Recursive simulation models

Ref-Nr: TDO0155

Technology abstract

Recursive simulation models are a technique of complex systems behavioural modelling that was initially developed to optimize the deployment and renewal of constellations of space satellites. This technology can model discrete state (Markovian or not) and hybrid (random and continuous) systems governed by complex logic.

- Private group -

- -

Read more about this broker

Technology Description

The original technique of recursive simulation modelling appears well adapted to the resolution of certain problems of discrete states systems, possibly hybrid, and provides an alternative tool to other methods such as Petri nets. These traditional methods are much more widely used but can sometimes lead to complex models that are difficult to validate: unlike the Petri nets modelling method, recursive Simulation models describe a single generic transition between the different states of a system. Based on recursive simulation, the program SIMCAB allows complex models to be developed and validated within a short time and at an extremely competitive cost in all fields of engineering (air transport, rail networks, telecommunications, energy...), in order to test the operational capacity of the systems and to optimize their characteristics and conditions of operation and maintenance, from the preliminary phases of design. The program has been used thus at CNES to size satellite constellations (reliability and lifespan of satellites, launcher capacity, renewal strategy, etc.) or complete orbital systems (satellites, RF link, ground segment with its logistic support, etc.) The results obtained in various application cases show that the advantage of the original coupling between optimization and Monte-Carlo simulation is very significant in computing times (giving a time gain of 30 approximately).


Innovations & Advantages

  • Efficiency and simplicity of implementation compared to other techniques (Petri net, etc.)
  • Excel environment
  • Monte Carlo simulation coupled with optimization (Genetic Algorithms & Simplex nonlinear)
  • Automatic generation of simulators: system architecture, maintenance policy and logistic support (with stocks shared parts)
  • Additional features associated with Monte Carlo simulation (adjustment of probability laws, statistical processing, graphics, confidence intervals, etc.)
  • The ability to simulate operation of the systems early in development, to assess the real availability of the expected service and identify potential feasibility problems, at relatively low cost.

Current and Potential Domains of Application

The method can be applied to any engineering field covering random processes and critic development:

  • Energy
  • Transport
  • Industry

A simulator of a transport or energy distribution network can be used to evaluate the availability of the service to the user by considering failures, recovery procedure, etc., and allows the best maintenance or modification actions to be determined.