Online Adaptive Controller Based on Dynamic Evolution Strategies

The majority of non-linear systems nowadays are controlled online using rapid PI-controllers with linear characteristics. Evolutionary algorithms are rarely used, especially for online adaptive control, due to their time complexity. This paper proposes an online adaptive controller based on a dynamic evolution strategy and attempts to overcome this performance problem. The main advantage of the evolution strategies over other gradient machine learning algorithms is that they are insensitive to becoming stuck into local optima. As a result, the proposed controller is capable of responding in real-time (sampling time between 1–5 ms) and was tested on a non-linear, single-degree-of-freedom robotic mechanism. To the extent of our knowledge, this is the first application of evolutionary algorithms in such an online control. In general, the results obtained were better than the results achieved using a traditional PI-controller.

[1]  Matthew Hughes,et al.  Investigating the effects Diversity Mechanisms have on Evolutionary Algorithms in Dynamic Environments , 2016, ArXiv.

[2]  Karel Jezernik,et al.  Neural network control for direct-drive robot mechanisms , 1998 .

[3]  J. G. Ziegler,et al.  Optimum Settings for Automatic Controllers , 1942, Journal of Fluids Engineering.

[4]  Ronald W. Morrison,et al.  Designing Evolutionary Algorithms for Dynamic Environments , 2004, Natural Computing Series.

[5]  Peter Norvig,et al.  Artificial Intelligence: A Modern Approach , 1995 .

[6]  Iztok Fister,et al.  Parameter tuning of PID controller with reactive nature-inspired algorithms , 2016, Robotics Auton. Syst..

[7]  Shengxiang Yang,et al.  Evolutionary dynamic optimization: A survey of the state of the art , 2012, Swarm Evol. Comput..

[8]  Stéphane Doncieux,et al.  Beyond black-box optimization: a review of selective pressures for evolutionary robotics , 2014, Evol. Intell..

[9]  David S. Johnson,et al.  Computers and Intractability: A Guide to the Theory of NP-Completeness , 1978 .

[10]  Thomas Bäck,et al.  Evolutionary algorithms in theory and practice - evolution strategies, evolutionary programming, genetic algorithms , 1996 .

[11]  Hans-Georg Beyer,et al.  Theory of evolution strategies - a tutorial , 2001 .

[12]  Suzana Uran,et al.  Neural-Network Estimation of the Variable Plant for Adaptive Sliding-Mode Controller , 2012 .

[13]  Ying Zhang,et al.  A New Method Evaluating Credit Risk with ES Based LS-SVM-MK , 2018 .

[14]  P. Stadler.  Fitness Landscapes , 1993 .

[15]  Iztok Fister,et al.  Using Novelty Search in Differential Evolution , 2018, PAAMS.

[16]  T. Back,et al.  On the behavior of evolutionary algorithms in dynamic environments , 1998, 1998 IEEE International Conference on Evolutionary Computation Proceedings. IEEE World Congress on Computational Intelligence (Cat. No.98TH8360).

[17]  Anabela Simões,et al.  Prediction in evolutionary algorithms for dynamic environments using markov chains and nonlinear regression , 2009, GECCO.

[18]  Hendrik Richter,et al.  Evolutionary Optimization and Dynamic Fitness Landscapes , 2010, Evolutionary Algorithms and Chaotic Systems.

[19]  Angela Sara Cacciapuoti,et al.  Self-Organizing Strategy Design for Heterogeneous Coexistence in the Sub-6 GHz , 2018, IEEE Transactions on Wireless Communications.

[20]  Miguel G. Villarreal-Cervantes,et al.  An adaptive control study for a DC motor using meta-heuristic algorithms , 2017 .

[21]  Miran Rodic,et al.  Neural network sliding mode robot control , 1997, Robotica.

[22]  Ian F. Akyildiz,et al.  On the Solution of the Steiner Tree NP-Hard Problem via Physarum BioNetwork , 2015, IEEE/ACM Transactions on Networking.

[23]  Wojciech Chmiel,et al.  Quantum-Inspired Evolutionary Approach for the Quadratic Assignment Problem , 2018, Entropy.

[24]  Andries P. Engelbrecht,et al.  Computational Intelligence: An Introduction , 2002 .

[25]  Tobias Glasmachers,et al.  Challenges in High-dimensional Reinforcement Learning with Evolution Strategies , 2018, PPSN.

[26]  K. Williams.  Applications of Genetic Algorithms to a Variety of Problems in Physics and Astronomy , 2005 .

[27]  Carlos Cotta,et al.  Evolutionary Computation in Astronomy and Astrophysics: A Review , 2012, ArXiv.

[28]  Carlos Cruz,et al.  Optimization in dynamic environments: a survey on problems, methods and measures , 2011, Soft Comput..

[29]  Eneko Osaba,et al.  A Hybrid Method for Short-Term Traffic Congestion Forecasting Using Genetic Algorithms and Cross Entropy , 2016, IEEE Transactions on Intelligent Transportation Systems.

[30]  L. Carro-Calvo,et al.  Constrained Single-Point Aerodynamic Shape Optimization of the DPW-W1 Wing Through Evolutionary Programming and Support Vector Machines , 2019 .

[31]  Swagatam Das,et al.  An Adaptive Differential Evolution Algorithm for Global Optimization in Dynamic Environments , 2014, IEEE Transactions on Cybernetics.

[32]  Janez Brest,et al.  Dynamic optimization using Self-Adaptive Differential Evolution , 2009, 2009 IEEE Congress on Evolutionary Computation.

[33]  A. E. Eiben,et al.  Introduction to Evolutionary Computing , 2003, Natural Computing Series.

[34]  Artem Vlasov,et al.  Application of the Evolutionary Algorithm for Creating Advanced Technologies for the Beneficial Use of Gas in Order to Increase the Efficiency and Reliability of the Operation of the Technological Equipment , 2018 .

[35]  Ingo Rechenberg,et al.  Evolutionsstrategie : Optimierung technischer Systeme nach Prinzipien der biologischen Evolution , 1973 .

[36]  A. E. Eiben,et al.  From evolutionary computation to the evolution of things , 2015, Nature.