verticale

Un approccio basato su MPC distribuito alla gestione energetica integrata negli edifici

(in lingua inglese)

Main concept of Smart-Grids
Example scenario (case study)
Proposed approach
Control
Test case
Advanteges of the proposed approach

Scarica il PDF Scarica il PDF
Aggiungi ai preferiti Aggiungi ai preferiti


Atti di convegni o presentazioni contenenti case history
HOME & BUILDING ottobre 2017 Smart Building: tecnologie per servizi innovativi

Pubblicato
da Benedetta Rampini
HOME & BUILDING 2017Segui aziendaSegui




Estratto del testo
Veronafiere 18-19 ottobre 2017 Gli atti dei convegni e pi di 8.000 contenuti su www.verticale.net Cogenerazione Termotecnica Industriale Pompe di Calore 27 ottobre Cogenerazione Termotecnica Industriale Pompe di Calore Alimentare Alimentare Petrolchimico Alimentare 28 ottobre Alimentare Petrolchimico Alimentare Alimentare Petrolchimico Visione e Tracciabilit 28 ottobre Luce Energia Domotica LED Luce Energia Domotica LED SMART BUILDING WORKSHOP S. Rastegarpour Distributed Automatic Systems Laboratory Soroush Rastegarpour, M. Ghaemi, L. Ferrarini 1 A distributed MPC approach to integrated energy management in buildings
(Un approccio basato su MPC distribuito alla gestione energetica integrata negli edifici) SMART BUILDING WORKSHOP S. Rastegarpour 2 Contents ' Introduction and problem description ' Proposed control approaches ' Concluding remarks and future works Hardware di base Architettura OS & Funz. Conclusioni Introduzione IEC61131 SMART BUILDING WORKSHOP S. Rastegarpour 3 Main goal Problem description Proposed approach Future works Main concept of Smart-Grids To deliver electricity in a controlled, smart way from points of generation to consumers Demand Side Management (DSM) Modification of consumer's purchasing patterns and behaviour For example: Load shaping SMART BUILDING WORKSHOP S. Rastegarpour 4 Load shaping ' I will exploit inertia and storages of energy to decouple load request to the electrical grid from the user comfort control problem. ' Examples of storages: building wal s (thermal energy), hot water tank (thermal energy), battery (electrical energy). ' But how much energy should I put in the storages to guarantee comfort when requested by the user, in spite of renewable future power generation and possible variable energy price' How can I coordinate all my energy resources' Possible solution: Distributed predictive approach. How can I shape the load according to grid constraints' F.e., shifting the demand away from peak hours Problem description Proposed approach Future works SMART BUILDING WORKSHOP S. Rastegarpour 5 Load shaping ' A Thermal Energy Storage (TES) ' A Heat Pump ' A building with floor radiant panels ' Renewable sources ' Electrical storage Coupling TES with air conditioning in building scan provide a spinning reserve on the demand, without sacrificing occupant thermal comfort. Example scenario (case study) HP Thermal Energy storage ''''' '''''''''''''''''''''''' P '''''''''''''''''''' '''''''''''''''''''''''''''' Problem description Proposed approach Future works SMART BUILDING WORKSHOP S. Rastegarpour 6 ' How big can be the impact of TES in system ' Possible questions ' How to model a TES in a realistic way' ' In case of presence of two storage elements, namely an electrical (Battery) and a thermal (TES) one: ' How to chose the charging/disharging process for each storage' ' What is the optimal decision' ' How can we control the energy to the TES (through the heat pump) and the building in order to assure user's comfort and economic convenience' Problem description Proposed approach Future works SMART BUILDING WORKSHOP S. Rastegarpour 7 Contents ' Problem description ' Proposed approach ' Future works Hardware di base Architettura OS & Funz. Conclusioni Introduzione IEC61131 SMART BUILDING WORKSHOP S. Rastegarpour Proposed approach 8 ' The proposed approach is based on two main pillars: Problem description Future works Proposed approach ' Modeling of the TES ' Modeling of the thermal behaviour of the building MODELING CONTROL ' Distributed control strategy ' Predictive control strategy (receding horizon control) ' A heuristic controller and a centralized MPC will be considered for comparison purposes SMART BUILDING WORKSHOP S. Rastegarpour 10 Thermal Energy Storage n(6) n(5) n(1) Heat Pump Input (u(k)) Ts(k) Tr(k) Tinlet(k) Tcw(k) Fs(k) Fr(k) Fcw(k) Finlet Water Supply direction Re tu rn w ate r di re cti on ' Valve position in water supply direction: ' = '''''''''''''''''''''''' - '''''''' '''''''' - '''''''' ' The model is nonlinear through variable circulated water flow rate Fcw k . '''' '''' + 1 = '''' '''' '''' , '''''''''''''''''''''''' '''' , '''''''' '''' + ''''. '''' '''' + '''''''' '''' . ''''('''') '''''''''''' '''' = '''''''' '''' = ' . '''''''''''''''''''''''''''''''' '''''''''''' , '''''''''''''''''''''''''''''''' ' In literature, application of TES is usually based on fixed Fcw k . Mode complex beahviour when we have a variable '''''''''''' '''' !! Modeling - TES Problem description Future works Proposed approach Model order RMSE (') 3rd order / 24th order 1.725 6th order /24th order 2.193 12th order /24th order 0.3950 15th order /24th order 0.254 21st order /24th order 0.073 Model-order selection SMART BUILDING WORKSHOP S. Rastegarpour 11 ' Valve OPEN: ' Valve CLOSED: ''''''''''''('''') >> ''''''''''''''''''''('''') ''''''''''''''''''''('''') is negligible. '''''''''''' '''' = 0 '''''''''''''''''''' '''' ' ''''*(5% '''''''' '''''''''''''''''''''''''''' '''''''''''') 1+'''''''' '''' = '''''''''''' '''''''''''''''' Tank heat transfer Convection Conduction t : depends on heat transfer coefficient 5% 10% 0% 5% nominal Fcw(k) Fcw(k) F''''''''''''''''(k) Valve ratio Modeling ' TES (Empact of valve positions) Problem description Future works Proposed approach 0 16 32 48 64 80 96 Time (Hour) 20 30 40 50 60 70 80 90 100 110 Temperature Warm-up & Cool-down process 3 order - Without convection effect
6 order - Without convection effect
12 order - Without convection effect
15 order - Without convection effect
3 order - With convection effect
6 order - With convection effect
12 order - With convection effect
15 order - With convection effect
* * The considered Convection effect formula has signifficant effect on load shifting in DSM system. SMART BUILDING WORKSHOP S. Rastegarpour 12 ' The model is based on Heating Balance Equetion for the walls, Ceiling and pavement. '''''' = ''''''''. '''' + ''''''''. '''' '''' = ''''''''. '''' + ''''''''. '''' '''''' = ''''''''. '''' + ''''''''. '''' '''' = ''''''''. '''' + ''''''''. '''' CLOSED OPEN d= 1, OPEN d= 0, CLOSED ''''''''''''''''' = '''''''''''''''''''' '''''''' - '''''''' + '''''''''''''''' '''''''''''''''' - ''''''''
''''''''''''''''''''''''''''' = '''''''''''''''''''' '''''''' - '''''''' +'''''''''''''''''''' '''''''''''' - ''''''''
'''''''''''''''''''''''''''''''' ' = '''''''''''''''' '''''''' - '''''''''''''''' + ''''1''''''''''''( 1 - ''''-'''' '''' )('''''''' - ''''''''''''''''' ''''''''''''''''''''''''' = [''''1'''''''''''' 1 - ''''-'''' '''' + ''''''''''''-''''''''(''''-'''' - 1 - ''''-'''' '''' )] '''''''''''''''' - '''''''' + ''''''''''''-'''''''' '''''''' - '''''''''''' ''''' = ''''''''' ''''''''' ''''''''''''''''' ''''''''''''' ; '''' = '''''''' '''''''' '''''''''''''''' '''''''''''' ; '''' = '''''''''''' '''''''''''''''','''' '''''''''''''''','''' '''''''''''''''','''' '''''''''''''''','''' '''''''''''''''','''''''''''''''' '''''''' '''''''''''''''' '''''''''''''''' '''''''' ; Modeling ' Radiant floor building Problem description Future works Proposed approach SMART BUILDING WORKSHOP S. Rastegarpour Control 13 ' The proposed approach is based on two main pillars: ' Modeling of the TES ' Modeling of the thermal behaviour of the building MODELING CONTROL ' Distributed control strategy ' Predictive control strategy (receding horizon control) ' A heuristic controller and a centralized MPC will be considered for comparison purposes Problem description Future works Proposed approach SMART BUILDING WORKSHOP S. Rastegarpour Control ' Studied scenario 14 MPC 1 MPC 2 Energy resources Demand (users) Fut ur e w or k Comfort level Energy consumption N -s te p pr ed ic tio n o f '''' '''' '''''''' '''''''' ('''' ) T''''(k) F''''(k) Tr(k) Fr(k) Valve Tinlet(k) Tcw(k) Fcw(k) ' Circulated water flow rate and circulated water temperature are fixed over N-step prediction of MPC.1 ' N-step prediction of '''''''''''''''''''''''' is available by MPC.2 Strong assumption ' Fixed water flow rate (valve position is fixed) HP TES ''''' '''''''''''''''''''''''' P '''''''''''''''''''' '''''''''''''''''''''''''''' MPC1 MPC2 ' Linear formulation of MPC.1 due to using fixed Circulated water flow rate ' Time invariant behavior of MPC.1 through fixed circulated water temperature ' Satisfying the comfort condition by using N-step prediction of '''''''''''''''''''''''' Consequences Problem description Future works Proposed approach ' Distributed Model Predictive Control (DMPC) SMART BUILDING WORKSHOP S. Rastegarpour 15 '''''''''''''''' ('''''''''''' - '''''''''''''''''''') = '[''''''''. ''''. '''' + ''''''''''''''''''''''''''''. ''''2. ''''''''''''''''''''''''] ''''-1 ''''=1 '''''''''''''''' (h''''''''h - '''''''''''''''''''') = '['''''''''. ''''. ''''' + ''''''''. ''''. '''' + ''''''''. ''''. '''' + ''''. ''''] ''''-1 ''''=1 Control ' Fixed Flow Rate Scenario Problem description Future works Proposed approach M PC 2 M PC 1 ' '''' is panic variable for soft constraints ' '''' is discomfort level (comfort level is evaluated by room temperature) ' '''' : Decision variables for Power-Exchange ' Optimization problem: Mixed integer quadratic ptograming ' Control Strategy: Non-Cooperative adaptive DMPC Discrete decision variable due to using energy storage ' The local controllers have different, possibly conflicting, objectives ' Non-iterative algorithms: information is transmitted (and received) once within each sampling time. SMART BUILDING WORKSHOP S. Rastegarpour 16 ' Non-Cooperative Distributed Adaptive MPC MPC 1 MPC 2 Energy resources Demand (users) Fut ur e w or k Comfort level Energy consumption N -s te p pr ed ic tio n o f '''' '''' '''''''' '''''''' ('''' ) T''''(k) F''''(k) Tr(k) Fr(k) Valve Tinlet(k) Tcw(k) Fcw(k) HP TES ''''' P '''''''''''''''''''' '''''''''''''''''''''''''''' MPC1 M PC 2 ' Fixed circulated water flow rate ' Fixed circulated water temperature Strong assumption Control ' Variable Flow Rate Scenario Consequences ' Tank temperature can vary freely; ' Tank circulated water flow rate is varying through the valve decision. Problem description Future works Proposed approach SMART BUILDING WORKSHOP S. Rastegarpour 17 ' Valve PID controller is capable to regulate requested room inlet water temperature during each sampling time. ' Tank Energy Storage is able to provide the minimum required energy for building heating system. Thermal Energy Storage Thermal energy Tcw(k) Ts(k) , Fs(k) Tr(k) Fr(k) Fcw(k) Tinlet(k) , Finlet('''') Room ''''''''''''1 '''''''''''' ''''''''''''2 Heat Pump PV- panel Grid PV- panel Estimator T''''''''''''(k) T inl et (k ) T '''''''' '''''''' (k ) Tref '''' '''''''''''''''''('''': '''' + '''') , '''''''''''''''''''''''''('''': '''' + '''') T '''' (k ) Economical control decision Control ' Variable Flow Rate Scenario Problem description Future works Proposed approach SMART BUILDING WORKSHOP S. Rastegarpour 19 Building information TES specification Heat-Pump characterization Battery specification PV panel characterization Dimention 3*8*3 (m) Capacity 700 (lit) COP: 2 High effecient Capacity 13500 (watt) Nominal production 800 (watt) Sampling time 15 (minutes) Sampling time 15 (minutes) Sampling time 15 (minutes) Accessible bound 70 % - 100 % Average production 114 (watt) Prediction Horizon 6 (hours) Prediction horizon 6 (hours) Prediction horizon 6 (hours) Exchangeable power 1000 (watt/sample) Production interval 12 (hours) 0 2.4 4.8 7.2 9.6 12 14.4 16.8 19.2 21.6 24 Time (Hour) 0 200 400 600 Power (Watt) PV-panel production X: 48 Y: 554.1 0 2.4 4.8 7.2 9.6 12 14.4 16.8 19.2 21.6 24 Time (Hour) -2 0 2 4 Temperature (C) External temperature during one day Test case ' Model information

Problem description Future works Proposed approach SMART BUILDING WORKSHOP S. Rastegarpour 20 0 2.4 4.8 7.2 9.6 12 14.4 16.8 19.2 21.6 24 Time (Hour) -600 -400 -200 0 200 400 600 800 Power (Watt) 0.12 0.14 0.16 0.18 0.2 0.22 0.24 0.26 0.28 0.3 0.32 Cost (Euro) Without load shifting phenomena Php
Pgrid
PV-production
Cost
0 2.4 4.8 7.2 9.6 12 14.4 16.8 19.2 21.6 24 Time (Hour) -600 -400 -200 0 200 400 600 800 1000 1200 Power (Watt) 0.12 0.14 0.16 0.18 0.2 0.22 0.24 0.26 0.28 0.3 0.32 Cost (Euro) Load shifting phenomenon Php
PV-production
Pgrid
Cost
Test case ' Performance analysis , Load shifting empact
Problem description Future works Proposed approach 0 8 16 24 0 8 16 24 Time (Hour) 20.5 21 21.5 22 22.5 23 23.5 Temperature(') Room temperature (Comfort analysis) Room temperature
Set-point
Ad ap tiv e di st ri bu ted M PC H eu ris tic c on tr olle r SMART BUILDING WORKSHOP S. Rastegarpour 21 Ad ap tiv e di st ri bu ted M PC H eu ris tic c on tr olle r 0 2.4 4.8 7.2 9.6 12 14.4 16.8 19.2 21.6 24 Time (Hour) -1500 -1000 -500 0 500 1000 1500 0.12 0.14 0.16 0.18 0.2 0.22 0.24 0.26 0.28 0.3 0.32 Cost (Euro) Power Exchange during one day HP-Power
Batt-Power
PV-production
Grid-Power
Cost
0 2.4 4.8 7.2 9.6 12 14.4 16.6 19.2 21.6 24 Time (Hour) -1000 -500 0 500 1000 0.12 0.14 0.16 0.18 0.2 0.22 0.24 0.26 0.28 0.3 0.32 Power exchange during one day HP-Power
Batt-Power
Grid-Power
PV-production
Cost
Test case ' Performance analysis of proposed approach
Problem description Future works Proposed approach 0 2.4 4.8 7.2 9.6 12 14.4 16.8 19.2 21.6 24 Time (Hour) 21 22 23 24 25 26 27 28 Temperature (C) Tank Temperature vs. Room inlet water temperature Room inlet water temp
Tank temp (6-node)
Tank temp (12-node)
0 2.4 4.8 7.2 9.6 12 14.4 16.8 19.2 21.6 24 Time (Hour) 21 22 23 24 25 26 27 28 29 Temperature (C) Tank temperature vs. Room inlet water temperature Room inlet water temp
Tank temp (6-node)
Tank temp (12-node)
SMART BUILDING WORKSHOP S. Rastegarpour 23 ' Successful Load shifting due to application of TES and Mixing-Valve ' Satisfactory comfort conditions ' Distributed solution for adapting the changes of configuration ' Nonlinear model of a TES for dynamic pricing application ' Modification of consumer's purchasing patterns and behaviour Advanteges of the proposed approach Problem description Future works Proposed approach With our approach we can save energy up to 46% (in terms of selling) and 22% (in terms of purchasing) in the analyzed scenarios (variable prices of electrical energy) with respect to heuristic approach, while the difference between centralized and distributed control is negligible, always preserving comfort SMART BUILDING WORKSHOP S. Rastegarpour 24 Contents ' Problem description ' Proposed approaches ' Future works Hardware di base Architettura OS & Funz. Conclusioni Introduzione IEC61131 SMART BUILDING WORKSHOP S. Rastegarpour 25 Inovation in Control part Component level Problem description Proposed approach Future works 1) Consensus-based MPC at the Load-Level MPC1 MPC2 Batt PV Room TES HP MPC3 MPC1 MPC2 MPC3 '''''''' '''''''' ''''''''''''''''''''''''' '''''''''''''''''''''''' ''''''''''''''''''''''''' '''''''''''''''''''' '''''''''''' , ''''''''''''''''''''-'''''''''''' ''''''''''''''''''''' '''''''''''''''''''' ' How to chose the charging/disharging process for each storage' ' Solution: negotiation at load level among all of the storages, to converge to the best agreement. ' Decomposing the whole plant into several homogenous parts (buildings, tank, HP&electrical equipment) makes it more flexible (eg easier to replace a part). ' Better real time pricing according to the best agreement amonge all strages. Advantages: SMART BUILDING WORKSHOP S. Rastegarpour 26 Inovation in Control part Problem description Proposed approach Future works 2) Model-plant mismatch detection in DMP application ' Comfort level is strongly affected by model mismatch. 0 2.4 4.8 7.2 9.6 12 14.4 16.8 19.2 21.6 24 Time (Hour) 21 22 23 24 25 26 27 28 Temperature (C) Model - Plant mismatch effect on comfort level comfort Level
Model - plant with mismatch
Plant
Losing comfort G ''''' ''''('''') ''''('''') ''''('''') '''''('''') - + + + ''''('''') Model-plant mismatch ' MPC decision is based on a wrong model ' Real plant is not always able to satisfy the comfort condition ' Model-plant mismatch can be detected earlier and can be considered in the MPC real-time decision. Problem: Solution: SMART BUILDING WORKSHOP S. Rastegarpour Distributed Automatic Systems Laboratory Thank you for your attention QUESTIONS '

Document Outline

Diapositiva numero 1 Contents Main goal Load shaping Load shaping Possible questions Contents Proposed approach Modeling - TES Modeling ' TES (Empact of valve positions) Modeling ' Radiant floor building Control Control ' Studied scenario Control ' Fixed Flow Rate Scenario Control ' Variable Flow Rate Scenario Control ' Variable Flow Rate Scenario Test case ' Model information Test case ' Performance analysis , Load shifting empact Test case ' Performance analysis of proposed approach Advanteges of the proposed approach Contents Inovation in Control part Inovation in Control part Diapositiva numero 24


In evidenza

ExxonMobil
Grassi Mobil™ - Formulati per fornire elevate prestazioni anche in condizioni operative estreme
2G Italia
Cogenerazione e trigenerazione dal leader tecnologico mondiale
SD Project
SPAC : Il Software per la progettazione Elettrica
select * from VERT_CONTENUTI as con left join VERT_UTENTI_AZIENDE as azi on con.iduteazienda = azi.iduteazienda where testataurl like '%%' or eventourl like '%'%' or testataurl like '%'eiomfiere.it%' or eventourl like '%'eiomfiere.it'%'
Array ( ) Array ( [0] => Array ( [0] => 42000 [SQLSTATE] => 42000 [1] => 102 [code] => 102 [2] => [Microsoft][ODBC Driver 13 for SQL Server][SQL Server]Incorrect syntax near 'eiomfiere'. [message] => [Microsoft][ODBC Driver 13 for SQL Server][SQL Server]Incorrect syntax near 'eiomfiere'. ) [1] => Array ( [0] => 01000 [SQLSTATE] => 01000 [1] => 16954 [code] => 16954 [2] => [Microsoft][ODBC Driver 13 for SQL Server][SQL Server]Executing SQL directly; no cursor. [message] => [Microsoft][ODBC Driver 13 for SQL Server][SQL Server]Executing SQL directly; no cursor. ) )
© Eiom - All rights Reserved     P.IVA 00850640186