Using system dynamics, neural nets, and eigenvalues to analyse supply chain behaviour. A case study

Modelling and simulation considerations for an end-to-end supply chain system

“…Often used to model dynamics in automobile SCs. (Anderson Jr, Fine, & Parker, 2000;Barlas & Gunduz, 2011;Das & Dutta, 2013;Georgiadis & Athanasiou, 2013;Helo, 2000;Higuchi & Troutt, 2004;Hofmann, 2017;Holweg, Disney, Hines, & Naim, 2005;Hussain & Drake, 2011;Martínez-Olvera, 2009;Moreno, Mula, & Campuzano-Bolarin, 2015;Özbayrak, Papadopoulou, & Akgun, 2007;Pierreval, Bruniaux, & Caux, 2007;Poles, 2013;Rabelo, Helal, Lertpattarapong, Moraga, & Sarmiento, 2008;Spengler & Schröter, 2003;Springer & Kim, 2010;Vlachos, Georgiadis, & Iakovou, 2007;Wangphanich, Kara, & Kayis, 2010) Powerful simulation packages such as Vensim, iThink, Powersim or Stella are used to enhance model functionality, capacity and performance. Although, the advancement in these tools capabilities allow for optimization and are geared more towards business managers, there is still lack of E2E-SC system models and guidance on how to develop such models.…”

“…Petri nets (PN)-hybrid, generalized, stochastic, deterministic and stochastic (Arns et al, 2002) Steady State Genetic Algorithm (ssGA) for SCN design (Altiparmak, Gen, Lin, & Karaoglan, 2009) Queuing Network (QN) (Arns et al, 2002) Markov decision process supported by Reinforced Learning (RL) to control inventory policies between multiple actors in SC ) Neural Nets (NN); Eigenvalue Analysis to evaluate SD model outputs (Rabelo et al, 2008) Complex Adaptive System (CAS) (Surana et al, 2005) SCN modelled as CAS and Fitness Landscape Theory to highlight evolutionary complexities of such SC systems (G. It has been noted that the computational complexity is derived from developments as well as a technological shift in modelling where a new era of combining modelling approaches prevails, particularly observed in hybrid and AM/SS models. Hierarchical approach to simulation modelling allows incorporating different levels of detail where each sub-model can be supported by different simulation modelling technique/s.…”

Economic simulation of a poultry supply chain facing a sanitary crisis

“…In addition and because of such complexity, many researchers have used a cybernetic vision based on feedback loops and considered that both the structure of the supply chain and the flow control determine its performance [2], which enables the modeller to capture the mode of dynamic behaviour for the whole system. Given these observations, the Forrester system dynamics theory using continuous simulation has been applied by numerous authors to supply chain modelling (Forrester, 1961;Senge and Sterman, 1992;Cheng, 1996;Sterman, 2000;Higuchi and Troutt, 2004;Kamath and Roy, 2007;Rabelo et al, 2008). Sterman (2006) states that a supply chain network is a complex dynamic system which generates multiple modes of behaviour including business cycles (oscillations in production and inventories), a boost in orders and production rates (see Bullwhip Effect), and a time gap between the beginning of the perturbation and its effect (shifts in the timing of the cycles from consumption to materials).…”

“…According to Sterman (2006), supply chain instability will be increased by the amplification ratio, defined as the ratio of the maximum change in the output to the maximum change in the input. We also refer our measuring method to Rabelo et al (2008) to perform a sensitivity analysis by changing the values of the control parameters to assess whether they can avoid or mitigate the undesirable behaviour of the supply chain. Frozen chicken inventory and external purchases 5.1.…”

Stability analysis of the supply chain by using neural networks and genetic algorithms

“…The integration of these techniques to detect and reduce oscillatory behavior of the SC has been proposed in the literature (Rabelo et al 2006;Moraga et al 2007). However, in most cases sensitivity analysis is used to determine the changes in the model parameters required to stabilize the system.…”