Robust control of nonlinear systems for external disturbances using second order derivatives of universal learning network

2002 ◽  
Author(s):  
M. Ohbayashi ◽  
K. Hirasawa ◽  
K. Nishimura
Keyword(s):  
Robust Control ◽  
Nonlinear Systems ◽  
Second Order ◽  
External Disturbances ◽  
Learning Network ◽  
Derivatives Of

“Loop-at-a-Time” Design of Dynamic Surface Controllers for Nonlinear Systems

1998 ◽  
Vol 124 (1) ◽  
pp. 104-110 ◽  
Author(s):  
J. Christian Gerdes ◽  
J. Karl Hedrick
Keyword(s):  
Robust Control ◽  
Nonlinear Systems ◽  
Engine Speed ◽  
Design Procedure ◽  
Sequential Design ◽  
Bounded Region ◽  
Dynamic Surface ◽  
Mismatched Uncertainties ◽  
The Common ◽  
Derivatives Of

The use of multiple surface sliding controllers for robust control of nonlinear systems with mismatched uncertainties has produced a number of impressive applications, but also raised a few theoretical questions. Among the latter are the use of numerical differencing to obtain derivatives of desired trajectories, robustness to uncertainties in the gain terms and the common practice of filtering desired trajectories for implementation. This paper seeks to address these issues through the concept of a Dynamic Surface Controller, in which filters form an integral part of the structure. This filtering removes the need for numerical differencing and guarantees a certain smoothness, enabling other assumptions of smoothness to be relaxed. In this paper, the Dynamic Surface Controller is coupled with a sequential design procedure that carves a system workspace out of the state space. Within this bounded region, bounded tracking performance can be rigorously guaranteed in the presence of uncertainties and constraints such as rate limits and saturation can be systematically avoided. The design of a Dynamic Surface Controller and the advantages of the workspace concept are demonstrated in the context of engine speed control.

scholarly journals Robust Control for Universal Learning Network Considering Fuzzy Criterion and Second Order Derivatives

1998 ◽  
Vol 34 (9) ◽  
pp. 1246-1254 ◽  
Author(s):  
Masanao OHBAYASHI ◽  
Kotaro HIRASAWA ◽  
Katsuyuki TOSHIMITSU ◽  
Junichi MURATA ◽  
Jinglu HU
Keyword(s):  
Robust Control ◽  
Second Order ◽  
Learning Network

scholarly journals Probabilistic Learning-Network-Based Robust Control Scheme for Nonlinear Systems

1999 ◽  
Vol 3 (6) ◽  
pp. 485-490
Author(s):  
Jinglu Hu ◽  
◽  
Kotaro Hirasawa ◽  
Junichi Murata ◽  
Chunzhi Jin ◽  
...  
Keyword(s):  
Control System ◽  
Robust Control ◽  
Nonlinear Systems ◽  
Controller Design ◽  
Probability Learning ◽  
Control Design ◽  
Probabilistic Learning ◽  
Criterion Function ◽  
Learning Network ◽  
Control Scheme

We present a control design scheme for nonlinear systems based on a probability learning network (ProNet). ProNet is a learning network equipped with the capability to deal with stochastic signals. A plant and its controllers are described by using a set of related equations and form a unified learning network-ProNet where disturbances are considered as external inputs. In this way, controller design is transferred to ProNet learning. By including an effort to reduce variances of ProNet output in the criterion function for training, the trained ProNet has different sensitivities to signals of different frequencies. A ProNet control system is designed by taking this advantage to increase its robustness against disturbances. Computer simulations confirm the effectiveness of the ProNet control scheme.

Sign in / Sign up

Export Citation Format

Share Document