Title :
Action-dependent heuristic dynamic programming for level control of three tanks
Author :
Shaojian, Song ; Jinzhi, Li ; Xiaofeng, Lin
Author_Institution :
Coll. of Electr. Eng., Guangxi Univ., Nanning
Abstract :
Action-dependent heuristic dynamic programming is an algorithm of approximate dynamic programming. It does not require an explicit model of system that is to be controlled and an explicitly defined system performance measure dasiaJpsila that is a function in the classical optimal control theory either. The structure and the implementation of this algorithm is analyzed and applied to the level control of three tanks, which is a strong coupling and non-linearity system. This algorithm is compared with single neuron adaptive PID. The results of simulation show that the action-dependent heuristic dynamic programming gets better in steady state error controlling, overshoot controlling, speed of response, antijamming capability and robustness.
Keywords :
adaptive control; control nonlinearities; dynamic programming; heuristic programming; level control; neurocontrollers; robust control; tanks (containers); three-term control; action-dependent heuristic dynamic programming; antijamming capability; nonlinearity system; overshoot control; robust control; single neuron adaptive PID; steady state error control; strong coupling system; tank level control; Algorithm design and analysis; Dynamic programming; Error correction; Heuristic algorithms; Level control; Neurons; Optimal control; Robust control; Steady-state; System performance; Action-dependent heuristic dynamic programming(ADHDP); Learning control; Neural network; adaptive pID;
Conference_Titel :
Control Conference, 2008. CCC 2008. 27th Chinese
Conference_Location :
Kunming
Print_ISBN :
978-7-900719-70-6
Electronic_ISBN :
978-7-900719-70-6
DOI :
10.1109/CHICC.2008.4605056
