One advantage of SysML for general-purpose modeling is that it is an object-oriented modeling language. We take advantage of it here to build RCS control point models, composed of Switch, LightSignal and JBox (communications junction box) building blocks. We can think of our RCS as a network of control points connecting blocks of bi-directional track. Each is controllable from a central location. A simple control point regulates train access to the next block of track with one or more light signals. More complex control points contain one or more switches, which can direct trains onto different tracks.
This post is the second in a series of blog posts on applying Model-Based Systems Engineering (MBSE) to Railway Control Systems (RCS). We describe some preliminary results for such an approach, showing sample diagrams from several different SysML modeling tools (MagicDraw from No Magic Inc, IBM Rational Rhapsody, and Enterprise Architect from Sparx Systems). These models will be made available for download with the last post of the series, which will also show how the Intercax MBSE platform Syndeia can connect these models with Simulink (The MathWorks, Inc.) for simulation.
Control Points of a Railway Control System (RCS)
From the fundamental building blocks, we can build up more complex elements. Figure 1 shows three types of control points, each connecting Blocks of track and characterized by a geographical location in latitude and longitude. A ControlPoint_2Way consists of two opposite-facing LightSignals (remember that Blocks are bi-directional; the train could be approaching from either direction) and a communications junction box (JBox2W) receiving signal control inputs from a remote source. In this example, the two LightSignals and the JBox2W are specified in SysML as Full Ports.
Figure 1 Control points (MagicDraw)
A ControlPoint_3Way consists of three LightSignals for three track inputs, a junction box and a Switch (indicated by a small image at the center of the block). The Switch can direct a train entering on the left, port 1, onto either of the two tracks on the right, port 2a or port 2b, or vice versa. (SysML note: In this assembly, signals and junction box remain as full ports, but the switch is treated as a part property).
ControlPort_4Way uses the same fundamental elements (four signals, two switches) to describe two parallel tracks with a crossover point and more complex subsystems could be developed in the same way.
For each element and subsystem, SysML also allows us to model the internal connectivity. Two examples are shown in Figure 2 and Figure 3. In Figure 2, an internal block diagram (IBD) for a ControlPoint_2Way, two light signals facing in opposite directions are controlled from an electrical junction box. The control lines have been shown with lighter weight than the track pass-through, a connector with a heavy dark line.
Figure 2 Internal block diagram, 2 Way Control Point (Rhapsody)
Figure 3 Internal block diagram, 3 Way Control Point (Enterprise Architect)
The ControlPoint_3Way IBD in Figure 3 is more complex, with connectors from the switch to the three signals guarding the inputs representing the tracks and lighter connectors showing internal comm links from the junction box to the signals and switch. Similar IBDs are created for other control point variants.
From these elements, we will build simple rail systems in Part 3. The final two parts planned for the blog series will consider behavior, as captured in SysML sequence diagrams, and generation of Simulink models for analyses.