.DESLib.DEVSLib.UsersGuide.Hybrid

Information

Hybrid system modeling using DEVSLib

Modelica is mainly designed for modeling and simulation of continuous-time systems. There are a lot of available Modelica components for developing continuous-time models in many different domains (electrical, mechanical, thermodynamical,...).

Using the DEVSlib implementation of the Parallel DEVS formalism and the Modelica capabilities for modeling continuous-time systems, it is possible to model and simulate hybrid systems. These kind of hybrid systems usually contain two parts, one continuous and another discrete, that must communicate somehow. This communication has been implemented in DEVSLib using the following interface models.

1. Continuous to Discrete

The communication from the continuous part to the discrete part is performed using three models: the Quantifier, the CrossUP and the CrossDOWN.

• Quantifier. This model behaves as a first order quantifier of the continuous-time signal. Every time the value of the signal changes a predefined value (in either positive or negative direction) the model sends an output message with value equal to the new value of the signal.
• CrossUP. This model is used to detect the signal crossing, in the upwards direction, a predefined value (i.e. zero-cross). Every time the value of the signal crosses the defined value, a message is sent with value equal to the signal value.
• CrossDOWN. This model behaves exactly like the CrossUP model, but detects crossings in the downwards direction.

An example of this kind of communication is shown in the next figure.

1. Discrete to Continuous

The communication from the discrete part to the continuous part is performed using the DICO model. This model translates the value of the received messages from the discrete part into a continuous (Real) signal. The generated continuous signal is a constant piecewise signal, since the changes in the value of the signal are generated by the arrived messages.

An example of this kind of communication is shown in the next figure.