Self-organised criticality models
olde Scheper, Tjeerd (2021), Self-organised criticality models, Dryad, Dataset, https://doi.org/10.5061/dryad.hhmgqnkdc
Complex biological systems are considered to be controlled using feedback mechanisms. Reduced systems modelling has been effective to describe these mechanisms, but this approach does not sufficiently encompass the required complexity that is needed to understand how localised control in a biological system can provide global stable states. Self-Organised Criticality (SOC) is a characteristic property of locally interacting physical systems, which readily emerges from changes to its dynamic state due to small nonlinear perturbations. These small changes in the local states, or in local interactions, can greatly affect the total system state of critical systems. It has long been conjectured that SOC is cardinal to biological systems, that show similar critical dynamics, and also may exhibit near power-law relations. Rate Control of Chaos (RCC) provides a suitable robust mechanism to generate SOC systems, which operates at the edge of chaos. The bio-inspired RCC method requires only local instantaneous knowledge of some of the variables of the system, and is capable of adapting to local perturbations. Importantly, connected RCC controlled oscillators can maintain global multi-stable states, and domains where power-law relations may emerge. The network of oscillators deterministically stabilises into different orbits for different perturbations, and the relation between the perturbation and amplitude can show exponential and power-law correlations. This can be considered to be representative of a basic mechanism of protein production and control, that underlies complex processes such as homeostasis. Providing feedback from the global state, the total system dynamic behaviour can be boosted or reduced. Controlled SOC can provide much greater understanding of biological control mechanisms, that are based on distributed local producers, with remote consumers of biological resources, and globally defined control.
These euqation files match the EuNeurone numerical simulation software. It can use different fixed step integration algorithms for the numerical integration of an entire set of equations. The Eq files contain the equation code, which is byte compiled for real time computation. External data can be presented to the models using a socket implementation. A client is provided that demonstrate this mechanism. This can be used to feed numerical models with real data.
Code is available at Github, but needs to be custom build by the user due to the dependencies.
Models are descibred as sets of differential equations with initial values, and parameters. These are parsed and compiled by the EuNeurone software for integration.