Evolutionary Computation and Nonlinear Programming in Multi-model Robust Control Design

An algorithmic parameter tuning methodology for controller design of complex systems is needed. This methodology should offer designers a great degree of flexibility and give insight into the potentials of the controller structure and the consequences of the design decisions that are made. Such a method is proposed here. For an exploratory phase a new pare to-ranked genetic algorithm is proposed to generate an evenly dispersed set of near optimal, global, solutions. By pair-wise preference statements on design alternatives a linear program is set up as a formal means for selecting the solution with best overall designer satisfaction. In a following interactive design phase using nonlinear programming techniques with a priori decisions on allowed quality levels, a best tuning compromise in competing requirements satisfaction is searched for while guaranteeing pare to-optimality. In particular, this two-phase tuning approach allows the designer to balance nominal control performance and multi-model control robustness.

[1]  Karl Heinz Kienitz.  Controller design using fuzzy logic - A case study , 1993, Autom..

[2]  William P. Birmingham,et al.  Preference-directed design , 1995, Artificial Intelligence for Engineering Design, Analysis and Manufacturing.

[3]  Reinhold Steinhauser,et al.  Application of vector performance optimization to a robust control loop design for a fighter aircraft , 1983 .

[4]  Peter J. Fleming,et al.  Multiobjective robust control using evolutionary algorithms , 1996, Proceedings of the IEEE International Conference on Industrial Technology (ICIT'96).

[5]  G. Grubel,et al.  Multi-objective design of controllers with fuzzy logic using the control engineering computation environment ANDECS , 1994, Proceedings of IEEE Symposium on Computer-Aided Control Systems Design (CACSD).

[6]  Hartmut Pohlheim,et al.  Genetic and evolutionary algorithm toolbox for use with matlab , 1994 .

[7]  Kaisa Miettinen,et al.  Nonlinear multiobjective optimization , 1998, International series in operations research and management science.

[8]  Kay Chen Tan,et al.  Control system design automation with robust tracking thumbprint performance using a multiobjective evolutionary algorithm , 1999, Proceedings of the 1999 IEEE International Symposium on Computer Aided Control System Design (Cat. No.99TH8404).

[9]  V. Zakian,et al.  Design of dynamical and control systems by the method of inequalities , 1973 .

[10]  G. Grübel,et al.  ANDECS: a computation environment for control applications of optimization , 1994 .

[11]  Yun Li,et al.  Performance indices in evolutionary CACSD automation with application to batch PID generation , 1999, Proceedings of the 1999 IEEE International Symposium on Computer Aided Control System Design (Cat. No.99TH8404).

[12]  Andras Varga,et al.  A graphical user interface for flight control development , 1999, Proceedings of the 1999 IEEE International Symposium on Computer Aided Control System Design (Cat. No.99TH8404).

[13]  G. Grubel.  Perspectives of CACSD: embedding the control system design process into a virtual engineering environment , 1999, Proceedings of the 1999 IEEE International Symposium on Computer Aided Control System Design (Cat. No.99TH8404).

[14]  Georg Grübel.  Another view on the Design Challenge achievements , 1997 .

[15]  Peter J. Fleming,et al.  An Overview of Evolutionary Algorithms in Multiobjective Optimization , 1995, Evolutionary Computation.

[16]  Samir Bennani,et al.  Robust flight control : a design challenge , 1997 .