Add benchmarks

We could run the either through cargo run --release or perhaps using cargo bench: https://doc.rust-lang.org/cargo/commands/cargo-bench.html

It would be useful to generate plots of how solution time scales with size of network for different types of network.