Acuasal modelling

Is acuasal modelling possible with MATLAB? I know simscape language is there to work with acuasal modelling. But I just want to know that acuasal modelling is possible with MATLAB itself or not! I have tried several models with modelica. It worked fine. But about doing the same with MATLAB (without using simscape language or simulink)?

[Walter Roberson]

Are you sure you do not want acasual modelling instead?

[Jan]

Or acausal modelling?

[Bill]

You can surely do what you like in MATLAB, as m-script is a Turing-complete programming language. Perhaps suboptimal to spell it however you like! :-)

Simscape is the MathWorks offering for acausal, where a component stipulates differential algebraic equations as invariants, eg resistor requires that V_in - V_out == I * R. Then the solver conceptually rearranges the equations before solving them. So you could vary V from a source, and I would vary in the model, or vary I from a source, and the voltage drop would change.

This is all very clever but you have to buy the extra licence.

If you want a taster of acausal you can get away with basic Simulink for simple quasi static situations, using GOTO blocks.

Simulink can suffice because a physical system generally has a flow variable (eg current) and a state variable (eg voltage) and you can have the flow go along the Simulink arrows (instead of causality), and the voltage travel against the flow via global gotos, the driving block controlling the flow variable based on the state feedback, and flow splitter blocks balancing against voltage feedback similarly. Now causality can travel backwards along the arrows because if you change the state variable the flow will react.

Heaven help you if the artificial dynamics of this circuit balancing interferes with the real dynamics of the physical system. Probably best to use the right tool for the job, especially as things get more complicated.