Files
brewpi/components/pid/temp_controller_smith.py
T
jensandClaude Sonnet 4.6 f1688d3c71 Derive Pot's L/Td from the Sud doc instead of a fixed server default
Sud now parses top-level L/Td fields from sude/*.json (defaulting to 0.2/30,
matching the old hardcoded server constants), and derive_plant_params()
returns them alongside M/C - constant for the whole brew, unlike M/C which
vary per step with grain/water mass. All sude/*.json docs gain explicit
L/Td fields.

Callers (tasks/sud.py, SudForecastEstimator, demo_sud.py) switch from the
narrow set_thermal_params(M, C)/set_model_params(M, C) to the full
set_plant_params(params)/set_model_plant_params(params), since
derive_plant_params() now always returns all four keys; both narrow setters
are removed as dead code.

Caught along the way: Pot.set_plant_params() unconditionally rebuilt the
Delay ring buffer, which was harmless when only ever called once at
construction - but now running on every Sud step change, it was wiping
in-flight delayed power at each step boundary. Fixed to only rebuild when
Td actually changes; verified this restores the exact original forecast
result for sud_0010.json.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DkkuG48uHFCGKe6dPSERFk
2026-06-22 18:45:07 +02:00

76 lines
2.7 KiB
Python
Executable File

from components.plant.pot import Pot
from components.pid.temp_controller_base import TempControllerBase, States
class TempController(TempControllerBase):
def __init__(self, dt):
TempControllerBase.__init__(self, dt)
self.dt = dt
self.last_theta_ist = 20
self.heatrate_ist = 0
# Fast model: same plant model but with zero transport delay, used
# to predict the current temperature without the dead time.
self.model = Pot(dt)
# Delayed model: keeps the plant's assumed transport delay, so it
# can be compared like-for-like against the real (delayed) measurement.
self.model_delay = Pot(dt)
self.theta_ist_plant = 0
self.theta_ist_model = 0
self.theta_ist_model_delay = 0
def set_model_plant_params(self, model_params):
self.model.set_plant_params({**model_params, 'Td': 0})
self.model_delay.set_plant_params(model_params)
def set_model_power(self, power):
self.model.set_power(power)
self.model_delay.set_power(power)
def set_ambient_temperature(self, theta_amb):
self.model.set_ambient_temperature(theta_amb)
self.model_delay.set_ambient_temperature(theta_amb)
def on_state_entered(self, state):
if state in (States.HEAT, States.COOL):
self.model.initial(self.theta_ist)
self.model_delay.initial(self.theta_ist)
def post_pid(self):
self.model.process()
self.model_delay.process()
def process(self):
self._require_params()
if not (self.model.is_configured() and self.model_delay.is_configured()):
raise RuntimeError(
"{}.process(): model plant params and/or ambient temperature not set - "
"call set_model_plant_params() and set_ambient_temperature() first".format(type(self).__name__))
self.theta_ist_plant = self.theta_ist_set
self.theta_ist_model = self.model.get_temperature()
self.theta_ist_model_delay = self.model_delay.get_temperature()
# Smith predictor: use the fast model's prediction, corrected by the
# mismatch between the real (delayed) plant and the delayed model,
# so the dead time drops out of the feedback path.
self.theta_ist = self.theta_ist_model + (self.theta_ist_plant - self.theta_ist_model_delay)
heatrate = (self.theta_ist - self.last_theta_ist)/self.dt*60
alpha = 0.1
self.heatrate_ist = (1-alpha) * self.heatrate_ist + alpha*heatrate
self.last_theta_ist = self.theta_ist
# Compensate for max heat rate to reduce overshoot
hold_scale = 1.0/self.heatrate_soll_set if self.heatrate_soll_set > 0 else 1.0
self.heatrate_soll = self.heatrate_soll_set * self.pid_hold.get_y()
theta_err = self.theta_soll_set - self.theta_ist
heatrate_err = self.heatrate_soll - self.heatrate_ist
diff = self.theta_soll_set - self.theta_ist
self.process_fsm(diff)
self.process_pid(theta_err, heatrate_err, hold_scale)