Classification of Dynamical Patterns in Autonomously Controlled Logistic Simulations Using Echo State Networks

The concept of autonomous control aims at improving the robustness and performance of logistic systems in a dynamical environment. In this context logistic objects are able to make and execute routing decisions autonomously. On the one hand this enables logistics systems to react promptly on dynamic changes and disturbances. On the other hand autonomous control causes an inherent dynamic systems behavior, which depends mainly on decision logic and initial system states. In order to analyze the interplay between these internal dynamics and the overall systems performance, new approaches for describing and classifying the observed behavior are needed. This paper presents an approach, based on Echo State Networks, for identifying and comparing different dynamics, gained by a simulation model of an autonomous controlled production network. It will be shown, that echo-state networks are able to distinguish different patterns of this exemplary autonomously controlled production network.

[1]  Helmut Hauser,et al.  Echo state networks with filter neurons and a delay&sum readout , 2010, Neural Networks.

[2]  Teodor Gabriel Crainic,et al.  Service network design in freight transportation , 2000, Eur. J. Oper. Res..

[3]  Bernd Scholz-Reiter,et al.  Effects of Autonomous Cooperation on the Robustness of International Supply Networks - Contributions and Limitations for the Management of External Dynamics in Complex Systems , 2007, LDIC.

[4]  Russell Y. Webb,et al.  MULTI-LAYER CORRECTIVE CASCADE ARCHITECTURE FOR ON-LINE PREDICTIVE ECHO STATE NETWORKS , 2008, Appl. Artif. Intell..

[5]  Herbert Jaeger,et al.  A tutorial on training recurrent neural networks , covering BPPT , RTRL , EKF and the " echo state network " approach - Semantic Scholar , 2005 .

[6]  John G. Harris,et al.  Automatic speech recognition using a predictive echo state network classifier , 2007, Neural Networks.

[7]  Katja Windt,et al.  Changing Paradigms in Logistics — Understanding the Shift from Conventional Control to Autonomous Cooperation and Control , 2007 .

[8]  H. Van Dyke Parunak,et al.  "Go to the ant": Engineering principles from natural multi-agent systems , 1997, Ann. Oper. Res..

[9]  Katja Windt,et al.  Autonomy in production logistics: Identification, characterisation and application , 2008 .

[10]  B Scholz-Reiter,et al.  Handling Dynamics in Logistics - Adoption of Dynamic Behaviour and Reduction of Dynamic Effects , 2009 .

[11]  B. Scholz-Reiter,et al.  CARV 2007 Analysing the Dynamics caused by Autonomously Controlled Logistic Objects , 2007 .

[12]  Joachim Hertzberg,et al.  Learning to Ground Fact Symbols in Behavior-Based Robots , 2002, ECAI.

[13]  J. Sauer.  Modeling and solving multi-site scheduling problems , 2001 .

[14]  Herbert Jaeger,et al.  Echo state network , 2007, Scholarpedia.

[15]  Hamideh Afsarmanesh,et al.  Elements of a base VE infrastructure , 2003, Comput. Ind..

[16]  Herbert Jaeger,et al.  Adaptive Nonlinear System Identification with Echo State Networks , 2002, NIPS.

[17]  Dmitry Ivanov,et al.  An adaptive framework for aligning (re)planning decisions on supply chain strategy, design, tactics, and operations , 2010 .

[18]  Bernd Scholz-Reiter,et al.  Modelling and Analysis of Autonomously Controlled Production Networks , 2009 .

[19]  K. Windt,et al.  Understanding Autonomous Cooperation & Control in Logistics: The Impact on Management, Information, Communication and Material Flow , 2007 .

[20]  Herbert Jaeger,et al.  The''echo state''approach to analysing and training recurrent neural networks , 2001 .

[21]  Thomas Jagalski,et al.  Autonomous control of production networks using a pheromone approach , 2006 .

[22]  H. Wiendahl,et al.  Production in Networks , 2002 .

[23]  Bernd Scholz-Reiter,et al.  Autonomously controlled production systems—Influence of autonomous control level on logistic performance , 2009 .