Planning-A into s2025_teamA

GEMstack/knowledge/defaults/computation_graph.yaml

@@ -31,7 +31,7 @@ components:
       inputs: [vehicle, roadgraph, agents]
       outputs: agent_intents
   - relations_estimation:
-      inputs: [vehicle, roadgraph, agents, obstacles]
+      inputs: all
       outputs: relations
   - predicate_evaluation:
       inputs: [vehicle, roadgraph, agents, obstacles]

GEMstack/knowledge/defaults/current.yaml

@@ -20,6 +20,18 @@ control:
         pid_p: 0.8
         pid_i: 0.03
         pid_d: 0.0
+planning:
+    longitudinal_plan:
+        mode:               'real' # 'real' or 'sim'
+        yielder:            'expert' # 'expert', 'analytic', or 'simulation'
+        planner:            'dt' # 'milestone', 'dt', or 'dx'
+        desired_speed:      1.0
+        acceleration:       0.5
+        max_deceleration:   6.0
+        deceleration:       2.0
+        min_deceleration:   0.5
+        yield_deceleration: 0.5
+
 
 #configure the simulator, if using
 simulator:

GEMstack/mathutils/quadratic_equation.py

@@ -1,6 +1,16 @@
 import math
 
 def quad_root(a : float, b : float, c : float) -> float:
+    """Calculates the roots of a quadratic equation
+
+    Args:
+        a (float): coefficient of x^2
+        b (float): coefficient of x
+        c (float): constant term
+
+    Returns:
+        float: returns all valid roots of the quadratic equation
+    """
     x1 = (-b + max(0,(b**2 - 4*a*c))**0.5)/(2*a)
     x2 = (-b - max(0,(b**2 - 4*a*c))**0.5)/(2*a)
 

GEMstack/onboard/planning/README.md

@@ -0,0 +1,50 @@
+# Planning-A HW1 Document
+
+This document serves as an overview for integrating pedestrian yielder into longitudinal planner. It describes how parameters from `pedetrian_detecion.yaml` are used in the planning system along with references to the `longitudinal_planning.py` script and the `pedestrian_yield_logic.py` script.
+
+Details on the algorithm and demos are availavle at the link below:<br>
+https://drive.google.com/drive/folders/1aSOWIrqXjf9-j6i1S_MJksATMyyoP31_?usp=sharing
+
+
+## Contribution
+Each member contributed to:
+### longitudinal_planning.py Contributions
+
+| Algorithm  | Contributor     |
+| :--------- | :-------------- |
+| milestone  | Simon (sk106)   |
+| dt         | Rohit (srm17)   |
+| dx         | Hansen (hl58)   |
+
+### pedestrian_yield_logic.py Contributions
+
+| Algorithm  | Contributor                                 |
+| :--------- | :------------------------------------------ |
+| expert     | Patrick (bohaowu2), Animesh (animesh8)      |
+| analytic   | Henry (weigang2)                            |
+| simulation | Yudai (yyamada2)                            |
+
+
+## Configuration Details
+The `pedetrian_detecion.yaml` file includes:
+```
+args:
+    mode: 'real'
+    params: {
+        'yielder':       'expert', # 'expert', 'analytic', or 'simulation'
+        'planner':       'milestone', # 'milestone', 'dt', or 'dx'
+        'desired_speed': 1.0,  # m/s
+        'acceleration':  0.75  # m/s2
+        }
+```
+
+
+## Usage
+### pedetrian_detecion.yaml
+Algorithm can be chosen from the above in this file. Also contains parameters for the logic.
+### Execution: Simulator
+`python3 main.py --variant=fake_sim launch/pedestrian_detection.yaml`
+### Testing
+- `testing/test_longitudinal_plan.py`
+- `testing/test_yield_logic_analytic.ipynb`
+- `testing/test_collision_detection.py`