close

Se connecter

Se connecter avec OpenID

ARX Model

IntégréTéléchargement
ADREAM: Energy Consumption Optimization
through Dynamic Energetic Simulations for an
Intelligent Management of Energy
Ilias PAPAS
Carlos GARCIA RODRIGUEZ
Bruno ESTIBALS
Christelle ECREPONT
Corinne ALONSO
Summary






Introduction
The ADREAM Project
Dynamic Thermal Simulation
Physical Model
ARX Model
Conclusion
Laboratoire d’analyse et d’architecture des systèmes du CNRS
2
Smart Grids and Intelligent Energy Management
Laboratoire d’analyse et d’architecture des systèmes du CNRS
3
ADREAM
Embedded Reconfigurable Dynamic Autonomous and Mobile Architectures
Laboratoire d’analyse et d’architecture des systèmes du CNRS
4
ADREAM
Embedded Reconfigurable Dynamic Autonomous and Mobile Architectures
 A multidisciplinary scientific project of LAAS-CNRS
• Research axes
-
Autonomous data processing systems,
Sensors, robots and mobility
Connected objects,
Energetic systems.
 Large-scale experimental platform
• Surface : 1 700 m²
• Experimental area: 500 m²
• Offices : 700 m²
Laboratoire d’analyse et d’architecture des systèmes du CNRS
5
5
Energy Systems
Laboratoire d’analyse et d’architecture des systèmes du CNRS
6
Lighting System Management
 Lighting system
• Intelligent management of lighting
• Automatic adjustment of light unit power according
to natural light
Office luminosity
Threshold of 400lux
% Light Unit Power
Laboratoire d’analyse et d’architecture des systèmes du CNRS
7
Supervision of electrical production and consumption
PV Production
HVAC Consumption
Total Consumption
Total Consumption –
PV Consumption
Laboratoire d’analyse et d’architecture des systèmes du CNRS
8
Objectives
 Energy Optimization of the ADREAM building
ADREAM
PRODUCTION
CONSUMPTIO
N
• PV Panels
• Geothermal
Energy
• HVAC
• Electronic Equipment
• Lighting
How can we optimize energy consumption with a focus on HVAC systems?
 Development of precise models
•
•
•
•
Identification of input and output of every model
Model calibration according to sensor data
Prediction of electrical consumption
Improvement on systems regulation
Laboratoire d’analyse et d’architecture des systèmes du CNRS
9
DTS Principles
Dynamic Thermal Simulation (DTS)
Output
Input
Materials
Pléiades + Comfie
Energy Needs
Systems
Simulation
Consumptions
Use
Weather Data
Laboratoire d’analyse et d’architecture des systèmes du CNRS
Temperatures
10
Thermal Modeling
3D Model of ADREAM (2)
3D Model of ADREAM (3)
Laboratoire d’analyse et d’architecture des systèmes du CNRS
11
Thermal Model Calibration
Calibration between real and simulated
temperature for a thermal zone (RRMSE =
Laboratoire d’analyse et d’architecture
des systèmes du CNRS
2,84%)
12
Thermal Model Calibration - Optimization
Consumpti
on Source
Heat Pump
1
Heat Pump
2
Heat Pump
3
Air
Handling
Unit
Distributio
n Pumps
Electronic
Equipment
Lighting
TOTAL
Real
Consumption
s [MWh/an]
Simulated
Consumptio
ns
[MWh/an]
Relative
Error [%]
36,49
3,05
9,03
16,71
35,41
9,67
17,79
17,55
1,35
69,64
63,63
8,63
65,77
63,43
3,56
12,03
200,64
11,89
192,88
1,16
3,82
Laboratoire d’analyse et d’architecture des systèmes du CNRS
Expected Energy Gain of
20% by replacing the
distribution pumps
13
Physical Model
 Software
• MatLab et Simulink
 Objective
• Heat Pumps Modeling
• Optimization of regulation
 Method
• Application of physical equations for every element
Laboratoire d’analyse et d’architecture des systèmes du CNRS
14
Physical Model
Modeling of all Heat Pump elements with Simulink
Laboratoire d’analyse et d’architecture des systèmes du CNRS
15
Physical Model – Mass flow through the Compressor
Mass flow rate of the refrigerant through the compressor
Laboratoire d’analyse et d’architecture des systèmes du CNRS
16
“Black Box” Method
 Principles of the “Black Box” method
• Information on internal functioning : Limited
• Input and Output : Observable and Associable
 Black Box model in MatLab
• Transfer functions
• State-space
• ARX (Auto Regressive model with eXternal inputs)
Laboratoire d’analyse et d’architecture des systèmes du CNRS
17
ARX Model
 ARX Model?
•
Construction of a linear differential equation which associates the current outputs [y(t)]
with a finite number of past outputs [y(t-k)] and inputs[u(t-k).
 Objective
•
•
Heat Pump Modeling
Evaluation of control strategies
 Entrées
•
•
•
Geothermal Water Temperature
Return Temperature
Compressor Power
 Sorties
•
•
Hot Water Temperature
Cold Water Temperature
Laboratoire d’analyse et d’architecture des systèmes du CNRS
18
ARX Model – Water Temperature Calibration
Water Temperature Prediction (RRMSE = 3,29% , NMBE = 0,82%)
Laboratoire d’analyse et d’architecture des systèmes du CNRS
19
ARX Model – Control Simulation
Simulation de la PAC avec Simulink et contrôle
Laboratoire d’analyse et d’architecture des systèmes du CNRS
20
Conclusion and Perspectives
 Dynamic Thermal Simulation
• Precise tool for preliminary energy consumption analysis
 Physical Model
• Possibility to analyse all parameters
• Precise modeling of systems elements
 ARX Model
• Modeling of complex interactions
• Fault detection
• High versatility
 Future Work
• Modeling of all HVAC systems
• Coupling of Consumption and Production Models
Laboratoire d’analyse et d’architecture des systèmes du CNRS
21
Conclusion and Perspectives
 What does this work mean for Smart Cities?
• Optimization of energy flow  Optimization of individual
entities
• Need for prototype models of Consumption and
Production
• Fast and accurate data processing
• Intelligent management of resources by adaptive
regulation
Laboratoire d’analyse et d’architecture des systèmes du CNRS
22
Thank you for your attention!
Laboratoire d’analyse et d’architecture des systèmes du CNRS
23
Annexe
Laboratoire d’analyse et d’architecture des systèmes du CNRS
24
ARX Model – Water Temperature Calibration
Calibration of the ARX Model for Cold Water production by the Heat Pump
Laboratoire d’analyse et d’architecture des systèmes du CNRS
25
Auteur
Документ
Catégorie
Без категории
Affichages
0
Taille du fichier
11 726 Кб
Étiquettes
1/--Pages
signaler