Model prediction-based control method for grid forming of multi-port autonomous reconfigurable solar plants
A model prediction-based control method for grid forming of multi-port autonomous reconfigurable solar plants includes: estimating and predicting system parameters by establishing a multi-port autonomous reconfigurable solar plant and a dynamic model of a synchronous generator model; converting an objective function into an unconstrained optimization problem, using Newton's method to achieve minimum computational burden within each calculation time step to find a solution in real time for obtaining an optimal angular frequency; based on results of the optimal angular frequency, updating an output voltage and a dq-axis current of the multi-port autonomous reconfigurable solar plant, and changing an arm modulation index of the plant, thereby realizing the plant's inertia and primary frequency modulation support. The model prediction-based control method provided by the present invention achieves rapid prediction during operation and improves frequency response, rapidity and system stability.
1 . A model prediction-based control method for grid forming of multi-port autonomous reconfigurable solar plant (MARS), comprising:
estimating and predicting system parameters by establishing the MARS and a dynamic model of a synchronous generator model;
converting an objective function into an unconstrained optimization problem, using Newton's method to achieve minimum computational burden within each calculation time step to find a solution in real time for obtaining an optimal angular frequency;
updating, based on results of the optimal angular frequency, an output voltage and a dq-axis current of the MARS, and changing an arm modulation index of the plant, thereby realizing the plant's inertia and primary frequency modulation support;
wherein the estimating and predicting the system parameters comprises measuring voltage and current data at a first time step [k] of an interconnection point between a MARS system and a transmission grid;
according to a dynamic continuous time model of the MARS system, estimating the voltage and current data at a subsequent time step [k+1];
setting an optimal angular frequency (ω sg ) at the subsequent time step [k+1] to the value of the previous step, and, according to a synchronous generator frequency dynamic model and the voltage and current data at the subsequent time step [k+1], estimating a rotor angle (θ sg ) to obtain an electromechanical torque at a next successive time step [k+2];
establishing quadratic term deviation of the [k+2] electromechanical torque relative to a mechanical torque and an optimization objective function of deviation of a [k+1] angular frequency relative to a reference angular frequency.
2 . The model prediction-based control method for grid forming of multi-port autonomous reconfigurable solar plant according to claim 1 , wherein the MARS system is a three-phase system with each phase composed of two arms, and each arm is composed of a normal module, a photovoltaic module, an energy storage module, an arm inductor (L o ), and an arm resistance (R o ), in which each arm comprises at least one normal module, one photovoltaic module, and one energy storage module, respectively.
3 . The model prediction-based control method for grid forming of multi-port autonomous reconfigurable solar plant according to claim 2 , wherein the dynamic continuous time model is expressed as,
{
(
L
s
+
L
o
2
)
d
i
j
d
t
+
(
R
s
+
R
o
2
)
i
j
=
e
j
-
v
j
e
i
=
-
v
p
,
j
+
v
n
,
j
2
T
e
[
k
+
2
]
=
M
f
i
f
[
k
]
(
i
a
[
k
+
2
]
cos
(
θ
s
g
[
k
+
2
]
)
+
i
b
[
k
+
2
]
cos
(
θ
s
g
[
k
+
2
]
-
2
π
/
3
)
+
i
c
[
k
+
2
]
cos
(
θ
s
g
[
k
+
2
]
-
4
π
/
3
)
)
e
a
[
k
+
1
]
=
ω
s
g
[
k
+
1
]
M
f
i
f
[
k
]
cos
(
θ
s
g
[
k
+
1
]
)
e
b
[
k
+
1
]
=
ω
s
g
[
k
+
1
]
M
f
i
f
[
k
]
cos
(
θ
s
g
[
k
+
1
]
-
2
π
/
3
)
e
c
[
k
+
1
]
=
ω
s
g
[
k
+
1
]
M
f
i
f
[
k
]
cos
(
θ
s
g
[
k
+
1
]
-
4
π
/
3
)
Q
[
k
]
=
-
ω
s
g
[
k
]
M
f
i
f
[
k
]
(
i
a
[
k
]
sin
(
θ
s
g
[
k
]
)
+
i
b
[
k
]
sin
(
θ
s
g
[
k
]
-
2
π
/
3
)
+
i
c
[
k
]
sin
(
θ
s
g
[
k
]
-
4
π
/
3
)
)
θ
s
g
[
k
+
2
]
=
θ
s
g
[
k
+
1
]
+
h
ω
sg
[
k
+
1
]
ω
s
g
[
k
+
2
]
=
ω
s
g
[
k
+
1
]
(
1
-
h
D
p
J
)
+
h
J
(
T
m
[
k
+
1
]
-
T
e
[
k
+
1
]
-
D
p
(
ω
n
[
k
+
1
]
)
)
where, i j and v j are grid current and voltage of the j-th phase (j∈a, b, c) respectively, e j represents an output voltage of the j-th phase of the MARS system, L o and R o are arm inductance and arm resistance, respectively, L s , R s are grid-side inductance and grid-side resistance, respectively, M fif is an excitation current of the synchronous generator, T e represents an electromechanical torque, θ sg represents a rotor angle, Q represents a reactive power, ω sg represents an optimal angular frequency, h represents an integration step, and k represents current time, D p represents a damping coefficient, J represents a moment of inertia, and T m represents a mechanical torque.
4 . The model prediction-based control method for grid forming of multi-port autonomous reconfigurable solar plant according to claim 3 , the objective function is expressed as,
J
(
x
)
=
J
1
(
x
)
+
J
2
(
x
)
x
=
(
ω
s
g
[
k
+
1
]
)
J
1
(
x
)
=
λ
1
(
T
e
[
k
+
2
]
-
T
m
[
k
+
2
]
)
2
J
2
(
x
)
=
λ
2
(
ω
n
[
k
+
1
]
-
x
)
2
where, λ 1 and λ 2 represent weights of cost functions, and On represents a reference angular frequency.
5 . The model prediction-based control method for grid forming of multi-port autonomous reconfigurable solar plant according to claim 4 , wherein a result of the optimal angular frequency is expressed as,
min
x
J
(
x
)
s
.
t
.
T
e
[
k
+
2
]
=
M
f
i
f
[
k
]
(
i
a
[
k
+
2
]
cos
(
θ
s
g
[
k
+
2
]
)
+
i
b
[
k
+
2
]
cos
(
θ
s
g
[
k
+
2
]
-
2
π
/
3
)
+
i
c
[
k
+
2
]
cos
(
θ
s
g
[
k
+
2
]
-
4
π
/
3
)
)
e
a
[
k
+
1
]
=
ω
s
g
[
k
+
1
]
M
f
i
f
[
k
]
cos
(
θ
s
g
[
k
+
1
]
)
e
b
[
k
+
1
]
=
ω
s
g
[
k
+
1
]
M
f
i
f
[
k
]
cos
(
θ
s
g
[
k
+
1
]
-
2
π
/
3
)
e
c
[
k
+
1
]
=
ω
sg
[
k
+
1
]
M
f
i
f
[
k
]
cos
(
θ
s
g
[
k
+
1
]
-
4
π
/
3
)
θ
s
g
[
k
+
2
]
=
θ
s
g
[
k
+
1
]
+
h
ω
s
g
[
k
+
1
]
T
m
[
k
+
2
]
=
P
a
c
,
r
e
f
[
k
+
2
]
ω
n
where, P ac,ref represents an AC side power scheduling command.
6 . The model prediction-based control method for grid forming of multi-port autonomous reconfigurable solar plant according to claim 5 , wherein the result for the optimal angular frequency comprises substituting equation terms into J(x), and the objective function J(x) is transformed into an unconstrained optimization problem, wherein the transformed objective function is minimized using Newton's method, and x at each time step is given by calculating (n+1) iterations by using a recursive formula;
wherein the recursive formula is expressed as,
x
(
n
+
1
)
=
x
(
n
)
-
F
(
x
(
n
)
)
-
1
g
(
x
(
n
)
)
g
(
x
(
n
)
)
=
[
∂
J
(
x
(
n
)
)
∂
x
]
F
(
x
(
n
)
)
=
[
∂
g
(
x
(
n
)
)
-
∂
x
]
where, x (n+1) is the optimal angular frequency ω sg after the (n+1)-th iteration, g(x (n) ) is derivative of the objective function J(x), and F (x (n) ) is derivative of function g(x (n) );
where the realizing the plant's inertia and primary frequency modulation support comprises updating the output voltage of the MARS system and the dq-axis current according to the results of the optimal [k+1] angular frequency, changing the arm modulation index of the system, and realizing the plant's inertia and primary frequency modulation support.
7 . A system applying a model prediction-based control method for grid forming of multi-port autonomous reconfigurable solar plant (MARS) according to claim 1 , comprising:
a power grid data acquisition module, a model prediction control module, an optimization-to-objective function module, and an unconstrained optimization solver module, and a control index updated module;
wherein the power grid data acquisition module is used to collect voltage and current data information of a power grid in real time;
wherein the model prediction control module uses the data obtained from the power grid data acquisition module to perform prediction through a dynamic model of a synchronous generator, so as to calculate the voltage, current and rotor angle at the next time;
wherein the optimization-to-objective function module is used to establish the optimization objective function for system control, comprising a difference between motor torque and mechanical torque and deviation of angular frequency relative to a reference angular frequency;
wherein the unconstrained optimization solver module uses a Newton method optimization algorithm, according to the optimization objective function, to convert the optimization problem into an unconstrained optimization problem, and solves the optimal angular frequency;
wherein the control index updated module updates the output voltage of the MARS system and the dq-axis current, according to the optimal angular frequency obtained by the unconstrained optimization solver module, and changes the arm modulation index of the system, and realizes the plant's inertia and primary frequency modulation support.
8 . A computer device comprising a memory and a processor, wherein the memory stores a computer program, and, when the processor executes the computer program, the method of claim 1 is implemented.
9 . A non-transitory computer-readable storage medium has a computer program stored thereon, wherein, when the computer program is executed by a processor, the method of claim 1 is implemented.