File size: 10,799 Bytes
5a22808 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 | """
Mock Data Server for Thermal Grid RL Agent Hackathon Demo
=========================================================
Loads the 3 real-world CSV datasets and exposes REST endpoints
that the RL environment can call instead of using hardcoded values.
Endpoints
---------
GET /ambient-temp?hour=<0-23>
GET /carbon-intensity?month=<1-12>
GET /pue-benchmark?region=india
GET /energy-price?hour=<0-23>
GET /off-peak?hour=<0-23>
GET /signals?hour=<0-23>&month=<1-12> ← all-in-one
Run: uvicorn mock_data_server:app --reload --port 8000
"""
from __future__ import annotations
import os
import sys
import random
from datetime import datetime
from pathlib import Path
from typing import Optional
import numpy as np
import pandas as pd
# ---------------------------------------------------------------------------
# FastAPI (optional – falls back to stdlib SimpleHTTPServer if not installed)
# ---------------------------------------------------------------------------
try:
from fastapi import FastAPI, Query
from fastapi.responses import JSONResponse
import uvicorn
_HAS_FASTAPI = True
except ImportError:
_HAS_FASTAPI = False
FastAPI = None # type: ignore
# ---------------------------------------------------------------------------
# Paths
# ---------------------------------------------------------------------------
DATA_DIR = Path(__file__).parent / "data"
WEATHER_CSV = DATA_DIR / "india_2000_2024_daily_weather.csv"
CARBON_CSV = DATA_DIR / "india_monthly_full_release.csv"
REGION_CSV = DATA_DIR / "region-metadata.csv"
# ---------------------------------------------------------------------------
# Lazy Data Loading
# ---------------------------------------------------------------------------
_DATA_LOADED = False
_monthly_temp: dict[int, float] = {}
_monthly_carbon: dict[int, float] = {}
INDIA_BENCHMARK_PUE: float = 1.58
def load_data():
"""Load and pre-process CSV data (once)."""
global _DATA_LOADED, _monthly_temp, _monthly_carbon, INDIA_BENCHMARK_PUE
if _DATA_LOADED:
return
try:
print("[MockServer] Loading weather data …")
if not WEATHER_CSV.exists():
raise FileNotFoundError(f"Weather CSV not found: {WEATHER_CSV}")
_weather_df = pd.read_csv(WEATHER_CSV, parse_dates=["date"])
# Focus on Delhi as representative Indian DC location
_delhi_weather = _weather_df[_weather_df["city"] == "Delhi"].copy()
_delhi_weather["month"] = _delhi_weather["date"].dt.month
# Build hourly ambient temp by month: use mean of max+min
for month in range(1, 13):
sub = _delhi_weather[_delhi_weather["month"] == month]
if len(sub):
_monthly_temp[month] = float(
(sub["temperature_2m_max"].mean() + sub["temperature_2m_min"].mean()) / 2
)
else:
_monthly_temp[month] = 25.0
print("[MockServer] Loading carbon intensity data …")
if not CARBON_CSV.exists():
raise FileNotFoundError(f"Carbon CSV not found: {CARBON_CSV}")
# Pre-filtered: only load CO2 intensity rows (headers: Area, Country code, Date, ..., Variable, Unit, Value, ...)
_carbon_df = pd.read_csv(CARBON_CSV, usecols=["Country code", "Variable", "Unit", "Date", "Value"])
_india_carbon = _carbon_df[
(_carbon_df["Variable"] == "CO2 intensity") &
(_carbon_df["Unit"] == "gCO2/kWh")
].copy()
_india_carbon["Date"] = pd.to_datetime(_india_carbon["Date"])
_india_carbon["month"] = _india_carbon["Date"].dt.month
# Monthly average carbon intensity for India
for month in range(1, 13):
sub = _india_carbon[_india_carbon["month"] == month]
if len(sub) and sub["Value"].notna().any():
_monthly_carbon[month] = float(sub["Value"].mean())
else:
_monthly_carbon[month] = 720.0 # India mean ~720 gCO2/kWh
print("[MockServer] Loading region benchmarks …")
if not REGION_CSV.exists():
raise FileNotFoundError(f"Region CSV not found: {REGION_CSV}")
_region_df = pd.read_csv(REGION_CSV)
# India rows
_india_regions = _region_df[_region_df["location"].str.contains("India|Mumbai|Delhi|Hyderabad", na=False)]
_india_pue_values = _india_regions["power-usage-efficiency"].dropna().astype(float)
INDIA_BENCHMARK_PUE = float(_india_pue_values.mean()) if len(_india_pue_values) else 1.58
print(f"[MockServer] India benchmark PUE = {INDIA_BENCHMARK_PUE:.3f}")
_DATA_LOADED = True
print("[MockServer] All data loaded successfully [DONE]")
except Exception as e:
print(f"[MockServer] ERROR: Failed to load data: {e}", file=sys.stderr)
raise
# ---------------------------------------------------------------------------
# Indian electricity tariff – time-of-use schedule (₹/kWh → $/kWh @84 INR/$)
# ---------------------------------------------------------------------------
_INR_TO_USD = 1.0 / 84.0
_TOU_INR: list[float] = [
5.5, 5.5, 5.5, 5.5, 5.5, 6.0, # 00-05 off-peak
7.0, 8.5, 9.5, 10.0, 10.0, 9.5, # 06-11 morning peak
9.0, 8.5, 9.0, 10.0, 11.0, 12.0, # 12-17 afternoon/evening peak
12.0, 11.5, 10.0, 8.5, 7.0, 6.0, # 18-23 evening taper
]
_TOU_USD = [p * _INR_TO_USD for p in _TOU_INR]
# Off-peak: midnight to 6 AM
_OFF_PEAK_HOURS = set(range(0, 6))
# Diurnal temperature adjustment (how much hotter/cooler relative to monthly mean)
_HOUR_TEMP_DELTA: list[float] = [
-4.0, -4.5, -5.0, -5.0, -4.5, -3.0, # 00-05 cool night
-2.0, 0.0, 2.0, 4.0, 5.0, 5.5, # 06-11 morning rise
5.5, 5.0, 4.5, 4.0, 3.0, 1.5, # 12-17 peak then fall
0.0, -1.0, -2.0, -2.5, -3.0, -3.5, # 18-23 evening drop
]
# Solar ramp factor for carbon intensity (lower at midday due to solar)
_SOLAR_FACTOR: list[float] = [
1.00, 1.00, 1.00, 1.00, 1.00, 0.98, # 00-05
0.95, 0.88, 0.80, 0.72, 0.68, 0.65, # 06-11
0.63, 0.65, 0.68, 0.75, 0.82, 0.90, # 12-17
0.95, 0.98, 1.00, 1.00, 1.00, 1.00, # 18-23
]
# ---------------------------------------------------------------------------
# Helper functions (used both by FastAPI routes and HybridSignalGenerator)
# ---------------------------------------------------------------------------
def _current_month() -> int:
return datetime.now().month
def get_ambient_temp(hour: int, month: Optional[int] = None) -> float:
"""Return ambient temperature in °C for given hour and month."""
load_data()
month = month or _current_month()
month = max(1, min(12, month))
hour = hour % 24
base = _monthly_temp.get(month, 25.0)
noise = random.gauss(0, 0.5)
return round(base + _HOUR_TEMP_DELTA[hour] + noise, 2)
def get_carbon_intensity(month: Optional[int] = None, hour: int = 12) -> float:
"""Return grid carbon intensity in gCO₂/kWh."""
load_data()
month = month or _current_month()
month = max(1, min(12, month))
hour = hour % 24
base = _monthly_carbon.get(month, 720.0)
noise = random.gauss(0, base * 0.03)
return round(float(np.clip(base * _SOLAR_FACTOR[hour] + noise, 300, 900)), 2)
def get_energy_price(hour: int) -> float:
"""Return energy price in $/kWh."""
hour = hour % 24
noise = random.gauss(0, 0.002)
return round(float(np.clip(_TOU_USD[hour] + noise, 0.04, 0.20)), 4)
def get_off_peak(hour: int) -> int:
"""Return 1 if off-peak, else 0."""
return int((hour % 24) in _OFF_PEAK_HOURS)
def get_pue_benchmark(region: str = "india") -> float:
"""Return real-world PUE benchmark for specified region."""
load_data()
return round(INDIA_BENCHMARK_PUE, 3)
def get_all_signals(hour: int, month: Optional[int] = None) -> dict:
"""Return all signals in a single dictionary."""
month = month or _current_month()
amb = get_ambient_temp(hour, month)
ci = get_carbon_intensity(month, hour)
price = get_energy_price(hour)
offpk = get_off_peak(hour)
dr = 1 if (hour in {17, 18, 19} and random.random() < 0.25) else 0
pue_b = get_pue_benchmark()
return {
"hour": hour,
"month": month,
"ambient_temp_c": amb,
"energy_price_per_kwh": price,
"grid_carbon_intensity_g_per_kwh": ci,
"off_peak_window": offpk,
"demand_response_signal": dr,
"pue_benchmark": pue_b,
}
# ---------------------------------------------------------------------------
# FastAPI app (only if fastapi is available)
# ---------------------------------------------------------------------------
if _HAS_FASTAPI:
app = FastAPI(
title="Thermal Grid RL – Mock Data Server",
description="Serves real India weather, carbon, and grid data for the RL environment.",
version="1.0.0",
)
@app.get("/health")
def health():
load_data()
return {"status": "ok", "india_benchmark_pue": INDIA_BENCHMARK_PUE}
@app.get("/ambient-temp")
def ambient_temp(
hour: int = Query(12, ge=0, le=23),
month: Optional[int] = Query(None, ge=1, le=12),
):
return {"ambient_temp_c": get_ambient_temp(hour, month)}
@app.get("/carbon-intensity")
def carbon_intensity(
hour: int = Query(12, ge=0, le=23),
month: Optional[int] = Query(None, ge=1, le=12),
):
return {"grid_carbon_intensity_g_per_kwh": get_carbon_intensity(month, hour)}
@app.get("/energy-price")
def energy_price(hour: int = Query(12, ge=0, le=23)):
return {"energy_price_per_kwh": get_energy_price(hour)}
@app.get("/off-peak")
def off_peak(hour: int = Query(12, ge=0, le=23)):
return {"off_peak_window": get_off_peak(hour), "is_off_peak": bool(get_off_peak(hour))}
@app.get("/pue-benchmark")
def pue_benchmark(region: str = Query("india")):
return {"pue_benchmark": get_pue_benchmark(region), "region": region}
@app.get("/signals")
def signals(
hour: int = Query(12, ge=0, le=23),
month: Optional[int] = Query(None, ge=1, le=12),
):
return get_all_signals(hour, month)
else:
print("[MockServer] WARNING: FastAPI not installed. "
"Install with: pip install fastapi uvicorn")
print("[MockServer] The HybridSignalGenerator can still run in 'direct' mode.")
app = None # type: ignore
# ---------------------------------------------------------------------------
# Fallback: direct-call mode (no HTTP, used by HybridSignalGenerator)
# ---------------------------------------------------------------------------
if __name__ == "__main__":
uvicorn.run("mock_data_server:app", host="0.0.0.0", port=8001, reload=True)
|