CN109066707A - One kind being based on NARMA-L2 model energy management method for micro-grid - Google Patents
One kind being based on NARMA-L2 model energy management method for micro-grid Download PDFInfo
- Publication number
- CN109066707A CN109066707A CN201811054866.5A CN201811054866A CN109066707A CN 109066707 A CN109066707 A CN 109066707A CN 201811054866 A CN201811054866 A CN 201811054866A CN 109066707 A CN109066707 A CN 109066707A
- Authority
- CN
- China
- Prior art keywords
- node
- indicate
- generation resource
- distributed generation
- distributed
- 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
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/12—Circuit arrangements for ac mains or ac distribution networks for adjusting voltage in ac networks by changing a characteristic of the network load
- H02J3/16—Circuit arrangements for ac mains or ac distribution networks for adjusting voltage in ac networks by changing a characteristic of the network load by adjustment of reactive power
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
- H02J3/381—Dispersed generators
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
-
- 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
- Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
- Y02E40/30—Reactive power compensation
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses one kind to be based on NARMA-L2 model energy management method for micro-grid, comprising the following steps: step 10) exports and control the relational expression between input using NARMA-L2 model foundation distributed generation resource；Step 20 is based on pinning control theory, is divided into target with reactive power, generates the input of control signal and voltage output data collection for distributed generation resource in system；The data set training step 10 that step 30) is obtained with step 20) in expression formula for characterizing the neural network of complex nonlinear function, fitting reactive power divides equally distributed generation resource nonlinear object dynamic characteristic under situation；Step 40) generates reference value using gradient descent search algorithm, completes micro-capacitance sensor reactive power Collaborative Control.For this method using nonlinear function complicated in neural network fitting input-output characteristic, that realizes micro-capacitance sensor removes modelling control；It is based on pinning control theory simultaneously, realizes the Collaborative Control of micro-capacitance sensor, promotes control response speed and accuracy.
Description
Technical field
The present invention relates to the energy management technical fields of micro-capacitance sensor, especially a kind of to be based on NARMA-L2 model micro-capacitance sensor energy
Quantity management method.
Background technique
In recent years, China's economic development is rapid, and electricity consumption total amount increases rapidly, and increasingly prominent energy demand out increases and can not
The renewable sources of energy are short, renewable energy utilization rate is beneath, the contradiction between sustainable development.Therefore, renewable energy benefit is improved
With rate, tap a new source of energy using technology, reinforce renewable energy utilization it is imperative.
Distributed power generation is increasingly mature, and the new energy power supply gradually increased is that power system stability operation brings new choose
War.Most important problem is: new energy is often " uncontrollable ", " not easily-controllable ", has large effect to electrical stability.By
The requirement of large-scale distributed plant-grid connection, the concept quilt of micro-capacitance sensor are not adapted in custom power distribution systems structure and operation reserve
It introduces as the reliable solution for utilizing new energy.Micro-capacitance sensor is relative to custom power distribution systems: small scale, topological structure are easy
Become, can be run independently of bulk power grid.It is economical, to stablize, micro-capacitance sensor control is rapidly completed be the mesh that micro-capacitance sensor manager wishes to reach
Mark.
The research for distributed micro-capacitance sensor Collaborative Control has been used to be modeled as micro-capacitance sensor using pinning control theory
Voltage Collaborative Control is transformed into complex network collaboration and neighbours' error tracing problem, ensure that system voltage by complex network
The stability of convergence and operation.But distributed generation resource dynamic mathematical models are used, have ignored high-frequency inverter mathematical model
Problem difficult to model may bring control error；Restraining theory simultaneously is to be passed to reference value using to a small number of nodes, in neighbours
Constantly transmitting forms containing between node and then whole system is harmonious, and stabilized speed is relatively slow.Traditional containing theory it is micro-
Power grid Collaborative Control be based on distributed generation resource dynamic mathematical models, there is a problem of modeling inaccuracy, and pinning control cause be
Convergence rate of uniting is slower.
Summary of the invention
It is provided the technical problem to be solved by the present invention is to overcome the deficiencies in the prior art a kind of based on NARMA-L2 mould
Type energy management method for micro-grid, the present invention are taken to solve the problems, such as distributed generation resource modeling error using NARMA-L2 model
For power supply dynamic mathematical models, and it is based on containing theory, generates data set and the artificial neural network in NARMA-L2 model is instructed
Practice, generates suitable non-linear expressions；For model, suitable reference is given using gradient descent search algorithm
Value, realization fast and accurately remove modelling control, reach load or burden without work and divide equally, the control target that voltage restores rapidly.
The present invention uses following technical scheme to solve above-mentioned technical problem:
The one kind proposed according to the present invention is based on NARMA-L2 model energy management method for micro-grid, comprising the following steps:
Step 10 exports the relational expression between control signal input using NARMA-L2 model foundation distributed generation resource；
Step 20 is based on pinning control theory, is divided into target with idle, obtains the input of control signal and voltage output
Data set；
For characterizing complex nonlinear in the relational expression that step 30, the data set training step 10 obtained with step 20 are established
The neural network of function, fitting reactive power divides equally distributed generation resource nonlinear object dynamic characteristic under situation, to be instructed
Practice the relational expression between the distributed generation resource output completed and control signal input；The distributed generation resource output completed according to training
Relational expression between control signal input constructs the distributed power controller based on NARMA-L2；
Step 40 provides target output using gradient descent search algorithm for the distributed power controller that step 30 constructs
Voltage reference value completes micro-capacitance sensor reactive power Collaborative Control.
Scheme is advanced optimized based on NARMA-L2 model energy management method for micro-grid as one kind of the present invention,
In step 10, distributed generation resource dynamic mathematical models are replaced using NARMA-L2 model, by artificial neural network to distribution
Complex nonlinear function part carries out approximate substitution inside power-supply system, between distributed generation resource output and control signal input
Relational expression is as follows:
Y (k+d)=f [y (k), y (k-1) ..., y (k-n+1), u (k), u (k-1), u (k-z+1)]+g [y (k), y (k-
1),…,y(k-n+1),u(k),u(k-1),u(k-z+1)]·u(k) (1)
In formula, u (k) is represented in the control signal input of kth moment distributed generation resource, and y (k) represents kth moment distributed electrical
The output voltage signal in source, y (k+d) represent the output voltage signal of kth+d moment distributed generation resource, and f [*] and g [*] are indicated
The complex nonlinear function of distributed generation system internal nonlinearity part；It is defeated that the distributed generation resource at k-n+1 moment is acquired altogether
Out, y (k-n+1) indicates the distributed generation resource output at (k-n+1) moment；The control signal input at k-z+1 moment is acquired altogether,
U (k-z+1) is that the control signal at (k-z+1) moment inputs.
Scheme is advanced optimized based on NARMA-L2 model energy management method for micro-grid as one kind of the present invention,
The detailed process of step 20 are as follows:
Step 201, for distributed AC servo system micro-capacitance sensor topology, collaboration targeted transformation is become into the tracking of neighbor information error:
In formula, eiIndicate the error between i-th of distributed generation resource node output voltage and reference value, yiIndicate i-th point
Cloth power supply node output voltage, j ∈ NiIndicate that j-th of node is the neighbor node of i-th of node, NiIndicate i-th of distribution
The set that all neighbor nodes of power supply node are constituted, aijIndicate adjacency matrix in distributed power supply systemIn i-th
The value of row, jth column, giIndicate the containing gain of i-th of distributed electrical source node, yjIndicate have in communication topology with i-th of node
The output voltage values of j-th of node of information interchange, y0Indicate the output voltage values of containing node, i.e. reference value；
Step 202 is to guarantee error ei0 is converged to, under conditions of controlling target is that load or burden without work is divided equally, by following public
Formula selection control signal inputs ui:
In formula, Q0It indicates to refer to load or burden without work, QiIndicate that i-th of node exports idle, ωiIndicate i-th of node whether
Through being restrained,Represent the degree of coupling between node；
Step 203, according to the sagging control characteristic of distributed generation resource, i.e., output voltage and it is idle between relational expression, by formula
(3) idle in replaces with output voltage, and control signal input expression formula becomes:
Wherein, nQiIndicate the idle sagging coefficient of i-th of node, y0The output voltage values for indicating containing node, that is, refer to
Value, u0Indicate that the control signal for restraining node inputs；It is defeated by the distributed generation resource for giving different on the basis of formula (4)
Out, it determines corresponding control signal input, and then forms the data set [y of control signal input and voltage outputi,ui]。
Scheme is advanced optimized based on NARMA-L2 model energy management method for micro-grid as one kind of the present invention,
The detailed process of step 30 are as follows: according to the obtained data set [y in step 203i,ui], in step 10 opening relationships formula
Complex nonlinear function f [*], g [*], training artificial neural network substitution f [*], g [*], and building is based on this basis
The distributed power controller of NARMA-L2；Distributed power controller is the specific implementation of formula (1), is different targets
Voltage output provides corresponding control signal input u (k).
Scheme is advanced optimized based on NARMA-L2 model energy management method for micro-grid as one kind of the present invention,
The detailed process of step 40 are as follows:
Step 401 sets distributed generation resource target output voltage reference value as yr(k+d), in order in step 2 with it is idle divide equally
For target, the expression formula of target output voltage reference value is found, the distributed power controller to construct in step 30 provides ginseng
Examine value；Target voltage output reference value generator is provided using gradient descent search algorithm:
In formula, j | (i, j) ∈ DGIndicate that i-th, j distributed electrical source node is neighbor node, DGIt is distributed power supply system
The set that interior all neighbor node combinations are constituted；J(yi) indicate the quadratic sum of i-th of node and neighbor node output without work difference
Average value, QjIndicate that the output of j-th of node is idle, m indicates that i-th of distributed electrical source node shares m neighbor node；
Step 402, according to formula (5), J (yi) there are i.e. all distributed generation resources when minimum value to export the difference between idle most
It is small, reach to realize in step 2 and is divided into target with idle；To find J (yi) it is minimum when distributed generation resource target output voltage
Reference value yr(k+d), following circulations are executed until convergence；
(7) formula is expressed as finding the smallest J (yi) value, continuous iteration yiValue, α indicate step-size in search,Indicate J
(yi) to yiDerivation, temp indicate previous cycle yiValue, the condition of convergence is set as recycling y twiceiThe variation of value is less than or equal to pre-
If value.
Scheme is advanced optimized based on NARMA-L2 model energy management method for micro-grid as one kind of the present invention,
Preset value is 0.001.
The invention adopts the above technical scheme compared with prior art, has following technical effect that
The present invention utilizes NARMA-L2 model substitutional theorem formula power supply dynamic mathematical models, realizes that micro-capacitance sensor goes modelling to control
System promotes control speed and precision；Followed by pinning control, is divided equally with reactive power and voltage reverts to rapidly target and generates
The data set of input control signal and voltage output signal is trained in expression formula for characterizing the artificial mind of complex nonlinear function
Through network, to reach good fit effect.Reference value generator is finally provided using gradient descent search algorithm, completes micro-capacitance sensor
Fast and stable.
Detailed description of the invention
Fig. 1 is the flow chart of the embodiment of the present invention.
Fig. 2 is the topology diagram of micro-capacitance sensor of the embodiment of the present invention.
Fig. 3 is the communication topology of micro-capacitance sensor in the embodiment of the present invention.
Fig. 4 is that system internal loading is uprushed the situation of change of each power supply；Wherein, (a) is that the idle power output of each distributed generation resource becomes
Change process is (b) each distributed generation resource voltage change process in the case of sudden load increase.
Fig. 5 is the situation of change of system interior power broken string remaining power supply；Wherein, (a) is distributed generation resource voltage change mistake
Journey (b) is the idle power output state change process of distributed generation resource.
Specific embodiment
To make the objectives, technical solutions, and advantages of the present invention clearer, below in conjunction with the accompanying drawings and the specific embodiments
The present invention will be described in detail.
System in the present invention refers to that distributed power supply system, node refer to distributed electrical source node.
As shown in Figure 1, the embodiment of the method for the present invention, the topological structure for flowing microgrid is as shown in Figure 2.This method includes following
Step:
Step 10 exports the relational expression between control signal input using NARMA-L2 model foundation distributed generation resource；
Step 20 is based on pinning control theory, is divided into target with idle, obtains the input of control signal and voltage output
Data set；
For characterizing complex nonlinear in the relational expression that step 30, the data set training step 10 obtained with step 20 are established
The neural network of function, fitting reactive power divides equally distributed generation resource nonlinear object dynamic characteristic under situation, to be instructed
Practice the relational expression between the distributed generation resource output completed and control signal input；The distributed generation resource output completed according to training
Relational expression between control signal input constructs the distributed power controller based on NARMA-L2；
Step 40 provides target output using gradient descent search algorithm for the distributed power controller that step 30 constructs
Voltage reference value completes micro-capacitance sensor reactive power Collaborative Control.
Scheme is advanced optimized based on NARMA-L2 model energy management method for micro-grid as one kind of the present invention,
In step 10, distributed generation resource dynamic mathematical models are replaced using NARMA-L2 model, by artificial neural network to distribution
Complex nonlinear function part carries out approximate substitution inside power-supply system, between distributed generation resource output and control signal input
Relational expression is as follows:
Y (k+d)=f [y (k), y (k-1) ..., y (k-n+1), u (k), u (k-1), u (k-z+1)]+g [y (k), y (k-
1),…,y(k-n+1),u(k),u(k-1),u(k-z+1)]·u(k) (1)
In formula, u (k) is represented in the control signal input of kth moment distributed generation resource, and y (k) represents kth moment distributed electrical
The output voltage signal in source, y (k+d) represent the output voltage signal of kth+d moment distributed generation resource, and f [*] and g [*] are indicated
The complex nonlinear function of distributed generation system internal nonlinearity part；It is defeated that the distributed generation resource at k-n+1 moment is acquired altogether
Out, y (k-n+1) indicates the distributed generation resource output at (k-n+1) moment；The control signal input at k-z+1 moment is acquired altogether,
U (k-z+1) is that the control signal at (k-z+1) moment inputs.
Scheme is advanced optimized based on NARMA-L2 model energy management method for micro-grid as one kind of the present invention,
The detailed process of step 20 are as follows:
Step 201, for distributed AC servo system micro-capacitance sensor topology, collaboration targeted transformation is become into the tracking of neighbor information error:
In formula, eiIndicate the error between i-th of distributed generation resource node output voltage and reference value, yiIndicate i-th point
Cloth power supply node output voltage, j ∈ NiIndicate that j-th of node is the neighbor node of i-th of node, NiIndicate i-th of distribution
The set that all neighbor nodes of power supply node are constituted, aijIndicate adjacency matrix in distributed power supply systemIn i-th
The value of row, jth column, giIndicate the containing gain of i-th of distributed electrical source node, yjIndicate have in communication topology with i-th of node
The output voltage values of j-th of node of information interchange, y0Indicate the output voltage values of containing node, i.e. reference value；
Step 202 is to guarantee error ei0 is converged to, under conditions of controlling target is that load or burden without work is divided equally, by following public
Formula selection control signal inputs ui:
In formula, Q0It indicates to refer to load or burden without work, QiIndicate that i-th of node exports idle, ωiIndicate i-th of node whether
Through being restrained,Represent the degree of coupling between node；
Step 203, according to the sagging control characteristic of distributed generation resource, i.e., output voltage and it is idle between relational expression, by formula
(3) idle in replaces with output voltage, and control signal input expression formula becomes:
Wherein, nQiIndicate the idle sagging coefficient of i-th of node, y0The output voltage values for indicating containing node, that is, refer to
Value, u0Indicate that the control signal for restraining node inputs；It is defeated by the distributed generation resource for giving different on the basis of formula (4)
Out, it determines corresponding control signal input, and then forms the data set [y of control signal input and voltage outputi,ui]。
Scheme is advanced optimized based on NARMA-L2 model energy management method for micro-grid as one kind of the present invention,
The detailed process of step 30 are as follows: according to the obtained data set [y in step 203i,ui], in step 10 opening relationships formula
Complex nonlinear function f [*], g [*], training artificial neural network substitution f [*], g [*], and building is based on this basis
The distributed power controller of NARMA-L2；Distributed power controller is the specific implementation of formula (1), is different targets
Voltage output provides corresponding control signal input u (k).
Scheme is advanced optimized based on NARMA-L2 model energy management method for micro-grid as one kind of the present invention,
The detailed process of step 40 are as follows:
Step 401 sets distributed generation resource target output voltage reference value as yr(k+d), in order in step 2 with it is idle divide equally
For target, the expression formula of target output voltage reference value is found, the distributed power controller to construct in step 30 provides ginseng
Examine value；Target voltage output reference value generator is provided using gradient descent search algorithm:
In formula, j | (i, j) ∈ DGIndicate that i-th, j distributed electrical source node is neighbor node, DGIt is distributed power supply system
The set that interior all neighbor node combinations are constituted；J(yi) indicate the quadratic sum of i-th of node and neighbor node output without work difference
Average value, QjIndicate that the output of j-th of node is idle, m indicates that i-th of distributed electrical source node shares m neighbor node；
Step 402, according to formula (5), J (yi) there are i.e. all distributed generation resources when minimum value to export the difference between idle most
It is small, reach to realize in step 2 and is divided into target with idle；To find J (yi) it is minimum when distributed generation resource target output voltage
Reference value yr(k+d), following circulations are executed until convergence；
(7) formula is expressed as finding the smallest J (yi) value, continuous iteration yiValue, α indicate step-size in search,Indicate J
(yi) to yiDerivation, temp indicate previous cycle yiValue, the condition of convergence is set as recycling y twiceiThe variation of value is less than or equal to pre-
If value, preset value can be 0.001.
The method of the embodiment of the present invention replaces distributed generation resource dynamic mathematical models using NAMRA-L2 model, realizes
Micro-capacitance sensor removes modelling control.Training dataset is generated using pinning control simultaneously, guarantees the convergence of micro-grid system and steady
It is fixed, modelling control device is finally gone based on above-mentioned generation, reference value is generated using gradient descent search algorithm, keeps system quick
Convergence.
A specific embodiment is enumerated below.
Certain self exchanges micro-capacitance sensor structure as shown in Fig. 2, carrying out microgrid energy management for the microgrid, and microgrid leads to
Sintop is flutterred as shown in figure 3, relevant parameter is as shown in table 1
1 system parameter of table
The application effect of microgrid energy management in the case of being shown below two kinds
Case 1: sudden load increase
Coupling gain c=4 in this case, microgrid operate in island mode, and emulation starts in t=0, and control strategy is in t
=0.5s starts, and internal system increases 40kW burden with power, 20kVA load or burden without work in t=2s.
Step-length α=0.02 of gradient descent search, strategy execution interval 0.1s, the condition of convergence is between neighbor node without work difference
Less than 100Var.
(a) in Fig. 4 is the idle power output change procedure of each distributed generation resource.In (a) in Fig. 4, NARAMA-L2 is used
The system of controller may be implemented it is idle divide equally, when load or burden without work increases suddenly in system, system can be with quick response.May be used also
To guarantee the balance of System Reactive Power distribution, idle circulation is reduced.Obviously, system can restrain in 1s.
(b) in Fig. 4 is each distributed generation resource voltage change process in the case of sudden load increase, if DG1 is to restrain node.It can
To find out, even if load increases, remaining distributed generation resource voltage will not be decreased below reference voltage in system.Namely
It says, restrains node and limited provided with voltage, the voltage output of all DGs will all be remained above the limitation.In supply district, it is
System is not in collapse of voltage problem.
Case 2: system interior power broken string
Coupling gain c=4 in this case, microgrid operate in island mode, and emulation starts in t=0, and control strategy is in t
=0.5s starts.DG3 power supply breaks in t=2s, remaining setting is identical as case 1.
As shown in (a) in Fig. 5, after DG3 broken string, the internal load of system is held by the rest part of distributed generation resource
Load.According to the sagging control characteristic of power generation, the voltage of other three kinds of power supplys should be reduced accordingly.Due to restraining the presence of node, electricity
Pressure is stablized.Remaining voltage changes in the reasonable scope, is not above limit value, and power quality is caused to decline.
(b) in Fig. 5 is the idle power output state change process of distributed generation resource, even if DG3 breaks, remaining distribution hair
The target of reactive power equilibrium still may be implemented in electricity, realizes the stability of power supply, reduces the influence of reactive circular power flow.
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, any
In the technical scope disclosed by the present invention, any changes or substitutions that can be easily thought of by those familiar with the art, all answers
It is included within the scope of protection of the present invention.
Claims (6)
1. one kind is based on NARMA-L2 model energy management method for micro-grid, which comprises the following steps:
Step 10 exports the relational expression between control signal input using NARMA-L2 model foundation distributed generation resource；
Step 20 is based on pinning control theory, is divided into target with idle, obtains the data of control signal input and voltage output
Collection；
For characterizing complex nonlinear function in the relational expression that step 30, the data set training step 10 obtained with step 20 are established
Neural network, fitting reactive power divide equally distributed generation resource nonlinear object dynamic characteristic under situation, to obtain having trained
At distributed generation resource output and control signal input between relational expression；The distributed generation resource output and control completed according to training
Relational expression between signal input processed constructs the distributed power controller based on NARMA-L2；
Step 40 provides target output voltage using the distributed power controller that gradient descent search algorithm constructs for step 30
Reference value completes micro-capacitance sensor reactive power Collaborative Control.
2. according to claim 1 a kind of based on NARMA-L2 model energy management method for micro-grid, which is characterized in that step
In rapid 10, distributed generation resource dynamic mathematical models are replaced using NARMA-L2 model, by artificial neural network to distributed electrical
Source internal system complex nonlinear function part carries out approximate substitution, the pass between distributed generation resource output and control signal input
It is that formula is as follows:
Y (k+d)=f [y (k), y (k-1) ..., y (k-n+1), u (k), u (k-1), u (k-z+1)]+g [y (k), y (k-1) ...
y(k-n+1),u(k),u(k-1),u(k-z+1)]·u(k) (1)
In formula, u (k) is represented in the control signal input of kth moment distributed generation resource, and y (k) represents kth moment distributed generation resource
Output voltage signal, y (k+d) represent the output voltage signal of kth+d moment distributed generation resource, and f [*] and g [*] indicate distribution
The complex nonlinear function of formula electricity generation system internal nonlinearity part；The distributed generation resource output at k-n+1 moment, y are acquired altogether
(k-n+1) the distributed generation resource output at (k-n+1) moment is indicated；The control signal input at k-z+1 moment, u (k- are acquired altogether
Z+1 it) is inputted for the control signal at (k-z+1) moment.
3. according to claim 2 a kind of based on NARMA-L2 model energy management method for micro-grid, which is characterized in that step
Rapid 20 detailed process are as follows:
Step 201, for distributed AC servo system micro-capacitance sensor topology, collaboration targeted transformation is become into the tracking of neighbor information error:
In formula, eiIndicate the error between i-th of distributed generation resource node output voltage and reference value, yiIndicate i-th of distribution
Power supply node output voltage, j ∈ NiIndicate that j-th of node is the neighbor node of i-th of node, NiIndicate i-th of distributed generation resource
The set that all neighbor nodes of node are constituted, aijIndicate adjacency matrix in distributed power supply systemIn the i-th row,
The value of jth column, giIndicate the containing gain of i-th of distributed electrical source node, yjIndicate there is letter with i-th of node in communication topology
Cease the output voltage values of j-th of node of exchange, y0Indicate the output voltage values of containing node, i.e. reference value；
Step 202 is to guarantee error ei0 is converged to, under conditions of controlling target is that load or burden without work is divided equally, is selected as follows
Select control signal input ui:
In formula, Q0It indicates to refer to load or burden without work, QiIndicate that i-th of node exports idle, ωiIndicate i-th of node whether by
It restrains,Represent the degree of coupling between node；
Step 203, according to the sagging control characteristic of distributed generation resource, i.e., output voltage and it is idle between relational expression, will be in formula (3)
It is idle replace with output voltage, control signal input expression formula becomes:
Wherein, nQiIndicate the idle sagging coefficient of i-th of node, y0Indicate the output voltage values of containing node, i.e. reference value, u0
Indicate that the control signal for restraining node inputs；On the basis of formula (4), exported by giving different distributed generation resources, really
Fixed corresponding control signal input, and then form the data set [y of control signal input and voltage outputi,ui]。
4. according to claim 3 a kind of based on NARMA-L2 model energy management method for micro-grid, which is characterized in that step
Rapid 30 detailed process are as follows: according to the obtained data set [y in step 203i,ui], for answering in step 10 opening relationships formula
Miscellaneous nonlinear function f [*], g [*], training artificial neural network substitution f [*], g [*], and building is based on this basis
The distributed power controller of NARMA-L2；Distributed power controller is the specific implementation of formula (1), is different targets
Voltage output provides corresponding control signal input u (k).
5. according to claim 3 a kind of based on NARMA-L2 model energy management method for micro-grid, which is characterized in that step
Rapid 40 detailed process are as follows:
Step 401 sets distributed generation resource target output voltage reference value as yr(k+d), in order to be divided into mesh in step 2 with idle
Mark, finds the expression formula of target output voltage reference value, the distributed power controller to construct in step 30 provides reference value；
Target voltage output reference value generator is provided using gradient descent search algorithm:
In formula, j | (i, j) ∈ DGIndicate that i-th, j distributed electrical source node is neighbor node, DGIt is institute in distributed power supply system
The set for thering is neighbor node combination to constitute；J(yi) indicate that i-th of node exports being averaged for the quadratic sum without work difference with neighbor node
Value, QjIndicate that the output of j-th of node is idle, m indicates that i-th of distributed electrical source node shares m neighbor node；
Step 402, according to formula (5), J (yi) have the difference that i.e. all distributed generation resources export between idle when minimum value minimum, reach
It realizes in step 2 and is divided into target with idle；To find J (yi) it is minimum when distributed generation resource target output voltage reference value yr
(k+d), following circulations are executed until convergence；
(7) formula is expressed as finding the smallest J (yi) value, continuous iteration yiValue, α indicate step-size in search,Indicate J (yi)
To yiDerivation, temp indicate previous cycle yiValue, the condition of convergence is set as recycling y twiceiThe variation of value is less than or equal to default
Value.
6. according to claim 5 a kind of based on NARMA-L2 model energy management method for micro-grid, which is characterized in that pre-
If value is 0.001.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811054866.5A CN109066707B (en) | 2018-09-11 | 2018-09-11 | Micro-grid energy management method based on NARMA-L2 model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811054866.5A CN109066707B (en) | 2018-09-11 | 2018-09-11 | Micro-grid energy management method based on NARMA-L2 model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109066707A true CN109066707A (en) | 2018-12-21 |
| CN109066707B CN109066707B (en) | 2021-03-02 |
Family
ID=64761199
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811054866.5A Expired - Fee Related CN109066707B (en) | 2018-09-11 | 2018-09-11 | Micro-grid energy management method based on NARMA-L2 model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109066707B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103268366A (en) * | 2013-03-06 | 2013-08-28 | 辽宁省电力有限公司电力科学研究院 | Combined wind power prediction method suitable for distributed wind power plant |
| CN103488869A (en) * | 2013-08-23 | 2014-01-01 | 上海交通大学 | Wind power generation short-term load forecast method of least squares support vector machine |
| CN104659811A (en) * | 2015-01-28 | 2015-05-27 | 东南大学 | Distributed cooperative control method of micro power grid on basis of holdback |
| CN104992248A (en) * | 2015-07-07 | 2015-10-21 | 中山大学 | Microgrid photovoltaic power station generating capacity combined forecasting method |
| US20160276830A1 (en) * | 2015-03-16 | 2016-09-22 | Board Of Regents, The University Of Texas System | System and Method for Distributed Control of an Electrical Network |
| CN107069776A (en) * | 2017-04-12 | 2017-08-18 | 东南大学 | A kind of energy storage prediction distributed control method of smooth microgrid dominant eigenvalues |
-
2018
- 2018-09-11 CN CN201811054866.5A patent/CN109066707B/en not_active Expired - Fee Related
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103268366A (en) * | 2013-03-06 | 2013-08-28 | 辽宁省电力有限公司电力科学研究院 | Combined wind power prediction method suitable for distributed wind power plant |
| CN103488869A (en) * | 2013-08-23 | 2014-01-01 | 上海交通大学 | Wind power generation short-term load forecast method of least squares support vector machine |
| CN104659811A (en) * | 2015-01-28 | 2015-05-27 | 东南大学 | Distributed cooperative control method of micro power grid on basis of holdback |
| US20160276830A1 (en) * | 2015-03-16 | 2016-09-22 | Board Of Regents, The University Of Texas System | System and Method for Distributed Control of an Electrical Network |
| CN104992248A (en) * | 2015-07-07 | 2015-10-21 | 中山大学 | Microgrid photovoltaic power station generating capacity combined forecasting method |
| CN107069776A (en) * | 2017-04-12 | 2017-08-18 | 东南大学 | A kind of energy storage prediction distributed control method of smooth microgrid dominant eigenvalues |
Non-Patent Citations (1)
| Title |
|---|
| 薛蕊 等: "基于NARMA_L2控制器的电力***稳定性分析", 《现代电子技术》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109066707B (en) | 2021-03-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Arya | Automatic generation control of two-area electrical power systems via optimal fuzzy classical controller | |
| Tungadio et al. | Load frequency controllers considering renewable energy integration in power system | |
| Li et al. | Coordinated load frequency control of multi-area integrated energy system using multi-agent deep reinforcement learning | |
| Li et al. | A distributed coordination control based on finite-time consensus algorithm for a cluster of DC microgrids | |
| Zhao et al. | Distributed robust frequency restoration and active power sharing for autonomous microgrids with event-triggered strategy | |
| CN106849112B (en) | Power distribution network multi-objective reactive optimization method based on non-dominant neighborhood immune algorithm | |
| CN110265991B (en) | Distributed coordination control method for direct-current micro-grid | |
| WO2018058804A1 (en) | Universal microgrid cluster distributed control method comprising constant power and droop control | |
| Long et al. | A hierarchical distributed MPC for HVAC systems | |
| Zou et al. | Practical predefined-time output-feedback consensus tracking control for multiagent systems | |
| Liu et al. | A distributed iterative learning framework for DC microgrids: Current sharing and voltage regulation | |
| El Helou et al. | Fully decentralized reinforcement learning-based control of photovoltaics in distribution grids for joint provision of real and reactive power | |
| Xi et al. | A virtual generation ecosystem control strategy for automatic generation control of interconnected microgrids | |
| CN107065556A (en) | A kind of automatic search method of reactor core unit Variable power optimization of operation strategy scheme | |
| Wang et al. | Toward balancing dynamic performance and system stability for DC microgrids: A new decentralized adaptive control strategy | |
| CN108471109B (en) | Unified distributed control method and system for direct-current multi-microgrid system | |
| CN109286204A (en) | A kind of distribution network black starting-up reconstructing method based on minimum expectation loss of outage | |
| CN105896547B (en) | A kind of bulk power grid hierarchical voltage control method under wind power integration | |
| CN111490551A (en) | Distributed Newton method-based power distribution network photovoltaic power generation cluster voltage control method | |
| Yin et al. | Expandable deep width learning for voltage control of three-state energy model based smart grids containing flexible energy sources | |
| CN110544960A (en) | distributed control method for improving reactive power sharing capability of island microgrid | |
| Chen et al. | Distributed optimization of single-integrator systems with prescribed-time convergence | |
| CN109066707A (en) | One kind being based on NARMA-L2 model energy management method for micro-grid | |
| Han et al. | Distributed Containment Control Strategy for the Dynamic Stabilization of Integrated Energy System With Multiple Virtual Leaders | |
| Cheng et al. | Optimal energy management of energy internet: A distributed actor-critic reinforcement learning method |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20200821 Address after: Four pailou Nanjing Xuanwu District of Jiangsu Province, No. 2 210096 Applicant after: SOUTHEAST University Applicant after: YANGZHOU POWER SUPPLY BRANCH OF STATE GRID JIANGSU ELECTRIC POWER Co.,Ltd. Address before: Four pailou Nanjing Xuanwu District of Jiangsu Province, No. 2 210096 Applicant before: SOUTHEAST University |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20210302 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |