Use ESPHOMe ans Dallas temperature sensors to compute hot water volume in belectrical boiler
Volume calculation is made with Dallas sensors sticked on the inner metallic surface of the electrical boiler.
Main assumptions are :
- boiler tank is a vertical cylinder
- hot water is extrated from top of the boiler tank
- Cold water supply is made ate the bottom of the boiler tank
- Temperature is stratified inside the tank whith high steep gradient making temperature front
- Boiler tank volume is divided in 4 subvolumes, using 5 Dallas temperature sensors set a 0%, 25%, 50%, 75% and 100% height of the tank
Data computed
- Available hot water volume (available volume of water that is above 40°C)
- Useful hot water volume (available volume of hot water diluted to 40°C at the tap/shower)
Calculation Strategy :
- top down calculation, from higher subvolume to lower
- for each subvolume :
- if upper temperature sensor < 40°c => subvolume = 0
- if upper ANd lower temperature sensor > 40°c => subvolume = full subvolume
- if upper temperaure sensor > 40°C and lower temperature sensor < 40°C => subvolume = linear interpolation between upper and lower TT and calculation of the part of the subvolume that is >40°C
Sensors calibration : Sensors are sticked against the inner metallic tank of the boiler using thermal paste
- 0% level is cold water supply => take some tapwater and check T° using the sensor then compare to the 0% level once the tank is at last 50% full of cold water (i.e. for example after daily usage)
- 100% level is max water temperature => take some hot water at the tap and put it in dewar bottle, and check temperature using the sensor. Compare to the 100% tempearture value measured on tank metal surface
From measurements build calibration :
filters:
- calibrate_linear:
method: least_squares
datapoints:
# Map 0.0 (from sensor) to 1.0 (true value)
- 14.9 -> 10.9 #at 0% surface tank -> at tap water
- 46.5 -> 54.5 #at 100% surfacetank -> at hot tap water
Sensor hardware addresses :
Dallas sensor addresses "address: 0xbd3335d446b8a228" are unique. Should be checked after first start of the ESPhome firmware to adjust config :
Parameters
Set of parameters used for th ecalculation, adjust according to your system :
float temp_ref = 40.0; # reference temperature => useful temperature at shower
float vol_ref = 25.0; # subvolume, depends on the number of sensors
float temp_reseau = 15.0; # cold water temperature
float temp_reseau_ref = 15.0; # default cold water temperature
float vol_total = 300; # total useful volume of the tank
NOTE : esp-idf framework is used here instead of arduino to eanble bluetooth proxy extendend capabilities and 801.11k / 802.11v wifi roaming capabilities. Works on arduino framework, adapt to your configuraition.
substitutions:
device_name: "Niveau Chauffe Eau"
dallas_hub_1_pin: GPIO22
dallas_hub_2_pin: GPIO19
dallas_hub_3_pin: GPIO21
dallas_hub_4_pin: GPIO25 #a changer
dallas_hub_5_pin: GPIO23
esphome:
name: "niveau-chauffe-eau"
min_version: 2025.10.4
build_path: build/niveu-chauffe-eau
esp32:
board: esp32dev
framework:
type: esp-idf
esp32_ble:
max_connections: 5
esp32_ble_tracker:
scan_parameters:
interval: 320ms
window: 300ms
active: true
bluetooth_proxy:
active: true
connection_slots: 5
# WiFi Configuration
wifi:
networks:
# - ssid: !secret iot_ssid
# password: !secret iot_password
ssid: !secret wifi_ssid
password: !secret wifi_password
manual_ip:
static_ip: 192.168.0.224 #a changer
gateway: 192.168.0.1
subnet: 255.255.255.0
#output_power: 20.5dB
#power_save_mode: NONE
# Wifi 802.11k Settings
# enable_btm: true
# enable_rrm: true
ap:
ssid: "${device_name}"
password: !secret temp_chauffe_eau_admin
# Enable Web Server for local access
#web_server:
# local: true
# port: 80
# Enable API for Home Assistant integration
api:
encryption:
key: !secret temp_chauffe_eau_api_key
# Logger Configuration
logger:
level: DEBUG
# OTA (Over-the-Air) updates
ota:
- platform: esphome
password: !secret temp_chauffe_eau_admin
# Time Configuration (sync with Home Assistant)
time:
- platform: homeassistant
timezone: "Europe/Paris"
id: homeassistant_time
# Binary Sensor for device status
binary_sensor:
- platform: status
name: "${device_name} Status"
# Button to restart the device
button:
- platform: restart
name: "${device_name} Restart"
# One-wire temperature sensors
one_wire:
- platform: gpio
id: dallas_hub_1
pin: ${dallas_hub_1_pin}
- platform: gpio
id: dallas_hub_2
pin: ${dallas_hub_2_pin}
- platform: gpio
id: dallas_hub_3
pin: ${dallas_hub_3_pin}
- platform: gpio
id: dallas_hub_4
pin: ${dallas_hub_4_pin}
- platform: gpio
id: dallas_hub_5
pin: ${dallas_hub_5_pin}
# # Dallas Temperature Sensors
sensor:
- platform: dallas_temp
one_wire_id: dallas_hub_1
address: 0xbd3335d446b8a228
id: sensor_1
name: "Temperature Niveau 75%"
update_interval: 60s
resolution: 11
device_class: "temperature"
state_class: "measurement"
unit_of_measurement: "°C"
icon: mdi:thermometer
filters:
- calibrate_linear:
method: least_squares
datapoints:
# Map 0.0 (from sensor) to 1.0 (true value)
- 14.9 -> 10.9
- 46.5 -> 54.5
- clamp:
min_value: 5
max_value: 65
ignore_out_of_range: true
- sliding_window_moving_average:
window_size: 10
send_every: 2
- round: 1
- platform: dallas_temp
one_wire_id: dallas_hub_2
# #address: 0xcf79ddd446e17728
id: sensor_2
name: "Temperature Niveau 50%"
update_interval: 60s
resolution: 11
device_class: "temperature"
state_class: "measurement"
unit_of_measurement: "°C"
icon: mdi:thermometer
filters:
- calibrate_linear:
method: least_squares
datapoints:
# Map 0.0 (from sensor) to 1.0 (true value)
- 14.9 -> 10.9
- 46.5 -> 54.5
- clamp:
min_value: 5
max_value: 65
ignore_out_of_range: true
- sliding_window_moving_average:
window_size: 10
send_every: 2
- round: 1
- platform: dallas_temp
one_wire_id: dallas_hub_3
address: 0xef1ff7d446a22628
id: sensor_3
name: "Temperature Niveau 25%"
update_interval: 60s
resolution: 11
device_class: "temperature"
state_class: "measurement"
unit_of_measurement: "°C"
icon: mdi:thermometer
filters:
- calibrate_linear:
method: least_squares
datapoints:
# Map 0.0 (from sensor) to 1.0 (true value)
- 14.9 -> 10.9
- 46.5 -> 54.5
- clamp:
min_value: 5
max_value: 65
ignore_out_of_range: true
- sliding_window_moving_average:
window_size: 10
send_every: 2
- round: 1
- platform: dallas_temp
one_wire_id: dallas_hub_4
# #address: 0xe13ce10457ac2128
id: sensor_4
name: "Temperature Niveau 100%"
update_interval: 60s
resolution: 11
device_class: "temperature"
state_class: "measurement"
unit_of_measurement: "°C"
icon: mdi:thermometer
filters:
- calibrate_linear:
method: least_squares
datapoints:
# Map 0.0 (from sensor) to 1.0 (true value)
- 14.9 -> 10.9
- 46.5 -> 54.5
- clamp:
min_value: 5
max_value: 65
ignore_out_of_range: true
- sliding_window_moving_average:
window_size: 10
send_every: 2
- round: 1
- platform: dallas_temp
one_wire_id: dallas_hub_5
# #address: 0x12839fb30a646128
id: sensor_5
name: "Temperature Niveau 0%"
update_interval: 60s
resolution: 11
device_class: "temperature"
state_class: "measurement"
unit_of_measurement: "°C"
icon: mdi:thermometer
filters:
- calibrate_linear:
method: least_squares
datapoints:
# Map 0.0 (from sensor) to 1.0 (true value)
- 14.9 -> 10.9
- 46.5 -> 54.5
- clamp:
min_value: 5
max_value: 65
ignore_out_of_range: true
- sliding_window_moving_average:
window_size: 10
send_every: 2
- round: 1
- platform: template
name: "Volume utile (> 40°C)"
unit_of_measurement: "L"
icon: mdi:gauge
device_class: "volume_storage"
state_class: "measurement"
update_interval: 60s
accuracy_decimals: 1
lambda: |-
float temp_ref = 40.0;
float temp_reseau = 15.0;
float temp_reseau_ref = 15.0;
float vol_total = 300;
float vol_ref_100_75 = (30.0/140.0)*100.0;
float vol_ref_75_50 = (40.0/140.0)*100.0;
float vol_ref_50_25 = (40.0/140.0)*100.0;
float vol_ref_25_0 = (30.0/140.0)*100.0;
float temp_75 = id(sensor_1).state;
float temp_50 = id(sensor_2).state;
float temp_25 = id(sensor_3).state;
float temp_100 = id(sensor_4).state;
float temp_0 = id(sensor_5).state;
float tangente_100_75 = (temp_75 - temp_100) / vol_ref_100_75;
float tangente_75_50 = (temp_50 - temp_75) / vol_ref_75_50;
float tangente_50_25 = (temp_25 - temp_50) / vol_ref_50_25;
float tangente_25_0 = (temp_0 - temp_25) / vol_ref_25_0;
float temp_moy_100_75 = (temp_100 + temp_75) / 2.0;
float temp_moy_75_50 = (temp_75 + temp_50) / 2.0;
float temp_moy_50_25 = (temp_50 + temp_25) / 2.0;
float temp_moy_25_0 = (temp_25 + temp_0) / 2.0;
float vol_utile_100_75 = 0.0;
float vol_utile_75_50 = 0.0;
float vol_utile_50_25 = 0.0;
float vol_utile_25_0 = 0.0;
float vol_utile = 0.0;
if (temp_0 < temp_reseau_ref) {
temp_reseau = temp_0;
} else {
temp_reseau = temp_reseau_ref;
}
if (temp_100 >= temp_ref) {
if (temp_75 >= temp_ref) {
vol_utile_100_75 = vol_ref_100_75;
} else {
vol_utile_100_75 = (temp_ref-temp_100)/tangente_100_75;
}
} else {
vol_utile_100_75 = 0.0;
}
if (temp_100 >=temp_ref) {
if (temp_75 >= temp_ref) {
if (temp_50 >= temp_ref) {
vol_utile_75_50 = vol_ref_75_50;
} else {
vol_utile_75_50 = (temp_ref-temp_75)/tangente_75_50;
}
} else {
vol_utile_75_50 = 0.0;
}
} else {
vol_utile_75_50 = 0.0;
}
if (temp_75 >= temp_ref and temp_100 >=temp_ref) {
if (temp_50 >= temp_ref) {
if (temp_25 >= temp_ref) {
vol_utile_50_25 = vol_ref_50_25;
} else {
vol_utile_50_25 = (temp_ref-temp_50)/tangente_50_25;
}
} else {
vol_utile_50_25 = 0.0;
}
} else {
vol_utile_50_25 = 0.0;
}
if (temp_50 >= temp_ref and temp_75 >= temp_ref and temp_100 >=temp_ref) {
if (temp_25 >= temp_ref) {
if (temp_0 >= temp_ref) {
vol_utile_25_0 = vol_ref_25_0;
} else {
vol_utile_25_0 = (temp_ref-temp_25)/tangente_25_0;
}
} else {
vol_utile_25_0 = 0.0;
}
} else {
vol_utile_25_0 = 0.0;
}
vol_utile = (vol_utile_100_75 + vol_utile_75_50 + vol_utile_50_25 + vol_utile_25_0)*vol_total/100.0;
return vol_utile;
- platform: template
name: "Volume exploitable (eq. 40°C)"
unit_of_measurement: "L"
icon: mdi:gauge
device_class: "volume_storage"
state_class: "measurement"
update_interval: 60s
accuracy_decimals: 1
lambda: |-
float temp_ref = 40.0;
float vol_ref_100_75 = (30.0/140.0)*100.0;
float vol_ref_75_50 = (40.0/140.0)*100.0;
float vol_ref_50_25 = (40.0/140.0)*100.0;
float vol_ref_25_0 = (30.0/140.0)*100.0;
float temp_reseau = 15.0;
float temp_reseau_ref = 15.0;
float vol_total = 300;
float temp_75 = id(sensor_1).state;
float temp_50 = id(sensor_2).state;
float temp_25 = id(sensor_3).state;
float temp_100 = id(sensor_4).state;
float temp_0 = id(sensor_5).state;
float tangente_100_75 = (temp_75 - temp_100) / vol_ref_100_75;
float tangente_75_50 = (temp_50 - temp_75) / vol_ref_75_50;
float tangente_50_25 = (temp_25 - temp_50) / vol_ref_50_25;
float tangente_25_0 = (temp_0 - temp_25) / vol_ref_25_0;
float temp_moy_100_75 = (temp_100 + temp_75) / 2.0;
float temp_moy_75_50 = (temp_75 + temp_50) / 2.0;
float temp_moy_50_25 = (temp_50 + temp_25) / 2.0;
float temp_moy_25_0 = (temp_25 + temp_0) / 2.0;
float vol_expl_100_75 = 0.0;
float vol_expl_75_50 = 0.0;
float vol_expl_50_25 = 0.0;
float vol_expl_25_0 = 0.0;
float vol_expl = 0.0;
if (temp_0 < temp_reseau_ref) {
temp_reseau = temp_0;
} else {
temp_reseau = temp_reseau_ref;
}
if (temp_100 >= temp_ref) {
if (temp_75 >= temp_ref) {
vol_expl_100_75 = vol_ref_100_75 * (1 + (temp_moy_100_75 - temp_ref)/(temp_ref - temp_reseau));
} else {
vol_expl_100_75 = ((temp_ref-temp_100)/tangente_100_75) * (1 + (temp_moy_100_75 - temp_ref)/(temp_ref - temp_reseau));
}
} else {
vol_expl_100_75 = 0.0;
}
if (temp_100 >=temp_ref) {
if (temp_75 >= temp_ref) {
if (temp_50 >= temp_ref) {
vol_expl_75_50 = vol_ref_75_50 * (1 + (temp_moy_75_50 - temp_ref)/(temp_ref - temp_reseau));
} else {
vol_expl_75_50 = ((temp_ref-temp_75)/tangente_75_50) * (1 + (temp_moy_75_50 - temp_ref)/(temp_ref - temp_reseau));
}
} else {
vol_expl_75_50 = 0.0;
}
} else {
vol_expl_75_50 = 0.0;
}
if (temp_75 >= temp_ref and temp_100 >=temp_ref) {
if (temp_50 >= temp_ref) {
if (temp_25 >= temp_ref) {
vol_expl_50_25 = vol_ref_50_25 * (1 + (temp_moy_50_25 - temp_ref)/(temp_ref - temp_reseau));
} else {
vol_expl_50_25 = ((temp_ref-temp_50)/tangente_50_25) * (1 + (temp_moy_50_25 - temp_ref)/(temp_ref - temp_reseau));
}
} else {
vol_expl_50_25 = 0.0;
}
} else {
vol_expl_50_25 = 0.0;
}
if (temp_50 >= temp_ref and temp_75 >= temp_ref and temp_100 >=temp_ref) {
if (temp_25 >= temp_ref) {
if (temp_0 >= temp_ref) {
vol_expl_25_0 = vol_ref_25_0 * (1 + (temp_moy_25_0 - temp_ref)/(temp_ref - temp_reseau));
} else {
vol_expl_25_0 = ((temp_ref-temp_25)/tangente_25_0) * (1 + (temp_moy_25_0 - temp_ref)/(temp_ref - temp_reseau));
}
} else {
vol_expl_25_0 = 0.0;
}
} else {
vol_expl_25_0 = 0.0;
}
vol_expl = (vol_expl_100_75 + vol_expl_75_50 + vol_expl_50_25 + vol_expl_25_0)*vol_total/100.0;
return vol_expl;
- platform: wifi_signal
name: "WiFi Signal"
update_interval: 60s
safe_mode:
disabled: False
# --- BUTTONS ---
button:
- platform: restart
name: "Restart"
- platform: safe_mode
name: "Restart Safe Mode"
entity_category: config
Using Gemini and Claude AI, i've upgraded the code to make it more faul tolerant, improve documentation and add a new calcuilation mathod based on Hyperbolic Tangent over the whole volume instead of piecewise linear interpolation.
It results in a more "physical" temperature front identification and a more robust energy calculation and availability :
I also changed the GPIO to avoid strapping pin
Here's the upgraded code :
substitutions:
device_name: "Chauffe Eau"
# Pin Definitions (Matching 1-Wire Buses)
dallas_hub_1_pin: GPIO22
dallas_hub_2_pin: GPIO19
dallas_hub_3_pin: GPIO21
dallas_hub_4_pin: GPIO25
dallas_hub_5_pin: GPIO23
esphome:
name: "sonde-chauffe-eau"
friendly_name: Sonde-Chauffe-Eau
min_version: 2024.11.2
build_path: build/niveau-chauffe-eau
esp32:
board: esp32dev
framework:
type: esp-idf
# --- BLUETOOTH CONFIGURATION ---
esp32_ble:
max_connections: 5
esp32_ble_tracker:
scan_parameters:
interval: 320ms
window: 300ms
active: true
bluetooth_proxy:
active: true
connection_slots: 5
# --- WIFI CONFIGURATION ---
wifi:
networks:
- ssid: !secret iot_ssid
password: !secret iot_password
- ssid: !secret wifi_ssid
password: !secret wifi_password
manual_ip:
static_ip: 192.168.0.24
gateway: 192.168.0.1
subnet: 255.255.255.0
# Power Save OFF for BT stability
power_save_mode: NONE
# Roaming Support
enable_btm: true
enable_rrm: true
ap:
ssid: "${device_name}"
password: !secret temp_chauffe_eau_admin
api:
encryption:
key: !secret temp_chauffe_eau_api_key
logger:
level: DEBUG
ota:
- platform: esphome
password: !secret temp_chauffe_eau_admin
time:
- platform: homeassistant
timezone: "Europe/Paris"
id: homeassistant_time
# --- ONE-WIRE BUSES (1 per Pin) ---
one_wire:
- platform: gpio
id: bus_1
pin: ${dallas_hub_1_pin} # 75%
- platform: gpio
id: bus_2
pin: ${dallas_hub_2_pin} # 50%
- platform: gpio
id: bus_3
pin: ${dallas_hub_3_pin} # 25%
- platform: gpio
id: bus_4
pin: ${dallas_hub_4_pin} # 100%
- platform: gpio
id: bus_5
pin: ${dallas_hub_5_pin} # 0%
# --- SENSORS ---
sensor:
# System Health Sensors
- platform: internal_temperature
name: "Température CPU ESP32"
update_interval: 60s
- platform: wifi_signal
name: "${device_name} WiFi Signal"
update_interval: 30s # 30-second update frequency
# Common Filter Block for all DS18B20 sensors
filters: &common_filters
# Calibration Filter (Applying your linear map)
- calibrate_linear:
method: least_squares
datapoints:
- 14.9 -> 10.9
- 46.5 -> 54.5
# Sanity Check for Temperature Range
- clamp:
min_value: 5
max_value: 85
ignore_out_of_range: true
# Smoothing Filter (Moving Average)
- sliding_window_moving_average:
window_size: 10
send_every: 2
- round: 1
# Dallas Temperature Sensors (5 Sensors)
- platform: dallas_temp
one_wire_id: bus_1
address: 0xbd3335d446b8a228
id: sensor_1
name: "Temperature Niveau 75%"
update_interval: 60s
resolution: 11
filters: *common_filters
- platform: dallas_temp
one_wire_id: bus_2
address: 0xcf79ddd446e17728
id: sensor_2
name: "Temperature Niveau 50%"
update_interval: 60s
resolution: 11
filters: *common_filters
- platform: dallas_temp
one_wire_id: bus_3
address: 0xef1ff7d446a22628
id: sensor_3
name: "Temperature Niveau 25%"
update_interval: 60s
resolution: 11
filters: *common_filters
- platform: dallas_temp
one_wire_id: bus_4
address: 0xe13ce10457ac2128
id: sensor_4
name: "Temperature Niveau 100%"
update_interval: 60s
resolution: 11
filters: *common_filters
- platform: dallas_temp
one_wire_id: bus_5
address: 0x12839fb30a646128
id: sensor_5
name: "Temperature Niveau 0%"
update_interval: 60s
resolution: 11
filters: *common_filters
# --- VIRTUAL VOLUME 1: USEFUL VOLUME (GEOMETRIC) ---
# Calculates the volume physically above 40°C using Tanh/Logit interpolation.
- platform: template
name: "Volume utile (> 40°C)"
unit_of_measurement: "L"
icon: mdi:chart-bell-curve-cumulative
device_class: "volume_storage"
state_class: "measurement"
update_interval: 60s
accuracy_decimals: 1
lambda: |-
const float TARGET_TEMP = 40.0;
const float TOTAL_VOL = 300.0;
// Sensor Positions (Height percentage 0.0 to 1.0)
// Sorted from Bottom (Cold) to Top (Hot)
const float heights[] = {0.0, 0.25, 0.50, 0.75, 1.0};
// Get Readings (Sorted Bottom to Top)
float temps[5];
temps[0] = id(sensor_5).state; // 0%
temps[1] = id(sensor_3).state; // 25%
temps[2] = id(sensor_2).state; // 50%
temps[3] = id(sensor_1).state; // 75%
temps[4] = id(sensor_4).state; // 100%
for (float t : temps) { if (isnan(t)) return {}; }
if (temps[4] < TARGET_TEMP) return 0.0;
if (temps[0] >= TARGET_TEMP) return TOTAL_VOL;
// Define Asymptotes for Normalization (Safety margin to avoid log(0))
float t_min = temps[0] - 1.0;
float t_max = temps[4] + 1.0;
// Function to transform Temperature to Logit Space (Linearizing the Sigmoid)
auto to_logit = [&](float t) -> float {
float norm = (t - t_min) / (t_max - t_min);
// Clamp to prevent log domain errors
if (norm <= 0.01) norm = 0.01;
if (norm >= 0.99) norm = 0.99;
return log(norm / (1.0 - norm));
};
// 1. Find the segment containing 40°C
int idx = -1;
for (int i = 0; i < 4; i++) {
if (temps[i] <= TARGET_TEMP && temps[i+1] >= TARGET_TEMP) {
idx = i;
break;
}
}
if (idx == -1) return 0.0;
// 2. Perform Linear Interpolation in Logit Space
float y_lower = to_logit(temps[idx]);
float y_upper = to_logit(temps[idx+1]);
float y_target = to_logit(TARGET_TEMP);
float progress_fraction;
if (abs(y_upper - y_lower) < 0.0001) {
progress_fraction = 0.5; // Avoid div/0 if temps are perfectly equal
} else {
progress_fraction = (y_target - y_lower) / (y_upper - y_lower);
}
// 3. Calculate Height of the 40°C point
float h_lower = heights[idx];
float h_upper = heights[idx+1];
float height_of_40 = h_lower + (h_upper - h_lower) * progress_fraction;
// Volume is the portion ABOVE this height (1.0 - height_of_40)
float useful_ratio = 1.0 - height_of_40;
// Final Clamp
if (useful_ratio < 0) useful_ratio = 0;
if (useful_ratio > 1) useful_ratio = 1;
return useful_ratio * TOTAL_VOL;
# --- VIRTUAL VOLUME 2: EXPLOITABLE VOLUME (THERMODYNAMIC) ---
# Uses Numerical Integration (Riemann Sum) of the Tanh curve to calculate
# the equivalent volume of 40°C water that can be obtained by mixing.
- platform: template
name: "Volume exploitable (eq. 40°C)"
unit_of_measurement: "L"
icon: mdi:function-variant
device_class: "volume_storage"
state_class: "measurement"
update_interval: 60s
accuracy_decimals: 1
lambda: |-
// --- CONSTANTS ---
const float TARGET_TEMP = 40.0;
const float TOTAL_VOL = 300.0;
const float TEMP_COLD_DEF = 15.0;
const int INTEGRATION_STEPS = 50;
// Sensor Heights and Temps (Sorted Bottom to Top)
const float h_sensors[] = {0.0, 0.25, 0.50, 0.75, 1.0};
float t_sensors[5];
t_sensors[0] = id(sensor_5).state; t_sensors[1] = id(sensor_3).state;
t_sensors[2] = id(sensor_2).state; t_sensors[3] = id(sensor_1).state;
t_sensors[4] = id(sensor_4).state;
for (float t : t_sensors) { if (isnan(t)) return {}; }
// --- DYNAMIC PARAMETERS ---
float t_cold = (t_sensors[0] < TEMP_COLD_DEF) ? t_sensors[0] : TEMP_COLD_DEF;
// Global Asymptotes for Sigmoid Modeling (Safety margin)
float t_min_limit = t_sensors[0] - 2.0;
float t_max_limit = t_sensors[4] + 2.0;
// Helper Function: The Inverse Logit (T -> L)
auto to_logit = [&](float t) -> float {
float norm = (t - t_min_limit) / (t_max_limit - t_min_limit);
if (norm <= 0.001) norm = 0.001;
if (norm >= 0.999) norm = 0.999;
return log(norm / (1.0 - norm));
};
// Helper Function: The Sigmoid (L -> T)
auto get_temp_at_height = [&](float h_query) -> float {
// 1. Find segment index
int idx = 0;
for(int i=0; i<4; i++) { if (h_query >= h_sensors[i]) idx = i; }
// 2. Convert endpoints to Logit Space
float L_lower = to_logit(t_sensors[idx]);
float L_upper = to_logit(t_sensors[idx+1]);
// 3. Linear Interpolation in Logit Space
float h_lower = h_sensors[idx];
float h_upper = h_sensors[idx+1];
float progress = (h_query - h_lower) / (h_upper - h_lower);
float L_interpolated = L_lower + progress * (L_upper - L_lower);
// 4. Convert back from Logit to Temp (Sigmoid)
float norm_interpolated = 1.0 / (1.0 + exp(-L_interpolated));
return t_min_limit + norm_interpolated * (t_max_limit - t_min_limit);
};
// --- NUMERICAL INTEGRATION (Riemann Sum) ---
float total_enthalpy_ratio = 0.0;
float slice_height = 1.0 / INTEGRATION_STEPS;
float denom = TARGET_TEMP - t_cold;
// Safety check for target temperature relative to cold water
if (denom < 1.0) denom = 1.0;
for (int k = 0; k < INTEGRATION_STEPS; k++) {
// Sample the temperature at the CENTER of the slice
float h_center = (k + 0.5) * slice_height;
float t_slice = get_temp_at_height(h_center);
// Calculate mixing potential of this slice
// Ratio = (T_slice - T_cold) / (T_target - T_cold)
float ratio = (t_slice - t_cold) / denom;
// Physics constraint: only positive energy contributes to 'Exploitable' volume
if (ratio < 0) ratio = 0;
total_enthalpy_ratio += ratio;
}
// Final Calculation: Average Ratio * Total Volume
float effective_volume = (total_enthalpy_ratio / INTEGRATION_STEPS) * TOTAL_VOL;
return effective_volume;
- platform: wifi_signal
name: "WiFi Signal"
update_interval: 60s
safe_mode:
disabled: False
# --- BUTTONS ---
button:
- platform: restart
name: "Restart"
- platform: safe_mode
name: "Restart Safe Mode"
entity_category: config