A/C Performance abstract class + graph improvement (#453)

* updated graph + PS + OPENAP * cleanup
airbus · Dec 20, 2024 · 02f54a1 · 02f54a1
1 parent a581668
commit 02f54a1
Show file tree

Hide file tree

Showing 62 changed files with 3,677 additions and 2,532 deletions.
diff --git a/notebooks/15_flightplanning_tuto.ipynb b/notebooks/15_flightplanning_tuto.ipynb
@@ -73,7 +73,20 @@
    "source": [
     "import datetime\n",
     "\n",
+    "from skdecide.hub.domain.flight_planning.aircraft_performance.bean.aircraft_state import (\n",
+    "    AircraftState,\n",
+    ")\n",
+    "from skdecide.hub.domain.flight_planning.aircraft_performance.performance.performance_model_enum import (\n",
+    "    PerformanceModelEnum,\n",
+    ")\n",
+    "from skdecide.hub.domain.flight_planning.aircraft_performance.performance.phase_enum import (\n",
+    "    PhaseEnum,\n",
+    ")\n",
+    "from skdecide.hub.domain.flight_planning.aircraft_performance.performance.rating_enum import (\n",
+    "    RatingEnum,\n",
+    ")\n",
     "from skdecide.hub.domain.flight_planning.domain import FlightPlanningDomain, WeatherDate\n",
+    "from skdecide.hub.domain.flight_planning.flightplanning_utils import plot_network\n",
     "from skdecide.hub.solver.astar import Astar\n",
     "\n",
     "# reload\n",
@@ -149,6 +162,35 @@
     "cost_function = \"fuel\""
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Definition of the aircraft initial state\n",
+    "Now, we define the aircraft state. This implies picking an aircraft performance model : `OPENAP` or `POLL_SCHUMANN`"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "acState = AircraftState(\n",
+    "    model_type=\"A320\",  # only for OPENAP and POLL_SCHUMANN\n",
+    "    performance_model_type=PerformanceModelEnum.POLL_SCHUMANN,  # PerformanceModelEnum.OPENAP\n",
+    "    gw_kg=80_000,\n",
+    "    zp_ft=10_000,\n",
+    "    mach_cruise=0.7,\n",
+    "    cas_climb_kts=250,\n",
+    "    cas_descent_kts=220,\n",
+    "    phase=PhaseEnum.CLIMB,\n",
+    "    rating_level=RatingEnum.MCL,\n",
+    "    cg=0.3,\n",
+    "    gamma_air_deg=0,\n",
+    ")"
+   ]
+  },
   {
    "attachments": {},
    "cell_type": "markdown",
@@ -166,29 +208,28 @@
    "outputs": [],
    "source": [
     "domain_factory = lambda: FlightPlanningDomain(\n",
-    "    origin,\n",
-    "    destination,\n",
-    "    aircraft,\n",
-    "    weather_date=weather_date,\n",
-    "    heuristic_name=heuristic,\n",
-    "    perf_model_name=\"openap\",  # a/c performance model\n",
+    "    origin=origin,\n",
+    "    destination=destination,\n",
+    "    aircraft_state=acState,\n",
     "    objective=cost_function,\n",
-    "    fuel_loop=False,\n",
-    "    graph_width=\"normal\",\n",
+    "    heuristic_name=heuristic,\n",
+    "    weather_date=weather_date,\n",
+    "    cruise_height_min=35_000,\n",
+    "    cruise_height_max=42_000,\n",
+    "    graph_width=\"medium\",  # small, medium, large, xlarge\n",
+    "    nb_vertical_points=5,  # flight levels in cruise\n",
+    "    nb_forward_points=5,\n",
+    "    nb_lateral_points=7,  # must be odd\n",
     ")\n",
     "\n",
     "domain = domain_factory()"
    ]
   },
   {
-   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "## Solving and rendering out the flight planning domain\n",
-    "\n",
-    "We use here an A* solver as mentionned before. \n",
-    "We also use the custom rollout proposed to have some visualization during the flight planning generation."
+    "We can plot the graph:"
    ]
   },
   {
@@ -197,22 +238,18 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "with Astar(\n",
-    "    heuristic=lambda d, s: d.heuristic(s), domain_factory=domain_factory, parallel=False\n",
-    ") as solver:\n",
-    "    solver.solve()\n",
-    "    domain.custom_rollout(solver=solver)"
+    "plot_network(domain)"
    ]
   },
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "## Changing aircraft performance model\n",
-    "\n",
-    "One of the new features of the flight planning domain is the addition of another aircraft performance model. Indeed, the default is OpenAP (`openap`) but we can now choose [Poll-Schumann](https://elib.dlr.de/135592/1/Poll_Schumann_estimation_method_fuel_burn_performance_aircraft_cruise_part_1_fundamentals_2020.pdf) (`PS`), which is more accurate.\n",
+    "## Solving and rendering out the flight planning domain\n",
     "\n",
-    "Thus, we can run the solver on the domain with a Poll-Schumann A/C performance model:"
+    "We use here an A* solver as mentionned before. \n",
+    "We also use the custom rollout proposed to have some visualization during the flight planning generation."
    ]
   },
   {
@@ -221,34 +258,20 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "domain_factory = lambda: FlightPlanningDomain(\n",
-    "    origin,\n",
-    "    destination,\n",
-    "    aircraft,\n",
-    "    weather_date=weather_date,\n",
-    "    heuristic_name=heuristic,\n",
-    "    perf_model_name=\"PS\",  # a/c performance model\n",
-    "    objective=cost_function,\n",
-    "    fuel_loop=False,\n",
-    "    graph_width=\"normal\",\n",
-    ")\n",
-    "\n",
-    "domain = domain_factory()\n",
-    "\n",
-    "with Astar(\n",
-    "    heuristic=lambda d, s: d.heuristic(s), domain_factory=domain_factory, parallel=False\n",
-    ") as solver:\n",
-    "    solver.solve()\n",
-    "    domain.custom_rollout(solver=solver)"
+    "solver = Astar(\n",
+    "    domain_factory=domain_factory,\n",
+    "    heuristic=lambda d, s: d.heuristic(s),\n",
+    "    parallel=False,\n",
+    ")"
    ]
   },
   {
-   "cell_type": "markdown",
+   "cell_type": "code",
+   "execution_count": null,
    "metadata": {},
+   "outputs": [],
    "source": [
-    "## Longer routes\n",
-    "\n",
-    "Finally, we will make a long haul flight: Cartagena (ICAO: SKCG) - Toulouse (ICAO: LFBO), with an A380. For computational purposes, we will also modify the heuristic function:"
+    "solver.solve()"
    ]
   },
   {
@@ -257,37 +280,20 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "heuristic = \"lazy_fuel\"\n",
-    "\n",
-    "origin = \"SKCG\"\n",
-    "destination = \"LFBO\"\n",
-    "aircraft = \"A388\"\n",
-    "\n",
-    "domain_factory = lambda: FlightPlanningDomain(\n",
-    "    origin,\n",
-    "    destination,\n",
-    "    aircraft,\n",
-    "    weather_date=weather_date,\n",
-    "    heuristic_name=heuristic,\n",
-    "    perf_model_name=\"PS\",  # a/c performance model\n",
-    "    objective=cost_function,\n",
-    "    fuel_loop=False,\n",
-    "    graph_width=\"normal\",\n",
-    ")\n",
-    "\n",
-    "domain = domain_factory()\n",
-    "\n",
-    "with Astar(\n",
-    "    heuristic=lambda d, s: d.heuristic(s), domain_factory=domain_factory, parallel=False\n",
-    ") as solver:\n",
-    "    solver.solve()\n",
-    "    domain.custom_rollout(solver=solver)"
+    "domain.custom_rollout(solver=solver, make_img=True)"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
   }
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "Python 3 (ipykernel)",
+   "display_name": "test_dev_skdecide",
    "language": "python",
    "name": "python3"
   },
@@ -301,7 +307,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.10.13"
+   "version": "3.9.19"
   }
  },
  "nbformat": 4,

diff --git a/skdecide/hub/domain/flight_planning/aircraft_performance/base.py b/skdecide/hub/domain/flight_planning/aircraft_performance/base.py