Robust adaptive critic based neural networks for speed-constrained agile missile control

Dong-Hoon, Han; Balakrishnan, S. N.

doi:10.2514/6.1999-4064

Cited by 11 publications

(8 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…where R ¼ relative range between the missile and PIP (m) u ¼ elevation angle of the range vector measured from the local horizontal (rad) d ¼ heading error (rad) Using the range R rather than time as the independent variable, the system of equations (12) to (14) can be reformulated as…”

Section: Midcourse Guidance In a Vertical Planementioning

confidence: 99%

“…In this section, general development on the optimal control of the non-linear systems is presented in an ADP framework. Detailed derivations of these conditions may also be found in Balakrishnan and Biega [8] and Han and Balakrishnan [10][11][12], which are repeated here for clarity and completeness. The development in this section will subsequently be used in synthesizing the neural networks for midcourse guidance.…”

Section: Approximate Dynamic Programmingmentioning

confidence: 99%

See 1 more Smart Citation

Midcourse guidance law with neural networks

Dong-Hoon

Balakrishnan

Ohlmeyer

2005

Proceedings of the Institution of Mechanical Engineers, Part G:

View full text Add to dashboard Cite

A dual neural network 'adaptive critic approach' is used in this study to generate midcourse guidance commands for a missile to reach a predicted impact point while maximizing its final velocity. The adaptive critic approach is based on approximate dynamic programming. The first network, called a 'critic', network, outputs the Lagrangian multipliers arising in an optimal control formulation while the second network, called an 'action' network, outputs the optimal guidance/control. While a typical adaptive critic structure consists of a single critic and a single controller, the midcourse guidance problem needs indexing in terms of the independent variable and therefore there is a cascade of critics and controllers each set for a different index. Every controller learns from the critic at the previous stage. Though the networks are trained off-line, the resulting control is in a feedback form. A midcourse guidance problem is the first testbed for this approach where the input is vector-valued. The numerical results for a number of scenarios show that the network performance is excellent. Corroboration for optimality is provided by comparisons of the numerical solutions using a shooting method for a number of scenarios. Numerical results demonstrate some attractive features of the adaptive critic approach and show that this formulation works very well in guiding the missile to its final conditions from an envelope of initial conditions. This application also demonstrates the use of adaptive critics as a tool to solve a class of 'free final time' problems in optimal control, which are usually very difficult.

show abstract

Section: Midcourse Guidance In a Vertical Planementioning

confidence: 99%

Section: Approximate Dynamic Programmingmentioning

confidence: 99%

Midcourse guidance law with neural networks

Dong-Hoon

Balakrishnan

Ohlmeyer

2005

Proceedings of the Institution of Mechanical Engineers, Part G:

View full text Add to dashboard Cite

show abstract

“…Until now several RL Flight Controllers have been proposed with different ACD frameworks. [12][13][14][15][16][17][18][19][20][21] One limitation of these controllers is that they have an exorbitant computational requirement. This requirement comes from learning two different functions with separate function approximation structures.…”

Section: Introductionmentioning

confidence: 99%

Adaptive Critic Control For Aircraft Lateral-Directional Dynamics

Ashraf

Kampen

2020

AIAA Scitech 2020 Forum

View full text Add to dashboard Cite

Loss of control-in flight (LOC-I) is one of the causes of catastrophic aircraft accidents. Fault-tolerant flight control (FTFC) systems can prevent LOC-I and recover aircraft from LOC-I precursors. One group of promising methods for developing Fault-Tolerant Control (FTC) system is the Adaptive Critic Designs (ACD). Recently one ACD algorithm, called value function based single network adaptive critic (J-SNAC), has emerged and it promises to make applications of ACD more practical by reducing the required amount of computations. This paper discusses the implementation of this framework for the design of a lateral-directional flight controller. The proposed flight controller is trained to perform coordinated-turns with an F16 simulation model. The trained controller was evaluated for tracking two different heading command signals, robustness against sensor noises and partial failure of the ailerons. The controller is found to be effective for the considered assessments.

show abstract

“…In our previous study we have used this methodology to solve linear and even nonlinear problems [1] [2][13] [14], some other people have also contributed to this research area [ll]. Although these papers have shown impressive results, so far there is no analysis on the mechanics of the method.…”

Section: Introductionmentioning

confidence: 99%

Convergence analysis of adaptive critic based optimal control

Liu

Balakrishnan

2000

Proceedings of the 2000 American Control Conference. ACC (IEEE Cat. No.00CH36334)

View full text Add to dashboard Cite

Adaptive critic based neural networks have been found to be powerful tools in solving various optimal control problems. The adaptive critic approach consists of two neural networks which output the control values and the Lagrangian multipliers associated with optimal control. These networks are trained successively and when the outputs of the two networks are mutually consistent and satisfy the differential constraints, the controller network output produces optimal control. In this paper, we analyze the mechanics of convergence of the network solutions. We establish the necessary conditions for the network solutions to converge and show that the converged solution is optimal.

show abstract

Robust adaptive critic based neural networks for speed-constrained agile missile control

Cited by 11 publications

References 5 publications

Midcourse guidance law with neural networks

Midcourse guidance law with neural networks

Adaptive Critic Control For Aircraft Lateral-Directional Dynamics

Convergence analysis of adaptive critic based optimal control

Contact Info

Product

Resources

About