Neuro-fuzzy control system for grid-connected photovoltaic systems

Patent metadata

Field	Value
Publication number	US-2016226253-A1
Application number	US-201514613302-A
Country	US
Kind code	A1
Filing date	Feb 3, 2015
Priority date	Feb 3, 2015
Publication date	Aug 4, 2016
Grant date	—

Abstract

The neuro-fuzzy control system for a grid-connected photovoltaic (PV) system includes an Adaptive Neuro-Fuzzy Inference System (ANFIS) implemented in real time. Independent active and reactive P-Q power control transfers the generated power to the grid using a voltage source inverter (VSI). The PV system includes a PV module, a buck converter, a VSI, a maximum power point tracking (MPPT) controller for the buck converter, and a VSI controller. The MPPT controller uses irradiation and temperature as inputs. A five-layer ANFIS processes these inputs and provides a control reference voltage as input to a PI controller connected to the buck converter to maintain the output voltage of the photovoltaic array with respect to the control reference voltage.

First claim

We claim:

1 . A neuro-fuzzy control system for a grid-connected photovoltaic (PV) system, comprising:
a solar power array outputting a panel voltage V PV ; a DC-DC converter operably connected to the solar power array; a proportional-integral (PI) controller operably connected to the DC-DC converter, the PI controller controlling a duty cycle of the DC-DC converter; a processor in operable communication with the solar power array, the processor including means for receiving temperature and irradiance information from the solar power array and means for operating an adaptive neuro-fuzzy network, the adaptive neuro-fuzzy network outputting a reference voltage V ref ; and means for applying a reference input to the PI controller, the reference input being characterized by the relation (V ref −V PV ); wherein the maximum power point (MPPT) of the PV array is successfully allocated under varying temperature and irradiance values of the PV array.

2 . The neuro-fuzzy control system according to claim 1 , further comprising means for training the adaptive neuro-fuzzy network, the training means having:
means for estimating parameters of the PV array, the PV array parameters being I L , the light generated current, I 0 , the diode saturation current, R S and R SH being the series and parallel resistance respectively, and a being the diode modified ideality factor; means for initializing size of the PV array, the PV array size being characterized by variables N ss , N pp , wherein N ss is the number of series-connected panels and N pp is the number of parallel-connected panels; means for modifying the estimated PV array parameters I L , I 0 , R S , R SH , and a based on the initialized PV array size defined by N ss , and N pp ; means for initializing training parameters, N MAX , being the number of training data points, T MIN , being the minimum temperature, T MAX , being the maximum temperature, S MIN being the minimum irradiation, and S MAX being the maximum irradiation; means for selecting a random temperature and irradiation operating condition; means for calculating the PV array parameter values given the selected operating condition; means for solving a PV array modeling equation, the PV array modeling equation being characterized by the relation:

I
D

=

I
L

-

I
0



{

exp


[

V
PV

+

I
PV



R
S

a

]

-
1

}

-

V
PV

+

I
PV



R
S

R
SH

,

where I PV and V PV represent the current and voltage generated from the PV panel, I L is the light generated current, I 0 is the diode saturation current, R S and R SH are the series and parallel resistance, respectively, and a is the diode modified ideality factor;
means for storing a maximum power point reference voltage V MP given the solution of the PV array modeling equation and corresponding to the reference voltage V ref at the selected operating condition; and
means for storing additional V MP values corresponding to additional randomly selected operating conditions until a stopping criterion has been met.

3 . The neuro-fuzzy control system according to claim 2 , further comprising means for displaying the training data once the stopping criterion has been met.

4 . The neuro-fuzzy control system according to claim 1 , further comprising means for converting DC power coming from the DC-DC converter to three-phase AC power, either to supply AC loads or for integration with the grid.

5 . The neuro-fuzzy control system according to claim 1 , further comprising a two-level three-phase inverter coupled to the DC-DC converter for converting DC power coming from the DC-DC converter to three-phase AC power, either to supply AC loads or for integration with the grid.

6 . The neuro-fuzzy control system according to claim 5 , further comprising a phase locked loop (PLL) connected to the two-level three-phase inverter, the PLL having outputs ω and θ to track frequency (ω=2nf) and phase angle θ of the grid, respectively, the PLL outputs being used for voltage ABC/DQ reference frame conversion and for current ABC/DQ reference frame conversion.

7 . The neuro-fuzzy control system according to claim 6 , wherein the two-level three-phase inverter further comprises a DC voltage controller maintaining a DC link voltage to its reference value.

8 . The neuro-fuzzy control system according to claim 6 , wherein the two-level three-phase inverter further comprises decoupled active and reactive current controllers for independent active (P) and reactive (Q) power control, respectively.

9 . The neuro-fuzzy control system according to claim 2 , further comprising means for optimizing the estimated PV array parameters.

10 . A computer software product, comprising a non-transitory medium readable by a processor, the non-transitory medium having stored thereon a set of instructions for training an adaptive neuro-fuzzy maximum power point (MPPT) controller for a grid-connected photovoltaic (PV) system, the set of instructions including:
(a) a first sequence of instructions which, when executed by the processor, causes said processor to estimate parameters of a PV array of the PV system, the PV array parameters being I L , the light generated current, I 0 , the diode saturation current, R S and R SH being the series and parallel resistance, respectively, and a being the diode modified ideality factor; (b) a second sequence of instructions which, when executed by the processor, causes said processor to initialize size of the PV array, the P

Assignees

• Univ King Fahd Pet & Minerals

Inventors

• Abido Mohammad Ali
• Worku Muhammed Y

Classifications

• H02J2101/25

  involving maximum power point tracking control for photovoltaic sources · CPC title

• H02J2101/24

  Photovoltaics · CPC title

• H02M3/04

  by static converters · CPC title

• H02J3/383 Primary

  Electricity · mapped topic

• H02M7/44

  by static converters · CPC title