Concept

The section outlines the underlying principles of smif.

Running a System-of-Systems Model

Once smif has been used to configure a system-of-systems model, all that is needed to run the model is the command smif run.

smif handles the loading of input data, spinning up the simulation models, extracting a graph of dependencies from the network of inputs and outputs, running the models in the order defined by this graph and finally persisting state and results from the simulation models to a data store.

Operational Simulation and Capacity Expansion

Fundamental to the design of smif is the distinction between the simulation of the operation of a particular system, and the long-term expansion of the capacity which underpin this operation.

The former is the domain of the simulation models, while the latter is handled by smif. smif provides the architecture to handle the capacity expansion problem using one of three approaches: a fully specified approach, a rule based approach and an optimisation approach.

In each of these three approaches, decisions regarding the increase or decrease in the capacity of an asset are propagated into the model inputs via a state transition function.

State

State refers to the information which must be persisted over time. Normally, this will refer to the capacity of an asset (e.g. number of wind turbines), the level of storage (e.g. the volume of water stored in a reservoir). Other information, including metrics, such as CO₂ emissions, or cumulative costs, may also be relevant.

smif handles State for the management of the capacity expansion. The process of passing state from one time-period to another is managed by smif. In this respect, note the distinction between time-steps for the capacity expansion problem, which will normally be measured in years or decades, versus the time-steps for each instance of a simulation model, which will run within a year or decade.

Wrapping Simulation Models

At the core of smif are the target simulation models which we wish to integrate into a system-of-systems model. A simple example simulation model is included in tests.fixtures.water_supply.ExampleWaterSupplySimulation. A simulation model has inputs, and produces outputs, which are a function of the inputs. The smif.abstract.SectorModel is used to wrap an individual simulation model, and provides a uniform API to other parts of smif.

An input can correspond to:

• model parameters, whose source is either from a scenario, or from the outputs from another model (a dependency)
• model state (not yet implemented)