Parallel Trajectory Training of Recurrent Neural Network Controllers With Levenberg-Marquardt and Forward Accumulation Through Time in Closed-Loop Control Systems.
Xingang FuJordan SturtzEduardo AlonsoRajab ChallooLetu QinggePublished in: IEEE Trans. Sustain. Comput. (2024)
Keyphrases
- closed loop
- control system
- recurrent neural networks
- levenberg marquardt
- training algorithm
- artificial neural networks
- control law
- back propagation
- recurrent networks
- neural network
- feedforward neural networks
- feed forward
- hidden layer
- trajectory tracking
- open loop
- fuzzy logic
- control scheme
- neural network structure
- least squares
- reference trajectory
- feedback control
- pid controller
- control loop
- bundle adjustment
- neural model
- control algorithm
- control method
- controller design
- training process
- nonlinear systems
- real time
- multi layer perceptron
- fuzzy control
- multilayer perceptron
- neural nets
- control strategy
- training set
- adaptive control
- support vector machine
- learning algorithm
- convergence rate
- structure from motion
- fuzzy sets
- learning rate
- fuzzy controller
- rbf neural network
- expert systems
- neural network controller