DREAM Progress Demand Response Electrical Appliance Manager Thermostat and Controls Group

1
DREAM ProgressDemand Response Electrical
Appliance ManagerThermostat and Controls Group

• Prof. Edward Arens
• Prof. David Auslander
• Research Specialist
• Charlie Huizenga
• Graduate Students
• Xue Chen, Alex Do, Jaehwi Jang, Florian Jourda,
  Anna LaRue, Therese Peffer, William Watts
• Undergraduate Students
• Po-kai Chen, Reman Childs, Yi Yuan

2
Goals of the New Thermostat

• Low cost
• Simple and easy to operate
• Receives price signal from utility
• Control other appliances
• Involve and inform occupant

3
DREAM Demand Response Electrical Appliance
Manager
Utility
Temperature sensors
Power sensor
Price Indicator
Price
Electricity used
Power actuators
Smart Meter
Motion sensors
Background Overview DR
4
DREAM Control Code

• Control of simulated, model, or real houses

Sensor/ActuatorModule
Simulation
XML
Controller
Java
Physical Model
RF
Real House
RF
Interface
5
Test House 2005
6
Test House 2005
_at_ vent
Base
Wind/sun
repeater
Outdoor temps, RH
G
HVAC relay
_at_ main breaker
RH
IR
G
G
G
_at_ vent
Temperature (air), G globe Temperature (air)
RH Occupancy Power sensing (outlet) Power sensing
(breaker panel)
Weather station on roof (not shown) Anemometer
(wind direction and speed) Pyranometer (total
horizontal radiation and diffuse
radiation) Outside Temperature (exposed to night
sky, not exposed) Outside RH
7
System Communication Structure
DREAM
Utility Price
House Environment
8
System Communication Structure
DREAM
Utility Price
9
System Communication Structure
DREAM
Utility Price
House Environment
10
System Communication Structure
Remote Database
House Environment
Utility Price
Controller
Wireless network
Local Database
DREAM
11
System Communication Structure
DREAM
Utility Price
House Environment
12
System Communication Structure
Price Generator
13
System Communication Structure
Remote Database
Internet
Sensors
Controller
Base Mote
Switch
HVAC
LED Indicator
Local Database
Message Transmitter
14
Online Realtime Database
15
Control Code and Hardware Connection
Sensors, Weather Station
Control Code

• Assign a unique sensor
• ID to each sensor
• Control HVAC system
• based on sensor data
• (Occupancy, Temp etc.)

• Coded with Java
• Contain control hierarchy
• Each layer has its own
• learning algorithm
• Optimize comfort and energy
• consumption

Price Indicator

• Send current utility price
• (Low, Med, High, Critical)

DREAM Controller
HVAC System

• Currently using a tablet PC and a PDA
• Connect a base mote to the USB port
• Control the system on GUI screen manually

• Send HVAC mode (AC, Fan, Heater, Off)
• when it changes.
• Cycle rate limit (at least greater than 4 min)

16
Learning and Control Information/Query Flow
Dynamic Pricing
Weather Forecasting
Operating
Occupancy preference Occupant behavior
Prediction accuracy
Prediction accuracy
Goal Seeking Layer
Optimize all
Q. What is the total power for each alternative
maneuver
Mode Setpoint
Current Setpoint Temperature Operating Device
A. 2500w for precooling mode, 800w for normal mode
Supervisory Layer
Cost in total power for various manuever
Query Flow
Control Information
Q. Which device uses the least power to maintain
the setpoint?
A. 500w (fan)
Setpoint
Operating Device Temperature
Coordination Layer
Appropriate device selection in a certain
condition
Q. How much power is required to maintain the
setpoint for each device?
A. 500w for fan, 2500w for AC
On/Off Setpoint
Current Process Temperature
Direct Control Layer
Tune itself, Identify the static properties of
control Objects, Identify of the transient
properties
A. 76.3F (Bedroom 1)
AC On/Off
Temperature
Q. What is the current temperature of a
controlled zone?
Mote Interface Layer
Operation of network, Optimal sample speed, Data
correlation
17
Control Strategies in Control Code
Control Code
Strategy Two Changing Setpoint
Cooling setpoint
More complicated control strategy needed.
Decision will be made based on total cost,
comfort level, occupancy, user preferences etc.
Med
High
Low Price
Strategy Three Precooling
Cooling setpoint
To maximize comfort, the setpoint which an
occupant feels most comfortable at can be used to
control the HVAC system
Changing precooling setpoint
High Price
Strategy One Single Setpoint
Strategy Four Multi-sensing Control
Temperature (F)
AC On
Cooling setpoint 72F
AC Off
M Bedroom
Living room
Bedroom 2
Time (hr)
18
DREAM Control Code

• Control of simulated, model, or real houses

Sensor/ActuatorModule
Simulation
XML
Controller
Java
Physical Model
RF
Real House
RF
Interface
19
MZEST Simulation

• Simulation will allow us to evaluate strategies
  for demand response

20
Model Calibration

• Independent datalogger-based monitoring system
  used to validate mote sensors and to provide data
  for initial simulation model calibration
• Weather station measures site microclimate
• Mote sensor data will be used to refine the
  simulation model

21
HOBO Locations
22
Measured Data
23
Simulation Results
24
Simulation Results
25
Hardware and Electronics Implementation

• Control of simulated, model, or real houses

Sensor/ActuatorModule
Simulation
XML
Controller
Java
Physical Model
RF
Real House
RF
Interface
26
Implementation of Demand-Response Enabling
Technology in a Residential Occupancy

• Integrate Devices
• Control HVAC System
• Simulate Dynamic Pricing
• Enable Data Collection

27
Integration of Devices

• Database schema all objects defined
• Every mote, sensor, and actuator has a unique ID
• Conversion information and actuation states
  defined in schema
• Extensible testbed any types of sensors and
  actuators can be added to the system

28
HVAC Control

• Replace traditional thermostat with
  remote-controlled switching relays
• HVAC mote receives system-state commands from
  controller

29
Dynamic Pricing

• Test system response to dynamic pricing
• Test residential response to price indication

Remote Price Indicator
30
Data Collection

• Increase measurement resolution through wireless
  sensors
• Environment (Indoor Outdoor)
• Single temperature data point ? multiple
  temperature zones, air and radiant temperature,
  interior humidity
• Local weather report ? on-site weather station
• Electricity
• Whole-house metering ? appliance metering
• Occupancy
• Motion sensors ? Control algorithm learning
• Also
• Internal airflow aperture of sliding doors
  windows, interior doors open/closed
• HVAC monitoring system status

31
Indoor Environment Occupancy Generic Telos
Sensor Platform
32
Indoor Environment Occupancy Generic Telos
Sensor Platform
motion detector

• Designed for rapid tool-less assembly and
  disassembly
• Universal input jacks for up to five
  analog/digital sensors
• Runs for weeks/months on AA alkaline batteries

mote
globe temperature
shielded air temperature
33
Whole House Power Sensor
34
Outdoor Environment Telos-Enabled Weather Station
35
Electricity Demand-Response AC Receptacle

• Inline outlet device for lamps and appliances
• Measure AC current, power, energy
• Power relay for appliance on/off
• TRIAC for lamp/appliance dimming

36
Hardware Development Summary

• Everything developed for the house implementation
  was made from COTS technology
• Commercialization of BSAC sensors, BWRC radios,
  and BMI power sources will allow miniaturization
  for low-cost, ubiquitous installations