CN109901659A - A kind of photovoltaic off-grid inverter MPPT control system - Google Patents
A kind of photovoltaic off-grid inverter MPPT control system Download PDFInfo
- Publication number
- CN109901659A CN109901659A CN201910156522.3A CN201910156522A CN109901659A CN 109901659 A CN109901659 A CN 109901659A CN 201910156522 A CN201910156522 A CN 201910156522A CN 109901659 A CN109901659 A CN 109901659A
- Authority
- CN
- China
- Prior art keywords
- subsystem
- control
- photovoltaic
- neuron node
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E10/00—Energy generation through renewable energy sources
- Y02E10/50—Photovoltaic [PV] energy
- Y02E10/56—Power conversion systems, e.g. maximum power point trackers
Landscapes
- Inverter Devices (AREA)
Abstract
The invention discloses a kind of photovoltaic off-grid inverter MPPT control systems, including data acquisition subsystem, control subsystem and executive subsystem；The data acquisition subsystem acquisition electric current and voltage data, and it is sent to control subsystem；The control subsystem receives the data that the data acquisition subsystem is sent, and building generates confrontation network and carries out data analysis modeling, forms control instruction signal using ADRC control method after model foundation, be sent to executive subsystem；The executive subsystem tracks maximum power point according to the control instruction signal received.Control system of the invention has very strong adaptive, self-adjusting ability, can track the maximum power point of photovoltaic DC-to-AC converter well, improves off-network type photovoltaic DC-to-AC converter power output capacity.
Description
Technical field
The invention belongs to the interleaving techniques fields of power electronic technique, big data technology and control theory, and in particular to one
Kind photovoltaic off-grid inverter MPPT control system.
Background technique
In photovoltaic reversals, it usually needs detection major loop DC voltage and output electric current calculate too
The output power of positive energy array, and realize the tracking to maximum power point.Existing technology is to detect the output electricity of photovoltaic cell
Galvanic electricity pressure, then according to the thought of feedback control, Current Voltage, which is adjusted, using pid control algorithm makes solar opto-electronic board
And inverter output reaches maximum power point.But there is many deficiencies pid algorithm, in many occasions to maximum power point with
Track cannot get a desired effect, for example, PID is usually to wait until to compensate control after error occurs, if proportionality coefficient is small,
Then reaction is slow, if proportionality coefficient is excessive, it is possible to can overshoot.
Summary of the invention
It is difficult to track the defect of photovoltaic DC-to-AC converter maximum power point well for the prior art, it is an object of the invention to
A kind of photovoltaic off-grid inverter MPPT control system based on data-driven and production confrontation network method is provided.
To achieve the above object, the technical solution adopted by the present invention are as follows:
A kind of photovoltaic off-grid inverter MPPT control system, including data acquisition subsystem, control subsystem and execute son
System；The data acquisition subsystem acquisition electric current and voltage data, and it is sent to control subsystem；The control subsystem connects
The data that the data acquisition subsystem is sent are received, building generates confrontation network and carries out data analysis modeling, after model foundation
Control instruction signal is formed using ADRC control method, is sent to executive subsystem；The executive subsystem is according to receiving
Control instruction signal tracks maximum power point.
Preferably, model is the neural network model of 3-tier architecture, wherein the neuron node quantity of input layer is 2,
The neuron node quantity of output layer is 1, and the neuron node quantity of hidden layer is 10, and the calculating of each neuron node is public
Formula is as follows:
yi=f (kixi+bi) (1),
Wherein, yiRepresent the output of neuron node, xiRepresent the input of neuron node, kiRepresent neuron node
Weight, biThe threshold value of neuron node is represented, f is Sigmoid function.
Preferably, the discriminant function after model foundation using convolution function as generation confrontation network, carries out Modifying model.
It is furthermore preferred that revised model carries out expectation maximization (EM) iteration optimization.
One embodiment of the present invention the utility model has the advantages that
Analysis modeling is carried out using countering type network is generated, control instruction is formed using the feedback control structure of ADRC and believes
Number, there is very strong adaptive, self-adjusting ability, the maximum power point of photovoltaic DC-to-AC converter can be tracked well, improve off-network type
Photovoltaic DC-to-AC converter power output capacity.
Detailed description of the invention
Fig. 1 is the structural schematic diagram of one embodiment of the present invention.
Fig. 2 is the flow diagram of the amendment neural network model of one embodiment of the present invention.
Fig. 3 is second order ADRC controller structure diagram.
Specific embodiment
With reference to the accompanying drawing and specific embodiment the present invention will be described in detail.It should be understood that this place
The specific embodiment of description is merely to illustrate and explain the present invention, and is not intended to restrict the invention.
The photovoltaic off-grid inverter MPPT control system of one embodiment of the present invention, including data acquisition subsystem, control
Subsystem and executive subsystem, as shown in Figure 1.The data acquisition subsystem is mainly arranged in inverter output end, is used for
Electric current and voltage data are acquired, acquisition data are sent to control subsystem.The control subsystem receives the data acquisition
The data that system is sent, building generate confrontation network and carry out data analysis modeling, and ADRC control method is used after model foundation
Control instruction signal is formed, executive subsystem is sent to.The executive subsystem is matched with off-network inverter, the control that will be received
Duty ratio of the command signal processed as electronic device PWM in executive subsystem, controls electronic device, realizes to maximum work
The tracking of rate point.
The control subsystem includes algoritic module, carries out data analysis, modeling, output control instruction signal.This implementation
In example, selected model is the neural network model of 3-tier architecture, wherein the neuron node quantity of input layer is 2, with electric current
And voltage data is corresponding, the neuron node quantity of output layer is 1, and hidden layer neuron number of nodes does not limit especially
It is fixed.
The calculation formula of each neuron node is as follows:
yi=f (kixi+bi) (1),
Wherein, yiRepresent the output of neuron node, xiRepresent the input of neuron node, kiRepresent neuron node
Weight, biThe threshold value of neuron node is represented, f is Sigmoid function.
Number of nodes in view of input and output and the approximation accuracy to sample function, hidden layer neuron number of nodes are with 10
Preferably.
After model foundation, it is preferred to use convolution function carries out loop iteration, repair as the discriminant function for generating confrontation network
The weight of positive neural network model, threshold value, as shown in Figure 2.In figure, data sample is the input data of each neuron node, raw
Weight, the threshold value that each neuron node is preset at confrontation model, are differentiated using convolution function, if differentiating result symbol
Convolution function requirement is closed, shows that the weight of each neuron node, threshold value have been corrected, then output model；If differentiating that result is not inconsistent
It closes and requires, then carry out loop iteration, correct confrontation model, until meeting the requirement of convolution function, the mind that is corrected
Through network model.
After Modifying model, EM iteration optimization can be carried out, step can be with are as follows:
(a) two inputs and an output that Xi represents revised neural network model, building Xi, input and output connection are set
Close probability density distribution ziWith the three-number set set { X of distribution parameter θi；zi；θ}.
(b) z is gone out according to the initial value of parameter or previous obtained model parameter calculationiPosterior probability:
Qi(zi)=p (zi|Xi, θ) and (2),
In formula, p is the probability density function for obeying t distribution.
(c) likelihood function of formula (2) probability density function is maximized, and carries out following iteration:
The adjacent acquired results error of iteration twice is controlled no more than 1 ‰, obtains parameter value θ.
(d) model analyzing: being non-analytical model through the model that above step is obtained, needs to carry out parsing.This reality
The analytic method for applying example use is to carry out time series modeling, obtains the analytic modell analytical model shaped like formula (4):
In formula, k refers to the delay step of model；N1, N2, N3 refer to the order of each linear, non-linear phase；h0Refer to model
Constant term；A refers to each ordered coefficients of output delay；B refers to each ordered coefficients of input delay；U ' is input signal；C refers to defeated
Enter the non-linear order coefficient of signal.
Model after parsing can be simplified, and then be controlled using second order ADRC as shown in Figure 3, formed control and referred to
Signal is enabled, executive subsystem is sent to, maximum power point is tracked by executive subsystem.In Fig. 3, y is the output of control object,
U is control amount, v0For input, z1, z2For the output of Nonlinear Tracking Differentiator, w is disturbance, the state variable z3 of extended state observer
Can tracking object state variable well, b0For the rough estimate value of the control input amplification coefficient of object.To make second order ADRC
General Third-order object can be preferably controlled, all parameter constants of expansion observer (ESO) can be first kept；Appropriate adjustment
Parameter in NLSEF.Then, the above-mentioned parameter adjusted is fixed, is realized by adjusting gain in ESO to different
The control of Third-order object.
The preferred embodiment of the present invention has been described above in detail, still, during present invention is not limited to the embodiments described above
Detail within the scope of the technical concept of the present invention can be with various simple variants of the technical solution of the present invention are made, this
A little simple variants all belong to the scope of protection of the present invention.
Claims (4)
1. a kind of photovoltaic off-grid inverter MPPT control system, which is characterized in that including data acquisition subsystem, control subsystem
And executive subsystem；
The data acquisition subsystem acquisition electric current and voltage data, and it is sent to control subsystem；The control subsystem connects
The data that the data acquisition subsystem is sent are received, building generates confrontation network and carries out data analysis modeling, after model foundation
Control instruction signal is formed using ADRC control method, is sent to executive subsystem；The executive subsystem is according to receiving
Control instruction signal tracks maximum power point.
2. photovoltaic off-grid inverter MPPT control system according to claim 1, which is characterized in that model is 3-tier architecture
Neural network model, wherein the neuron node quantity of input layer is 2, and the neuron node quantity of output layer is 1,
The neuron node quantity of hidden layer is 10, and the calculation formula of each neuron node is as follows:
yi=f (kixi+bi) (1),
Wherein, yiRepresent the output of neuron node, xiRepresent the input of neuron node, kiThe weight of neuron node is represented,
biThe threshold value of neuron node is represented, f is Sigmoid function.
3. photovoltaic off-grid inverter MPPT control system according to claim 1 or 2, which is characterized in that after model foundation
Using convolution function as the discriminant function for generating confrontation network, Modifying model is carried out.
4. photovoltaic off-grid inverter MPPT control system according to claim 3, which is characterized in that revised model into
Row EM iteration optimization.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910156522.3A CN109901659B (en) | 2019-03-01 | 2019-03-01 | Photovoltaic off-grid inverter MPPT control system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910156522.3A CN109901659B (en) | 2019-03-01 | 2019-03-01 | Photovoltaic off-grid inverter MPPT control system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109901659A true CN109901659A (en) | 2019-06-18 |
| CN109901659B CN109901659B (en) | 2021-09-07 |
Family
ID=66945994
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910156522.3A Active CN109901659B (en) | 2019-03-01 | 2019-03-01 | Photovoltaic off-grid inverter MPPT control system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109901659B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113708405A (en) * | 2021-08-24 | 2021-11-26 | 深圳硕日新能源科技有限公司 | Energy storage off-grid inverter, control method thereof and energy storage off-grid inverter system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104113133A (en) * | 2014-07-01 | 2014-10-22 | 广东易事特电源股份有限公司 | Intelligent photovoltaic off-grid inverter system and power utilization control method thereof |
| CN106292828A (en) * | 2016-08-30 | 2017-01-04 | 天津理工大学 | A kind of photovoltaic system maximum power point tracking device and control method |
| CN106532744A (en) * | 2016-12-01 | 2017-03-22 | 华北电力大学（保定） | Photovoltaic power station low-frequency oscillation suppression method based on active disturbance rejection control |
| CN109167386A (en) * | 2018-09-07 | 2019-01-08 | 中北大学 | Core base fights network photovoltaic off-grid inverter loop current suppression system |
-
2019
- 2019-03-01 CN CN201910156522.3A patent/CN109901659B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104113133A (en) * | 2014-07-01 | 2014-10-22 | 广东易事特电源股份有限公司 | Intelligent photovoltaic off-grid inverter system and power utilization control method thereof |
| CN106292828A (en) * | 2016-08-30 | 2017-01-04 | 天津理工大学 | A kind of photovoltaic system maximum power point tracking device and control method |
| CN106532744A (en) * | 2016-12-01 | 2017-03-22 | 华北电力大学（保定） | Photovoltaic power station low-frequency oscillation suppression method based on active disturbance rejection control |
| CN109167386A (en) * | 2018-09-07 | 2019-01-08 | 中北大学 | Core base fights network photovoltaic off-grid inverter loop current suppression system |
Non-Patent Citations (1)
| Title |
|---|
| 乔淑娟: "风力发电机组控制系统故障诊断研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113708405A (en) * | 2021-08-24 | 2021-11-26 | 深圳硕日新能源科技有限公司 | Energy storage off-grid inverter, control method thereof and energy storage off-grid inverter system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109901659B (en) | 2021-09-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Al-Dhaifallah et al. | Optimal parameter design of fractional order control based INC-MPPT for PV system | |
| Ali et al. | Variable step size perturb and observe MPPT controller by applying θ-modified krill herd algorithm-sliding mode controller under partially shaded conditions | |
| Jouda et al. | Optimization of scaling factors of fuzzy–MPPT controller for stand-alone photovoltaic system by particle swarm optimization | |
| Kler et al. | A nonlinear PID controller based novel maximum power point tracker for PV systems | |
| Saidi et al. | A novel approach in stand-alone photovoltaic system using MPPT controllers & NNE | |
| Doubabi et al. | High Performance MPPT based on TS Fuzzy–integral backstepping control for PV system under rapid varying irradiance—Experimental validation | |
| Ounnas et al. | An efficient maximum power point tracking controller for photovoltaic systems using Takagi–Sugeno fuzzy models | |
| Aslam et al. | Observer–based control for a new stochastic maximum power point tracking for photovoltaic systems with networked control system | |
| Idrissi et al. | Backstepping Integral Sliding Mode Control Method for Maximum Power Point Tracking for Optimization of PV System Operation Based on High-Gain Observer. | |
| Zhang et al. | A new MPPT algorithm based on ANN in solar PV systems | |
| Osman et al. | Highly efficient MPP tracker based on adaptive neuro-fuzzy inference system for stand-alone photovoltaic generator system | |
| Noman et al. | Scaled conjugate gradient artificial neural network-based ripple current correlation MPPT algorithms for PV system | |
| Alaas et al. | Analysis and enhancement of MPPT technique to increase accuracy and speed in photovoltaic systems under different conditions | |
| CN109901659A (en) | A kind of photovoltaic off-grid inverter MPPT control system | |
| Arora et al. | AI based MPPT methods for grid connected PV systems under non linear changing solar irradiation | |
| CN109960307A (en) | A kind of photovoltaic off-grid inverter MPPT Auto-disturbance-rejection Control | |
| Hong et al. | A model-free control strategy for battery energy storage with an application to power accommodation | |
| Ghazi et al. | Dandelion optimizer-based reinforcement learning techniques for MPPT of grid-connected photovoltaic systems | |
| Shinde et al. | Design and Simulation of Self-tuning PID controller with MPPT for Solar Array using MATLAB/Simulink | |
| Sarhan et al. | Performance study of neural network and ANFIS based MPPT methods for grid connected PV system | |
| Dib et al. | Contribution to the performance of GPV systems by an efficient MPPT control | |
| Belova et al. | Maximum power point tracking methods for the solar battery | |
| Sefriti et al. | Neural network Incremental conductance MPPT algorithm for photovoltaic water pumping system | |
| Yadav et al. | Deep Q reinforcement learning to improve the MPPT in Solar cell | |
| Ameziane et al. | Neural network sliding mode control for a photovoltaic pumping system. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |