Systems and methods for stochastically using electric vehicles as mobile energy storage
View Patent ↗Systems and methods for energy management includes receiving parameters from commercial building management system components; generating a stochastic programming model of electric vehicles (EVs) as mobile energy storage (MES) for optimal planning, operation, and control purposes; and controlling operation of EVs according to the stochastic programming model to lower operating cost and carbon emission.
1. A method for energy management, comprising:
receiving, by electronics in a microgrid, parameters from energy management system components;
generating a stochastic model for electric vehicle (EV) operation as mobile energy storage (MES) for optimal planning, operation, and control purposes; solving an optimization problem for optimal planning, operation, and control purposes; controlling operation of EVs according to the optimization model to lower operating cost and emissions;
controlling operation of EVs according to the optimization model to ensure a predetermined level of reliability; and
modeling the EVs for optimal planning, operation, and control using an MES energy balance equation:
e
mes
,
t
,
s
=
(
1
-
Φ
mes
)
e
mes
,
t
-
1
,
s
+
τ
(
p
mes
,
t
,
s
chg
η
mes
chg
-
p
mes
,
t
,
s
dch
η
mes
dch
)
+
E
mes
,
t
,
s
conn
-
E
mes
,
t
,
s
disc
where E mes,ts conn and E mes,ts disc represent stochastic energy level of EVs connected to and disconnected from at time t in scenario s, respectively and wherein energy storage levels of EVs are limited by minimum and maximum available capacities of the EVs at each time interval in each scenario, E mes,ts and Ē mes,ts respectively, as follows:
SOC mes,s Ē mes,ts ≦e mes,ts ≦ SOC mes,s Ē mes,ts
where minimum and maximum available capacities of the EVs are calculated using following equations:
Ē mes,ts =( Ē mes,t− 1,s +Ē mes,ts conn − Ē mes,ts disc )
E mes,ts =( Ē mes,t− 1,s +Ē mes,ts conn − Ē mes,ts conn ).
2. The method of claim 1 , wherein the stochastic programming model comprises a two-stage stochastic formulation and selection of first and second-stage variables.
3. The method of claim 1 , comprising generating a probabilistic model of EVs, including arrival and departure times, and charging and discharging energy and power capacities.
4. The method of claim 1 , comprising performing probabilistic modeling of contribution of EVs in spinning reserve requirements.
5. The method of claim 1 , comprising constructing discretized probability distribution functions for random quantities.
6. The method of claim 1 , comprising determining
min xε n c′x+ [Q ( x ,ξ)]
s.t. Σ j a ij x j ≦b i
where Q(x, ξ) is the optimal value of a second-stage problem:
min yε m q′y
s.t. Σ k V ks x k +Σ k w ks y s ≦h s
where ξ:=(q, h, V, W) are the data of a second-stage problem, and elements of vector ξ are random, [Q(x, ξ)] is the expected operator with respect to probability distribution of ξ,
where ξ s :=(q s , h s , V s , W s ) with respective probabilities p s :
[ Q ( x ,ξ)]=Σ s=1 S p s Q ( x,ξ s ).
7. The method of claim 1 , comprising applying charge/discharge constraints of EVs to ensure that p mes,t chg and p mes,t dch are less than maximum charging and discharging power of the EVs at each time interval:
0 ≦p mes,t,s chg ≦u mes,t,s chg P mes,t,s
0 ≦p mes,t,s dch ≦u mes,t,s dch P mes,t,s
where P mes,t,s and P mes,t,s are calculated as follows:
P mes,t,s =( P mes,t-1,s + P mes,t,s conn − P mes,t,s disc )
and wherein operational and maintenance costs of EVs includes degradation costs and an effect of charging and discharging cycles on capacity loss of the EVS, as follows:
v
ses
,
t
,
s
chg
≥
u
ses
,
t
,
s
chg
-
u
ses
,
t
,
s
-
1
chg
v
ses
,
t
,
s
dch
≥
u
ses
,
t
,
s
dch
-
u
ses
,
t
-
1
,
s
dch
C
mes
,
t
,
s
=
C
mes
dg
1
2
(
v
mes
,
t
,
s
chg
+
v
mes
,
t
,
s
dch
)
+
C
mes
c
E
_
mes
,
t
,
s
+
p
mes
,
t
,
s
dch
η
mes
dch
C
mes
,
t
,
s
s
-
p
mes
,
t
,
s
chg
η
mes
chg
C
mes
,
t
,
s
d
.
where C mes,s dg represents costs of the EVs degradation per cycle to be paid to EV owners to reimburse battery degradation due to charge and discharge, and C mes,s c denotes capacity costs to be paid to EV owners for hours connecting their vehicles in each scenario, and C mes,t,s s and C mes,t,s d represent the selling and buying energy price of the EV.
8. The method of claim 1 , comprising modeling available charging/discharging energy capacity of EVs is formulated as a random variable.
9. The method of claim 1 , comprising modeling degradation costs of EV batteries.
10. The method of claim 1 , comprising modeling of the grid connection and peak demand charges for grid connection.
11. The method of claim 1 , comprising modeling contribution of EVs is Spinning Reserve requirements as follows:
p
mes
,
t
,
s
sp
=
min
{
(
e
mes
,
t
,
s
-
SOC
_
mes
,
s
E
_
mes
,
t
,
s
)
τ
,
P
_
mes
,
t
,
s
-
p
mes
,
t
,
s
dch
}
with constraints reformulated as linear constraints in the mode as follows:
p
mes
,
t
,
s
sp
≤
(
e
mes
,
t
,
s
-
SOC
_
mes
,
s
E
_
mes
,
t
,
s
)
τ
and
P
mes
,
t
,
s
sp
≤
P
_
mes
,
t
,
s
-
p
mes
,
t
,
s
dch
.
12. The method of claim 1 , comprising modeling uncertainty in energy prices for grid connection.
13. The method of claim 1 , comprising modeling stochastic optimization techniques to model electric vehicles as mobile energy storage.
14. The method of claim 1 , comprising modeling scenario based stochastic programming approach to model electric vehicles as mobile energy storage.
15. The method of claim 1 , comprising modeling stochastic scenario based MILP modeling.
16. The method of claim 1 , comprising modeling single objective MILP model with maximization of daily profit, minimization of GHG emissions, and minimization of total costs.
17. The method of claim 1 , comprising modeling single objective or multiple objective stochastic scenario based MILP modeling.
18. The method of claim 1 , comprising modeling non-linear energy balance equation for EVs as:
e
mes
,
t
,
s
=
(
1
-
Φ
mes
)
e
mes
,
t
-
1
,
s
+
τ
(
p
mes
,
t
,
s
chg
η
mes
,
t
,
s
chg
-
p
mes
,
t
,
s
dch
η
mes
,
t
,
s
dch
)
+
E
mes
,
t
,
s
conn
-
E
mes
,
t
,
s
disc
where and η mes,t,s chg and η mes,t,s dch are functions of p mes,t,s chg and p mes,t,s dch at each time and scenario, respectively.
19. The method of claim 1 , comprising modeling of uncertain parameters in scenario based stochastic programming approach to model electric vehicles as mobile energy storage.
20. The method of claim 1 , comprising selecting first and second stage variables in a scenario based stochastic programming approach to model electric vehicles as mobile energy storage.