Update README.md

tutorials/MarlAlgorithms/README.md

@@ -0,0 +1,31 @@
+# Implementation of Multi-agent reinforcement learning algorithms on RouteRL framework
+
+---
+
+In these notebooks, we utilize the Csömör network in our simulator [SUMO](https://eclipse.dev/sumo/), where agents-vehicles will determine their route choices.
+
+<img src="../../docs/img/csomor2.png" alt="Csömör network" width="700" />
+
+---
+
+## Path Generation
+
+Before running the notebooks we need to determine the paths that exist between destination and origin points, which can be adjusted by the user. In these notebooks, we defined 2 origin-destination pairs.
+The following commands will generate the available paths:
+
+```bash
+cd path_generation
+python3 generate_paths.py
+```
+
+Example of path generation:
+
+<img src="../../docs/img/paths.png" alt="Path generation" width="700" />
+
+
+--- 
+
+## Number and type of agents
+
+The number and type of agents in our experiments can be adjusted as needed. Agent information is provided in the `RouteRL/network_and_config/agents_data.csv` file.
+