Research on Electric Vehicle Road Identification Method Based on RBF Neural Network

2014 ◽  
Vol 543-547 ◽  
pp. 1413-1416
Author(s):  
Zhi Yu Huang ◽  
Jia Li
Keyword(s):  
Neural Network ◽  
Control System ◽  
Simulation Model ◽  
Rbf Neural Network ◽  
Relevant Information ◽  
Permanent Magnet Synchronous Motors ◽  
Synchronous Motors ◽  
The Road ◽  
Road Condition ◽  
Motor Control System

Accurately identifying road condition can send relevant information to the motor control system, so that control system of the motor can adjust the control strategy timely, eventually, the intelligent and optimal control of electric vehicles is realized. In this paper, according to these mathematical model, the permanent magnet synchronous motors simulation model and vehicles simulation model are proposed. Then, output torque of motor and speed of motor are served as the input of RBF neural network, which helps road condition to be identified. The simulation result shows that the road condition is well identified by proposed method based on RBF neural network.

Barrier Lyapunov function–based adaptive neural network control for incommensurate fractional-order chaotic permanent magnet synchronous motors with full-state constraints via command filtering

2020 ◽  
pp. 107754632096263
Author(s):  
Senkui Lu ◽  
Xingcheng Wang
Keyword(s):  
Neural Network ◽  
Fractional Order ◽  
State Constraints ◽  
Network Control ◽  
Neural Network Control ◽  
Permanent Magnet Synchronous Motors ◽  
Synchronous Motors ◽  
Adaptive Neural Network Control ◽  
Full State ◽  
Full State Constraints

This article considers the problem of adaptive neural network control via command filtering for incommensurate fractional-order chaotic permanent magnet synchronous motors with full-state constraints and parameter uncertainties. First, a neural network state observer based on a K-filter is established to reconstruct unmeasured feedback information. Then, the command filtered technology is used to overcome the inherent “explosion of complexity” problem under fractional-order framework. Furthermore, to eliminate the errors generated by filters, an error compensation system is used. Meanwhile, the nonlinear unknown functions are approximated by using neural networks. In addition, the barrier Lyapunov functions are designed to avoid the violation of the state constraints. Finally, the availability of the proposed control algorithm is revealed by numerical simulations.

Sign in / Sign up

Export Citation Format

Share Document