When you need to build a model that closely reflects a real-world facility, you might end up with large-scale?networks. Those are usually found in ports, terminals, warehouses, and factories – facilities with complex internal logistics.
In such cases, the IDE (integrated development environment) performance of the chosen simulation tool becomes crucial. In AnyLogic 8.8.5, we made several notable changes that improved the performance of the IDE.
Here’s an overview for advanced AnyLogic users.
Now, panning, zooming, and?navigating?(through Search or Ctrl + F/Command + F) to a particular element in the models with large networks have become smoother. Any modifications, such as adding new elements to the model or moving the entire network around, are also much faster.
This is also true for the flowchart blocks added to large-scale models. Before, there were two options: either have a fast IDE or code completion for newly added elements. You would need to decide which one was the priority at that moment.
In AnyLogic 8.8.5, you can enjoy both. You can easily add flowchart blocks with code completion accessible right after the element is dropped on the Canvas.
While developing large models, some of our users sometimes faced the “Java 64k limit” error. It was caused by the limitations that exist in the Java language and could be avoided only by dividing an agent that has a huge network into smaller agents.
Our developers made a range of improvements to the AnyLogic internal code generator. As a result, there’s a possibility to go around the limits of Java and keep that single agent whole. So now massive Process Modeling and Material Handling networks, railroads, car roads, and complex layouts with numerous walls in Pedestrian models can be safely created within a single agent with an extremely low chance of hitting the limit.
Note:?Although you can now create huge single agents, keep in mind that for?proper model structure?and easier model maintenance, it might be better to have several smaller agents instead.
When custom library elements are dropped on the Canvas, their 3D objects
appear in the upper left corner (by default) and can be dragged to the animation area
Advanced users who frequently work with large-scale models tend to reuse their parts in new projects,?creating custom libraries. Sometimes agents that may be used as elements of the new library have rich 3D animation that may also contain moving parts.
In the 8.8.6 release, 3D objects representing the animation of agents can be exported during the creation of a custom library. Therefore, when you drop an element from the custom library on the Canvas, the animation associated with that element will automatically be added to the model as well.
Also, you can use these 3D objects as icons for the custom library elements.
Charts and plots that are used for visualization have also faced some changes. Now, when you hover the mouse cursor over a chart during a model run, you’ll see more precise information.
Hovering over charts to see more precise information
For working with multiple agents, plot replication has also been improved. You might find it particularly useful when working with agent populations. For example, you might have a fleet of transporters in your model and want to have a plot showing the current speed of each transporter.
There’s no need to multiply that plot manually. Instead, use the local variable “index” in the Value field of time plots, time stack charts, and time color charts.
Plot settings for replication
Plot replications for each transporter’s current speed
These were major improvements that you might find helpful when working on huge projects with large-scale networks. More details on all the updates and bugfixes released