CN112600244A - Photovoltaic power station voltage control method based on neural network - Google Patents

Photovoltaic power station voltage control method based on neural network Download PDF

Info

Publication number
CN112600244A
CN112600244A CN202011434418.5A CN202011434418A CN112600244A CN 112600244 A CN112600244 A CN 112600244A CN 202011434418 A CN202011434418 A CN 202011434418A CN 112600244 A CN112600244 A CN 112600244A
Authority
CN
China
Prior art keywords
photovoltaic
voltage
power
neural network
bus
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
Application number
CN202011434418.5A
Other languages
Chinese (zh)
Other versions
CN112600244B (en
Inventor
陈之轩
窦春霞
赵昕
马建川
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangsu Paiergao Intelligent Technology Co ltd
Original Assignee
Jiangsu Paiergao Intelligent Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Jiangsu Paiergao Intelligent Technology Co ltd filed Critical Jiangsu Paiergao Intelligent Technology Co ltd
Priority to CN202011434418.5A priority Critical patent/CN112600244B/en
Publication of CN112600244A publication Critical patent/CN112600244A/en
Application granted granted Critical
Publication of CN112600244B publication Critical patent/CN112600244B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

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
    • H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
    • H02J3/381—Dispersed generators
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply
    • 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
    • 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
    • H02J3/46—Controlling of the sharing of output between the generators, converters, or transformers
    • H02J3/466—Scheduling the operation of the generators, e.g. connecting or disconnecting generators to meet a given demand
    • 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
    • H02J3/46—Controlling of the sharing of output between the generators, converters, or transformers
    • H02J3/48—Controlling the sharing of the in-phase component
    • 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
    • H02J3/46—Controlling of the sharing of output between the generators, converters, or transformers
    • H02J3/50—Controlling the sharing of the out-of-phase component
    • 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]
    • 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
    • H02J2300/00—Systems for supplying or distributing electric power characterised by decentralized, dispersed, or local generation
    • H02J2300/20—The dispersed energy generation being of renewable origin
    • H02J2300/22—The renewable source being solar energy
    • H02J2300/24—The renewable source being solar energy of photovoltaic origin
    • 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
    • H02J2300/00—Systems for supplying or distributing electric power characterised by decentralized, dispersed, or local generation
    • H02J2300/40—Systems for supplying or distributing electric power characterised by decentralized, dispersed, or local generation wherein a plurality of decentralised, dispersed or local energy generation technologies are operated simultaneously
    • 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

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Power Engineering (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • General Health & Medical Sciences (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • General Engineering & Computer Science (AREA)
  • Molecular Biology (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Evolutionary Computation (AREA)
  • Marketing (AREA)
  • Biophysics (AREA)
  • General Business, Economics & Management (AREA)
  • Biomedical Technology (AREA)
  • Computing Systems (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Tourism & Hospitality (AREA)
  • Game Theory and Decision Science (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Development Economics (AREA)
  • Public Health (AREA)
  • Water Supply & Treatment (AREA)
  • Primary Health Care (AREA)
  • Control Of Electrical Variables (AREA)

Abstract

The invention discloses a photovoltaic power station voltage control method based on a neural network, which comprises the following specific steps: the method comprises the following steps: the bus voltage active power neural network and the bus voltage reactive power neural network respectively output the sensitivity coefficients of the corresponding bus voltage active power and the corresponding bus voltage reactive power to the model predictive controller; step two: and the model prediction controller constructs a mathematical model of the bus voltage and the node injection active power and reactive power through the sensitivity coefficients of the bus voltage active power and the bus voltage reactive power issued by the upper-layer control framework. The photovoltaic power station voltage control method based on the neural network can solve the problems of over voltage and under voltage of a photovoltaic power station bus, effectively improves the practicability and expandability of the method based on sensitivity analysis, has important significance on safe operation of a power distribution network, and is used for boosting energy Internet construction, improving power service quality and optimally utilizing various resources.

Description

Photovoltaic power station voltage control method based on neural network
Technical Field
The invention relates to the field of voltage regulation and control of a photovoltaic power station of a power distribution network, in particular to a photovoltaic power station voltage control method based on a neural network.
Background
Along with the continuous penetration of new energy in a power distribution network, the traditional power distribution network is gradually developed into an active power distribution network, the gradual depletion of fossil energy and the improvement of environmental friendliness are solved, meanwhile, a plurality of challenges are provided for the smooth operation and planning of the power distribution network, a photovoltaic power station gathers a large number of distributed photovoltaics, is an important component element in the power distribution network, and has an irreplaceable effect on the stable operation of supporting the load storage of a source network of the whole power distribution network, and the intermittence and randomness of the distributed photovoltaics can cause the fluctuation of the bus voltage of the whole photovoltaic power station, seriously harm the safety of the whole power distribution network, even cause the splitting of the power grid, effectively solve the problems of the over-voltage and under-voltage of the bus of the photovoltaic power station, and is an important research;
in the existing photovoltaic power station voltage control method based on the neural network, in the actual operation of a photovoltaic power station, the topology and network parameters of a system are not easy to obtain, so that the practical application of the method based on sensitivity analysis has a plurality of difficulties, and the problems of over-voltage and under-voltage of a photovoltaic power station bus cannot be solved.
Disclosure of Invention
The invention aims to provide a photovoltaic power station voltage control method based on a neural network, and aims to solve the problems that in the existing photovoltaic power station voltage control method based on the neural network, in the actual operation of a photovoltaic power station, the topology and network parameters of a system are not easy to obtain, so that the actual application of the method based on sensitivity analysis has a plurality of difficulties, and the problems of over-voltage and under-voltage of a photovoltaic power station bus cannot be solved.
In order to achieve the purpose, the invention provides the following technical scheme: a photovoltaic power station voltage control method based on a neural network comprises the following specific steps:
the method comprises the following steps: the bus voltage active power neural network and the bus voltage reactive power neural network respectively output the sensitivity coefficients of the corresponding bus voltage active power and the corresponding bus voltage reactive power to the model predictive controller;
step two: and the model prediction controller constructs a mathematical model of the bus voltage and the node injection active power and reactive power through the sensitivity coefficients of the bus voltage active power and the bus voltage reactive power issued by the upper-layer control framework.
Furthermore, the bus voltage active power neural network and the bus voltage reactive power neural network form an upper-layer control framework, and the model prediction controller is a component of a lower-layer control framework;
the bus voltage active power neural network comprises an input layer, a hidden layer and an output layer, input data of the input layer of the bus voltage active power neural network comprises active power data of nodes and voltage deviation data of buses, and output data of the output layer of the bus voltage active power neural network is a sensitivity coefficient of bus voltage active power of photovoltaic nodes;
further, the active power data of the node can be obtained by equation (1):
Figure BDA0002827720250000021
in the formula (1), UjAnd IjThe voltage and current of the jth photovoltaic node are respectively marked by a conjugate operation and Re is an operation of an arithmetic part, and the voltage deviation data of the bus can be obtained by the formula (2)Obtaining:
Figure BDA0002827720250000022
in the formula (2), the reaction mixture is,
Figure BDA0002827720250000023
and
Figure BDA0002827720250000024
the voltage and the reference voltage value of the ith photovoltaic power station bus are respectively;
further, the reactive power data of the node may be obtained by equation (3):
Figure BDA0002827720250000025
in the formula (3), UjAnd IjThe voltage and the current of the jth photovoltaic node are respectively marked by a conjugate operation, and Im is an imaginary part calculation.
Further, the specific steps of obtaining the sensitivity coefficient of the active power of the bus voltage and the sensitivity coefficient of the reactive power of the bus voltage are as follows:
the method comprises the following steps: preprocessing active power data of the photovoltaic nodes, voltage deviation data of the bus and reactive power data of the photovoltaic nodes obtained through calculation, wherein the preprocessing modes are respectively shown as formulas (4), (5) and (6):
Figure BDA0002827720250000031
Figure BDA0002827720250000032
Figure BDA0002827720250000033
in the formula (4), P'jAnd
Figure BDA0002827720250000034
respectively obtaining an active power normalized value and an active power rated value of the jth photovoltaic node; in the formula (5), the reaction mixture is,
Figure BDA0002827720250000035
and
Figure BDA0002827720250000036
respectively obtaining a voltage deviation normalized value and a maximum value of the voltage deviation of the ith bus; q 'in the formula (6)'jAnd
Figure BDA0002827720250000037
respectively obtaining a reactive power normalized value and a reactive power rated value of the jth photovoltaic node;
step two: after active power data of the photovoltaic nodes, voltage deviation data of the bus and reactive power data of the photovoltaic nodes are preprocessed, the active power data of the photovoltaic nodes and the voltage deviation data of the bus are used as input of a bus voltage active power neural network, the reactive power data of the photovoltaic nodes and the voltage deviation data of the bus are used as input of a bus voltage reactive power neural network, the input of the bus voltage reactive power neural network is respectively input into corresponding neural networks for training, and weight values of the corresponding neural networks are corrected according to a formula (7):
Figure BDA0002827720250000038
in the formula (7), Ep(k)、wp(k)、Δwp(k) Respectively an output quantity error value, a weight matrix and a weight updating quantity when the k-th weight of the bus voltage/active power neural network is updated, Eq(k)、wq(k)、Δwq(k) The output quantity error value, the weight matrix and the weight updating quantity at the time of the kth weight updating of the bus voltage/reactive power neural network are respectively, and the method comprises the following steps:
Figure BDA0002827720250000041
in the formula (8), Yp(k)、Y'p(k) Respectively the real value and the actual value of the output quantity during the k-th weight updating of the bus voltage/active power neural network, Yq(k)、Y'q(k) Respectively obtaining the real value and the actual value of the output quantity when the k-th weight of the bus voltage/reactive power neural network is updated;
step three: when the errors respectively meet corresponding conditions, the updating of the weight matrix is stopped, and the final output quantity is sent to the model prediction controller of the lower layer,
Figure BDA0002827720250000042
in the formula (9), γpAnd gammaqError thresholds are preset for the bus voltage/active power neural network and the bus voltage/reactive power neural network respectively, and there are:
Figure BDA0002827720250000043
further, the model predictive controller solves the corresponding optimal control quantity through real-time rolling optimization, the objective function of the model predictive controller is composed of a multi-objective optimization function, the multi-objective optimization function specifically comprises two groups of objective functions, one group is a bus voltage deviation value of a minimized photovoltaic power station, the other group is an active control cost of a minimized photovoltaic node, and the objective function is a first item of the multi-objective optimization function as shown in formula (11):
Figure BDA0002827720250000044
in the formula (11), the reaction mixture is,
Figure BDA0002827720250000045
for the voltage deviation value of the kth step of the ith bus predicted based on the current time, NpAnd NbRespectively predicting the domain length and the total number of the photovoltaic power station buses;
the second term of the objective function of the lower model predictive controller is shown as the formula (12):
Figure BDA0002827720250000051
in the formula (12), Fj(k) Active control cost for kth step of jth photovoltaic node predicted based on current time, NpAnd NpvRespectively the predicted domain length and the total number of photovoltaic nodes, Fj(k) Is a quadratic function, as shown in equation (13):
Figure BDA0002827720250000052
in the formula (13), aj、bj、cjRespectively, the active control cost coefficient, delta P, of the jth photovoltaic nodej(k) The active control quantity of the kth step of the jth photovoltaic node predicted based on the current moment is obtained;
the objective function of the lower model predictive controller is formed by weighting objective functions in an equation (11) and an equation (12), specifically as shown in an equation (14), and is constrained as shown in equations (15) to (20):
min(λ1·Obj1+λ2·Obj2) (14)
Figure BDA0002827720250000053
Figure BDA0002827720250000054
Figure BDA0002827720250000055
Figure BDA0002827720250000056
Figure BDA0002827720250000057
Figure BDA0002827720250000058
in formula (14), λ1、λ2Are respectively the corresponding weight matrix, NcIs the control field length; equations (15) and (16) represent the ramp constraints for the active and reactive power of the jth photovoltaic node, respectively, wherein,
Figure BDA0002827720250000059
respectively carrying out climbing upper limit constraint, climbing lower limit constraint, reactive power climbing upper limit constraint and climbing lower limit constraint on active power of the jth photovoltaic node; equations (17) and (18) represent the upper and lower constraints on the active power and reactive power output of the jth photovoltaic node, respectively, wherein,
Figure BDA0002827720250000061
respectively are the output upper limit constraint, the output lower limit constraint, the output upper limit constraint and the output lower limit constraint of the active power, the output lower limit constraint, the reactive power of the jth photovoltaic node, Pj(t0)、Qj(t0) Respectively outputting active power and reactive power of the jth photovoltaic node at the current moment; equation (19) is the relationship between the bus node voltage change and the injection of active and reactive power into each photovoltaic node, wherein,
Figure BDA0002827720250000062
for bus node voltage variation/lightThe sensitivity coefficient of active power injected into the voltage node and the final output quantity from the bus voltage/active power neural network in the upper-layer control
Figure BDA0002827720250000063
Sensitivity coefficient for bus node voltage variation/photovoltaic node injection reactive power, final output from bus voltage/reactive power neural network in upper layer control
Figure BDA0002827720250000064
(according to the formula (10)); equation (20) represents the output active power constraint of the entire photovoltaic power plant, wherein,
Figure BDA0002827720250000065
and the active power reference output value of the whole photovoltaic power station comes from a power grid dispatching center.
Furthermore, the weight matrix of the objective function of the model predictive controller can be changed through the actual value of the photovoltaic power station bus voltage, and the multi-objective model is controlled through multiple modes, wherein the multiple modes refer to whether the photovoltaic bus voltage is in a normal operation range.
The multi-mode is that whether the photovoltaic bus voltage is in a normal operation range or not, and the specific judgment method is as follows:
Figure BDA0002827720250000066
in the formula (21), the compound represented by the formula,
Figure BDA0002827720250000067
is a reference voltage value, U, of the ith bus of the photovoltaic power stationi(t0) Is the actual voltage value, U, of the ith bus of the photovoltaic power station at the current momenti(t0) Is the safety margin coefficient of the voltage of the ith bus of the photovoltaic power station, sig (t)0) Is a modal indication function of the photovoltaic power station, wherein when the value of the modal indication function is 1, the modal indication function represents that the photovoltaic power station generates over/under voltage phenomenon, and when the value of the modal indication function is 0, the modal indication function represents the photovoltaic power stationAnd (5) normally running. Lambda [ alpha ]1And λ2The value of (b) satisfies the following formula:
Figure BDA0002827720250000071
λ2＝I (23)
equation (22) represents that the greater the extent to which the photovoltaic plant busbar voltage exceeds its normal operating range, the corresponding diagonal matrix λ1The larger the element value of (A), the corresponding diagonal matrix lambda is if the photovoltaic power station normally operates1The lower layer model predictive controller has an objective function degenerated into a single objective function from a multi-objective function, namely, the optimization objective is to minimize the active output cost of the photovoltaic power station under the normal operation condition of the photovoltaic power station, wherein lambda is1(m) is a diagonal matrix lambda1The m-th element of (a) is,
Figure BDA0002827720250000072
is a diagonal matrix lambda1Dimension of (a), λεIs a weight coefficient; in the formula (23), I is an identity matrix, the objective function and the constraint of the model predictive controller are solved through a particle swarm algorithm, and the solved optimal control quantity of the active power and the reactive power
Figure BDA0002827720250000073
And
Figure BDA0002827720250000074
and respectively transmitting the signals to the corresponding inverter terminal controllers.
Compared with the prior art, the invention has the beneficial effects that: the photovoltaic power station voltage control method based on the neural network carries out rolling optimization by designing a corresponding objective function, can solve the optimal output reference values of active power and reactive power of the inverter in real time and further sends the output reference values to the corresponding photovoltaic node inverter, and the solved optimal control quantity of the active power and the reactive power is respectively issued to the corresponding inverter terminal controller, thereby solving the problems of over voltage and under voltage of the bus of the photovoltaic power station, obtaining the relation between the voltage change of the photovoltaic node and the injection power in a data driving mode, training the neural network according to historical operating data can obtain the active and reactive sensitivity coefficients, therefore, the bus voltage control of the photovoltaic power station is realized, the practicability and expandability of the method based on sensitivity analysis are effectively improved, and the method has important significance on the safe operation of a power distribution network;
the photovoltaic power station voltage control method based on the neural network is characterized in that the neural network for acquiring the sensitivity coefficient is correspondingly designed based on a data-driven thought, the problem that the sensitivity coefficient is difficult to acquire due to the fact that parameters such as impedance in an actual power grid for sensitivity analysis are difficult to measure is solved, meanwhile, a model prediction controller under multi-mode switching is designed, multi-objective optimization of the photovoltaic power station under different operating conditions is achieved, the integrated design of safety control and steady-state control of the photovoltaic power station is achieved, a new method is provided for photovoltaic power station bus voltage regulation, and the control method has important significance in boosting energy internet construction, achieving national energy development strategy, improving power service quality, optimizing utilization of various resources and the like.
Drawings
FIG. 1 is a block diagram of a bus voltage active power neural network of the present invention;
FIG. 2 is a block diagram of the bus voltage reactive power neural network of the present invention;
fig. 3 is a bus voltage hierarchical control architecture of a photovoltaic power station of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1-3, the present invention provides a technical solution: a photovoltaic power station voltage control method based on a neural network comprises the following specific steps:
the method comprises the following steps: the bus voltage active power neural network and the bus voltage reactive power neural network respectively output the sensitivity coefficients of the corresponding bus voltage active power and the corresponding bus voltage reactive power to the model predictive controller;
step two: and the model prediction controller constructs a mathematical model of the bus voltage and the node injection active power and reactive power through the sensitivity coefficients of the bus voltage active power and the bus voltage reactive power issued by the upper-layer control framework.
The bus voltage active power neural network and the bus voltage reactive power neural network form an upper-layer control framework, and the model prediction controller is a component of a lower-layer control framework;
the bus voltage active power neural network comprises an input layer, a hidden layer and an output layer, input data of the input layer of the bus voltage active power neural network comprises active power data of nodes and voltage deviation data of buses, output data of the output layer of the bus voltage active power neural network is a sensitivity coefficient of bus voltage active power of photovoltaic nodes, and the active power data of the nodes can be obtained through the formula (1):
Figure BDA0002827720250000091
in formula (24), UjAnd IjThe voltage and the current of the jth photovoltaic node are respectively marked by a conjugate operation, and Re is an operation of an actual calculation part. The voltage deviation data of the bus can be obtained by equation (2):
Figure BDA0002827720250000092
in the formula (25), the reaction mixture,
Figure BDA0002827720250000093
and
Figure BDA0002827720250000094
the voltage and the reference voltage value of the ith photovoltaic power station bus are respectively.
And the output data of the bus voltage active power neural network output layer is the sensitivity coefficient of the bus voltage active power of the photovoltaic node.
As shown in fig. 2, the bus voltage reactive power neural network in the upper control architecture is composed of an input layer, a hidden layer and an output layer, input data of the bus voltage reactive power neural network input layer includes reactive power data of nodes and voltage deviation data of the bus, and the reactive power data of the nodes can be obtained by equation (3):
Figure BDA0002827720250000095
in the formula (3), UjAnd IjThe voltage and the current of the jth photovoltaic node are respectively marked by a conjugate operation, and Im is an imaginary part calculation.
And the output data of the bus voltage reactive power neural network output layer is the sensitivity coefficient of the bus voltage reactive power of the photovoltaic node.
The method specifically comprises the following three steps of obtaining the sensitivity coefficient of the active power of the bus voltage and the sensitivity coefficient of the reactive power of the bus voltage:
the method comprises the following steps: preprocessing active power data of the photovoltaic nodes, voltage deviation data of the buses and reactive power data of the photovoltaic nodes obtained by calculation of the formulas (1), (2) and (3), wherein the preprocessing modes are respectively shown as formulas (4), (5) and (6):
Figure BDA0002827720250000101
Figure BDA0002827720250000102
Figure BDA0002827720250000103
in the formula (4), P'jAnd
Figure BDA0002827720250000104
respectively obtaining an active power normalized value and an active power rated value of the jth photovoltaic node; in the formula (5), the reaction mixture is,
Figure BDA0002827720250000105
and
Figure BDA0002827720250000106
respectively obtaining a voltage deviation normalized value and a maximum value of the voltage deviation of the ith bus; q 'in the formula (6)'jAnd
Figure BDA0002827720250000107
respectively, the reactive power normalized value and the reactive power rated value of the jth photovoltaic node.
Step two: after active power data of the photovoltaic nodes, voltage deviation data of the bus and reactive power data of the photovoltaic nodes are preprocessed, the active power data of the photovoltaic nodes and the voltage deviation data of the bus are used as input of a bus voltage active power neural network, the reactive power data of the photovoltaic nodes and the voltage deviation data of the bus are used as input of a bus voltage reactive power neural network, the input of the bus voltage reactive power neural network is respectively input into corresponding neural networks for training, and weight values of the corresponding neural networks are corrected according to a formula (7):
Figure BDA0002827720250000108
in the formula (7), Ep(k)、wp(k)、Δwp(k) Respectively an output quantity error value, a weight matrix and a weight updating quantity when the k-th weight of the bus voltage active power neural network is updated, Eq(k)、wq(k)、Δwq(k) Are respectively provided withThe output quantity error value, the weight matrix and the weight updating quantity at the time of the kth weight updating of the bus voltage reactive power neural network are shown, and the method comprises the following steps:
Figure BDA0002827720250000111
in the formula (8), Yp(k)、Y'p(k) Respectively the real value and the actual value of the output quantity during the k-th weight updating of the bus voltage active power neural network, Yq(k)、Y'q(k) And respectively representing the real value and the actual value of the output quantity when the k-th weight of the bus voltage reactive power neural network is updated.
Step three: when error Ep(k)、Eq(k) When the conditions of formula (9) are satisfied, the updating of the weight matrix w is stoppedp(k) And wq(k) And will output the final output
Figure BDA0002827720250000112
And
Figure BDA0002827720250000113
and sending the data to a model predictive controller.
Figure BDA0002827720250000114
In the formula (10), γpAnd gammaqError thresholds preset for the bus voltage active power neural network and the bus voltage reactive power neural network respectively, and having:
Figure BDA0002827720250000115
as shown in fig. 3, the model predictive controller solves the corresponding optimal control quantity through real-time rolling optimization, the objective function of the model predictive controller is formed by a multi-objective optimization function, the multi-objective optimization function specifically includes two sets of objective functions, one of the two sets is to minimize the bus voltage deviation value of the photovoltaic power station, and the other set is to minimize the active control cost of the photovoltaic node, as shown in formula (11), the first item of the multi-objective optimization function is:
Figure BDA0002827720250000116
in the formula (11), the reaction mixture is,
Figure BDA0002827720250000117
for the voltage deviation value of the kth step of the ith bus predicted based on the current time, NpAnd NbRespectively the total number of the prediction domain length and the photovoltaic power station bus.
The second term of the objective function of the model predictive controller is as shown in equation (12):
Figure BDA0002827720250000121
in the formula (12), Fj(k) Active control cost for kth step of jth photovoltaic node predicted based on current time, NpAnd NpvRespectively the predicted domain length and the total number of photovoltaic nodes, Fj(k) Is a quadratic function, as shown in equation (13):
Figure BDA0002827720250000122
in the formula (13), aj、bj、cjRespectively, the active control cost coefficient, delta P, of the jth photovoltaic nodej(k) The active control quantity of the kth step of the jth photovoltaic node predicted based on the current moment is obtained.
The objective function of the model predictive controller is formed by weighting the objective functions in the formula (11) and the formula (12), specifically as shown in the formula (14), and is constrained as shown in the formulas (15) to (20):
min(λ1·Obj1+λ2·Obj2) (14)
s.t.
Figure BDA0002827720250000123
Figure BDA0002827720250000124
Figure BDA0002827720250000125
Figure BDA0002827720250000126
Figure BDA0002827720250000127
Figure BDA0002827720250000128
in formula (14), λ1、λ2Are respectively the corresponding weight matrix, NcIs the control field length; equations (15) and (16) represent the ramp constraints for the active and reactive power of the jth photovoltaic node, respectively, wherein,
Figure BDA0002827720250000131
respectively representing the climbing upper limit constraint, the climbing lower limit constraint, the climbing upper limit constraint and the climbing lower limit constraint of the active power, the reactive power and the climbing lower limit constraint of the jth photovoltaic node, wherein the equations (17) and (18) respectively represent the output upper and lower limit constraints of the active power and the reactive power of the jth photovoltaic node,
Figure BDA0002827720250000132
the output upper limit constraint, the output lower limit constraint and the reactive power of the active power of the jth photovoltaic node are respectivelyUpper and lower force limits, Pj(t0)、Qj(t0) Respectively output the active power and the reactive power of the jth photovoltaic node at the current moment, and the equation (19) is the relation between the voltage change of the bus node and the injection of the active power and the reactive power of each photovoltaic node, wherein,
Figure BDA0002827720250000133
the sensitivity coefficient of active power is injected into the photovoltaic node for the voltage change of the bus node, and the final output quantity of the bus voltage active power neural network in the upper control framework
Figure BDA0002827720250000134
The sensitivity coefficient of reactive power injected into the photovoltaic node for the voltage change of the bus node is the final output quantity from the bus voltage reactive power neural network in the upper control framework
Figure BDA0002827720250000135
(as can be seen from equation (10)), equation (20) represents the output active power constraint of the entire photovoltaic power plant, wherein,
Figure BDA0002827720250000136
and the active power reference output value of the whole photovoltaic power station comes from a power grid dispatching center.
Weight matrix lambda of an objective function of a model predictive controller1And λ2The method can be changed according to the actual value of the bus voltage of the photovoltaic power station, so that multi-target model predictive control under multi-mode switching is realized, the multi-mode means whether the photovoltaic bus voltage is in a normal operation range, and the specific judgment method comprises the following steps:
Figure BDA0002827720250000137
in the formula (21), the compound represented by the formula,
Figure BDA0002827720250000138
is a reference voltage value, U, of the ith bus of the photovoltaic power stationi(t0) Is the actual voltage value, U, of the ith bus of the photovoltaic power station at the current momenti(t0) Is the safety margin coefficient of the voltage of the ith bus of the photovoltaic power station, sig (t)0) Is a modal indication function of the photovoltaic power station, when the value of the modal indication function is 1, the modal indication function represents that the photovoltaic power station generates an over-under voltage phenomenon, when the value of the modal indication function is 0, the modal indication function represents that the photovoltaic power station normally operates, and lambda1And λ2The value of (b) satisfies the following formula:
Figure BDA0002827720250000141
λ2＝I (23)
equation (22) represents that the greater the extent to which the photovoltaic plant busbar voltage exceeds its normal operating range, the corresponding diagonal matrix λ1The larger the element value of (A), the corresponding diagonal matrix lambda is if the photovoltaic power station normally operates1The model predictive controller has an all-zero matrix, the objective function of the model predictive controller is degenerated from a multi-objective function to a single objective function, namely the optimization objective is to minimize the active output cost of the photovoltaic power station under the normal operation condition, wherein lambda1(m) is a diagonal matrix lambda1The m-th element of (a) is,
Figure BDA0002827720250000142
is a diagonal matrix lambda1Dimension of (a), λεIs a weight coefficient; in formula (23), I is an identity matrix.
Solving the objective function and the constraint of the model predictive controller by a particle swarm algorithm, and solving the optimal control quantity of the active power and the reactive power
Figure BDA0002827720250000143
And
Figure BDA0002827720250000144
respectively issued to corresponding inverter terminal controllers, thereby solving the problems of excessive shortage of the photovoltaic power station busVoltage problems.
The working principle is as follows: for the photovoltaic power station voltage control method based on the neural network, firstly, a deep neural network is utilized to respectively construct a photovoltaic node voltage change and inject active and reactive network models in a data driving mode, corresponding quantitative relations are mined, impedance parameters of an actual measurement system are not needed, the problem that reactive sensitivity in an actual power grid is difficult to obtain can be effectively solved, secondly, a bus voltage control scheme based on model predictive control is designed based on obtained active and reactive sensitivity coefficients, an optimal inverter output active power and reactive power reference value is solved in real time through online rolling optimization, and therefore a feasible, effective and accurate control method is provided for the bus voltage control problem of the photovoltaic power station, active power data of the photovoltaic nodes, voltage deviation data of the buses and reactive power data of the photovoltaic nodes are obtained through calculation and are preprocessed, after active power data of photovoltaic nodes, voltage deviation data of buses and reactive power data of the photovoltaic nodes are preprocessed, the active power data of the photovoltaic nodes and the voltage deviation data of the buses are used as input of a bus voltage active power neural network, the reactive power data of the photovoltaic nodes and the voltage deviation data of the buses are used as input of a bus voltage reactive power neural network, the active power data of the photovoltaic nodes and the voltage deviation data of the buses are respectively input into corresponding neural networks for training, when errors respectively meet corresponding conditions, updating of a weight matrix is stopped, final output quantity is sent to a model prediction controller at a lower layer, a sensitivity coefficient of line voltage active power and a sensitivity coefficient of bus voltage reactive power can be obtained, a target function and the constraint of the model prediction controller are solved by a particle swarm algorithm, and the solved optimal control quantities of the active power and the reactive power are respectively sent to corresponding inverter terminal controllers, thereby solve the overvoltage, undervoltage problem of photovoltaic power plant generating line.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.

Claims (9)

1. A photovoltaic power station voltage control method based on a neural network is characterized in that: the control method comprises the following specific steps:
the method comprises the following steps: the bus voltage active power neural network and the bus voltage reactive power neural network respectively output the sensitivity coefficients of the corresponding bus voltage active power and the corresponding bus voltage reactive power to the model predictive controller;
step two: and the model prediction controller constructs a mathematical model of the bus voltage and the node injection active power and reactive power through the sensitivity coefficients of the bus voltage active power and the bus voltage reactive power issued by the upper-layer control framework.
2. The photovoltaic power station voltage control method based on the neural network as claimed in claim 1, wherein: the bus voltage active power neural network and the bus voltage reactive power neural network form an upper-layer control framework, and the model prediction controller is a component of a lower-layer control framework;
the bus voltage active power neural network comprises an input layer, a hidden layer and an output layer, input data of the input layer of the bus voltage active power neural network comprises active power data of nodes and voltage deviation data of buses, and output data of the output layer of the bus voltage active power neural network is a sensitivity coefficient of bus voltage active power of photovoltaic nodes.
3. The photovoltaic power station voltage control method based on the neural network as claimed in claim 2, wherein: the above-mentioned
The active power data of the node can be obtained by the following formula (1):
Figure FDA0002827720240000011
in the formula (1), UjAnd IjThe voltage and the current of the jth photovoltaic node are respectively marked by a conjugate operation, Re is an operation of an arithmetic unit, and the voltage deviation data of the bus can be obtained by an equation (2):
Figure FDA0002827720240000012
in the formula (2), the reaction mixture is,
Figure FDA0002827720240000013
and
Figure FDA0002827720240000014
the voltage and the reference voltage value of the ith photovoltaic power station bus are respectively.
4. The photovoltaic power station voltage control method based on the neural network as claimed in claim 2, wherein: the reactive power data of the node can be obtained by equation (3):
Figure FDA0002827720240000021
in the formula (3), UjAnd IjThe voltage and the current of the jth photovoltaic node are respectively marked by a conjugate operation, and Im is an imaginary part calculation.
5. The photovoltaic power station voltage control method based on the neural network as claimed in claim 1, wherein: the method comprises the following specific steps of obtaining the sensitivity coefficient of the active power of the bus voltage and the sensitivity coefficient of the reactive power of the bus voltage:
the method comprises the following steps: preprocessing active power data of the photovoltaic nodes, voltage deviation data of the bus and reactive power data of the photovoltaic nodes obtained through calculation, wherein the preprocessing modes are respectively shown as formulas (4), (5) and (6):
Figure FDA0002827720240000022
Figure FDA0002827720240000023
Figure FDA0002827720240000024
in the formula (4), Pj' and
Figure FDA0002827720240000025
respectively obtaining an active power normalized value and an active power rated value of the jth photovoltaic node; in the formula (5), the reaction mixture is,
Figure FDA0002827720240000026
and
Figure FDA0002827720240000027
respectively obtaining a voltage deviation normalized value and a maximum value of the voltage deviation of the ith bus; q 'in the formula (6)'jAnd
Figure FDA0002827720240000028
respectively obtaining a reactive power normalized value and a reactive power rated value of the jth photovoltaic node;
step two: after active power data of the photovoltaic nodes, voltage deviation data of the bus and reactive power data of the photovoltaic nodes are preprocessed, the active power data of the photovoltaic nodes and the voltage deviation data of the bus are used as input of a bus voltage active power neural network, the reactive power data of the photovoltaic nodes and the voltage deviation data of the bus are used as input of a bus voltage reactive power neural network, the input of the bus voltage reactive power neural network is respectively input into corresponding neural networks for training, and weight values of the corresponding neural networks are corrected according to a formula (7):
Figure FDA0002827720240000031
in the formula (7), Ep(k)、wp(k)、Δwp(k) Respectively an output quantity error value, a weight matrix and a weight updating quantity when the k-th weight of the bus voltage/active power neural network is updated, Eq(k)、wq(k)、Δwq(k) The output quantity error value, the weight matrix and the weight updating quantity at the time of the kth weight updating of the bus voltage/reactive power neural network are respectively, and the method comprises the following steps:
Figure FDA0002827720240000032
in the formula (8), Yp(k)、Y′p(k) Respectively the real value and the actual value of the output quantity during the k-th weight updating of the bus voltage/active power neural network, Yq(k)、Y′q(k) Respectively obtaining the real value and the actual value of the output quantity when the k-th weight of the bus voltage/reactive power neural network is updated;
step three: when the errors respectively meet corresponding conditions, the updating of the weight matrix is stopped, and the final output quantity is sent to the model prediction controller of the lower layer,
Figure FDA0002827720240000033
in the formula (9), γpAnd gammaqError thresholds are preset for the bus voltage/active power neural network and the bus voltage/reactive power neural network respectively, and there are:
Figure FDA0002827720240000034
6. the photovoltaic power station voltage control method based on the neural network as claimed in claim 1, wherein: the model prediction controller solves the corresponding optimal control quantity through real-time rolling optimization, an objective function of the model prediction controller is composed of a multi-objective optimization function, the multi-objective optimization function specifically comprises two groups of objective functions, one group is a bus voltage deviation value of a minimized photovoltaic power station, the other group is an active control cost of a minimized photovoltaic node, and the objective function is a first item of the multi-objective optimization function as shown in a formula (11):
Figure FDA0002827720240000041
in the formula (11), the reaction mixture is,
Figure FDA0002827720240000042
for the voltage deviation value of the kth step of the ith bus predicted based on the current time, NpAnd NbRespectively predicting the domain length and the total number of the photovoltaic power station buses;
the second term of the objective function of the lower model predictive controller is shown as the formula (12):
Figure FDA0002827720240000043
in the formula (12), Fj(k) Active control cost for kth step of jth photovoltaic node predicted based on current time, NpAnd NpvRespectively the predicted domain length and the total number of photovoltaic nodes, Fj(k) Is a quadratic function, as shown in equation (13):
Figure FDA0002827720240000044
in the formula (13), aj、bj、cjRespectively, the active control cost coefficient, delta P, of the jth photovoltaic nodej(k) The active control quantity of the kth step of the jth photovoltaic node predicted based on the current moment is obtained.
7. The photovoltaic power station voltage control method based on the neural network as claimed in claim 6, wherein: the objective function of the lower model predictive controller is formed by weighting objective functions in an equation (11) and an equation (12), specifically as shown in an equation (14), and is constrained as shown in equations (15) to (20):
min(λ1·Obj1+λ2·Obj2) (14)
Figure FDA0002827720240000045
Figure FDA0002827720240000051
Figure FDA0002827720240000052
Figure FDA0002827720240000053
Figure FDA0002827720240000054
Figure FDA0002827720240000055
in formula (14), λ1、λ2Are respectively the corresponding weight matrix, NcIs the control field length; formula (15)) And equation (16) represents the ramp constraints of the active power and reactive power of the jth photovoltaic node, respectively, wherein,
Figure FDA0002827720240000056
respectively carrying out climbing upper limit constraint, climbing lower limit constraint, reactive power climbing upper limit constraint and climbing lower limit constraint on active power of the jth photovoltaic node; equations (17) and (18) represent the upper and lower constraints on the active power and reactive power output of the jth photovoltaic node, respectively, wherein,
Figure FDA0002827720240000057
respectively are the output upper limit constraint, the output lower limit constraint, the output upper limit constraint and the output lower limit constraint of the active power, the output lower limit constraint, the reactive power of the jth photovoltaic node, Pj(t0)、Qj(t0) Respectively outputting active power and reactive power of the jth photovoltaic node at the current moment; equation (19) is the relationship between the bus node voltage change and the injection of active and reactive power into each photovoltaic node, wherein,
Figure FDA0002827720240000058
sensitivity coefficient for bus node voltage change/photovoltaic node injection active power, final output from bus voltage/active power neural network in upper layer control
Figure FDA0002827720240000059
Sensitivity coefficient for bus node voltage variation/photovoltaic node injection reactive power, final output from bus voltage/reactive power neural network in upper layer control
Figure FDA00028277202400000510
(according to the formula (10)); equation (20) represents the output active power constraint of the entire photovoltaic power plant, wherein,
Figure FDA00028277202400000511
and the active power reference output value of the whole photovoltaic power station comes from a power grid dispatching center.
8. The photovoltaic power station voltage control method based on the neural network as claimed in claim 1, wherein: the weight matrix of the objective function of the model predictive controller can be changed through the actual value of the photovoltaic power station bus voltage, and the multi-objective model is controlled through multiple modes, wherein the multiple modes refer to whether the photovoltaic bus voltage is in a normal operation range.
9. The photovoltaic power station voltage control method based on the neural network as claimed in claim 8, wherein: the multi-mode is that whether the photovoltaic bus voltage is in a normal operation range or not, and the specific judgment method is as follows:
Figure FDA0002827720240000061
in the formula (21), the compound represented by the formula,
Figure FDA0002827720240000062
is a reference voltage value, U, of the ith bus of the photovoltaic power stationi(t0) Is the actual voltage value, U, of the ith bus of the photovoltaic power station at the current momenti(t0) Is the safety margin coefficient of the voltage of the ith bus of the photovoltaic power station, sig (t)0) The modal indication function of the photovoltaic power station represents that the photovoltaic power station has over/under voltage when the value of the modal indication function is 1, and represents that the photovoltaic power station normally operates when the value of the modal indication function is 0. Lambda [ alpha ]1And λ2The value of (b) satisfies the following formula:
Figure FDA0002827720240000063
λ2＝I (23)
equation (22) represents the photovoltaic plant bus voltage exceeding its normal operationThe greater the extent of the row range, the corresponding diagonal matrix λ1The larger the element value of (A), the corresponding diagonal matrix lambda is if the photovoltaic power station normally operates1The lower layer model predictive controller has an objective function degenerated into a single objective function from a multi-objective function, namely, the optimization objective is to minimize the active output cost of the photovoltaic power station under the normal operation condition of the photovoltaic power station, wherein lambda is1(m) is a diagonal matrix lambda1The m-th element of (a) is,
Figure FDA0002827720240000064
is a diagonal matrix lambda1Dimension of (a), λεIs a weight coefficient; in the formula (23), I is an identity matrix, the objective function and the constraint of the model predictive controller are solved through a particle swarm algorithm, and the solved optimal control quantity of the active power and the reactive power
Figure FDA0002827720240000071
And
Figure FDA0002827720240000072
and respectively transmitting the signals to the corresponding inverter terminal controllers.
CN202011434418.5A 2020-12-10 2020-12-10 Photovoltaic power station voltage control method based on neural network Active CN112600244B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011434418.5A CN112600244B (en) 2020-12-10 2020-12-10 Photovoltaic power station voltage control method based on neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011434418.5A CN112600244B (en) 2020-12-10 2020-12-10 Photovoltaic power station voltage control method based on neural network

Publications (2)

Publication Number Publication Date
CN112600244A true CN112600244A (en) 2021-04-02
CN112600244B CN112600244B (en) 2023-04-28

Family

ID=75191424

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011434418.5A Active CN112600244B (en) 2020-12-10 2020-12-10 Photovoltaic power station voltage control method based on neural network

Country Status (1)

Country Link
CN (1) CN112600244B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114123200A (en) * 2022-01-24 2022-03-01 国网山西省电力公司晋城供电公司 Photovoltaic power station dynamic modeling method based on data driving and storage device
CN115912372A (en) * 2022-11-30 2023-04-04 国网四川省电力公司电力科学研究院 Voltage control method and system for distribution network with high-proportion distributed photovoltaic access

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0643948A (en) * 1992-07-27 1994-02-18 Tohoku Electric Power Co Inc Voltage reactive power controller
WO2015081771A1 (en) * 2013-12-06 2015-06-11 国家电网公司 Adaptive emergency control method for voltage security and stability based on synchronous measurement information
CN105281360A (en) * 2015-09-14 2016-01-27 国家电网公司 Distributed photovoltaic automatic generating control method based on sensitivity
CN109462254A (en) * 2018-11-16 2019-03-12 国网辽宁省电力有限公司电力科学研究院 A method of photovoltaic digestion capability is promoted based on voltage sensibility
CN110601252A (en) * 2019-06-18 2019-12-20 武汉大学 MPC-based feeder-level rapid voltage control method for distribution-type photovoltaic power distribution network
CN112018755A (en) * 2020-07-03 2020-12-01 国网浙江省电力有限公司电力科学研究院 Photovoltaic power distribution network reactive voltage prediction method and system based on cyclic neural network

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0643948A (en) * 1992-07-27 1994-02-18 Tohoku Electric Power Co Inc Voltage reactive power controller
WO2015081771A1 (en) * 2013-12-06 2015-06-11 国家电网公司 Adaptive emergency control method for voltage security and stability based on synchronous measurement information
CN105281360A (en) * 2015-09-14 2016-01-27 国家电网公司 Distributed photovoltaic automatic generating control method based on sensitivity
CN109462254A (en) * 2018-11-16 2019-03-12 国网辽宁省电力有限公司电力科学研究院 A method of photovoltaic digestion capability is promoted based on voltage sensibility
CN110601252A (en) * 2019-06-18 2019-12-20 武汉大学 MPC-based feeder-level rapid voltage control method for distribution-type photovoltaic power distribution network
CN112018755A (en) * 2020-07-03 2020-12-01 国网浙江省电力有限公司电力科学研究院 Photovoltaic power distribution network reactive voltage prediction method and system based on cyclic neural network

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114123200A (en) * 2022-01-24 2022-03-01 国网山西省电力公司晋城供电公司 Photovoltaic power station dynamic modeling method based on data driving and storage device
CN114123200B (en) * 2022-01-24 2022-04-12 国网山西省电力公司晋城供电公司 Photovoltaic power station dynamic modeling method based on data driving and storage device
CN115912372A (en) * 2022-11-30 2023-04-04 国网四川省电力公司电力科学研究院 Voltage control method and system for distribution network with high-proportion distributed photovoltaic access
CN115912372B (en) * 2022-11-30 2023-10-03 国网四川省电力公司电力科学研究院 Voltage control method and system for high-proportion distributed photovoltaic access distribution network

Also Published As

Publication number Publication date
CN112600244B (en) 2023-04-28

Similar Documents

Publication Publication Date Title
Mahmoud et al. The role of intelligent generation control algorithms in optimizing battery energy storage systems size in microgrids: A case study from Western Australia
CN106505635B (en) Active scheduling model and scheduling system with minimum wind abandon
CN108173293B (en) New energy active output maximum value on-line calculation method considering power grid operation constraint
CN113313613B (en) Power distribution network modularized movable battery energy storage MMBES optimal configuration method
CN111950900B (en) Power system source network load storage intrinsic safety risk assessment method
CN105160459A (en) Evaluation method for stable operation state of power system
CN112600244A (en) Photovoltaic power station voltage control method based on neural network
CN115545290A (en) Distributed energy storage economic optimization configuration method in power distribution network containing photovoltaic power generation
Reddy et al. Placement of distributed generator, capacitor and DG and capacitor in distribution system for loss reduction and reliability improvement
CN114430165A (en) Micro-grid group intelligent coordination control method and device based on depth model prediction
Jaleel et al. Reliability Evaluation of Electric Distribution Network with Distributed Generation Integrated
Wang et al. Optimal power flow in distribution networks with high penetration of photovoltaic units
Akbari-Zadeh et al. Dstatcom allocation in the distribution system considering load uncertainty
CN116581759A (en) Power transmission section active correction control method, system and equipment
Chen et al. An evaluation method of distributed generation credible capacity based on island partition
Xiao et al. Operation risk assessment of AC/DC hybrid distribution system based on two-point estimation method
Alzaareer et al. Voltage and congestion control in active distribution networks using fast sensitivity analysis
Shu et al. Probabilistic Power Flow Optimization of Source Network Load Storage in AC/DC Systems Based on SOC Characteristics
CN112182952A (en) Multi-objective optimization scheduling method for improving elasticity of power system
CN105811423A (en) Reactive automatic compensation method for microgrid system
CN114006415B (en) Evaluation method, system and medium for improving new energy admission capacity of distribution network
CN112446521B (en) Multi-objective planning method for wind power plant access system considering economy and safety
Xu et al. Accommodation Capacity Evaluation of Distributed PV Considering Security Constraints of Feeders and Substation Areas
CN116384625A (en) Friendly interactive rapid evaluation index system and calculation method for power distribution network
Jiahao et al. Energy Reports

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