CN115173491A - Distributed power coordination control method and device for power distribution network containing photo-thermal power generation - Google Patents
Distributed power coordination control method and device for power distribution network containing photo-thermal power generation Download PDFInfo
- Publication number
- CN115173491A CN115173491A CN202210739336.4A CN202210739336A CN115173491A CN 115173491 A CN115173491 A CN 115173491A CN 202210739336 A CN202210739336 A CN 202210739336A CN 115173491 A CN115173491 A CN 115173491A
- Authority
- CN
- China
- Prior art keywords
- power
- power generation
- generation unit
- active
- active power
- 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.)
- Pending
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
- H02J3/46—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
- H02J3/004—Generation forecast, e.g. methods or systems for forecasting future energy generation
-
- 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]
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a distributed power coordination control method and a distributed power coordination control device for a power distribution network containing photo-thermal power generation, wherein the method comprises the steps of utilizing an LSTM network to predict the power generation power of a light and heat power generation unit in the power distribution network in real time according to historical data of the light and heat power generation unit in the power distribution network to obtain the power prediction value of the light and heat power generation unit; based on the obtained power predicted value, the error correction is carried out on the predicted value by utilizing a deep neural network, so that the power prediction precision of the light and heat power generation unit is improved; classifying the power regulation potentials of the light and heat power generation units based on a fuzzy C-means clustering algorithm and the corrected power prediction value; the active power distribution function of the distribution network containing the photo-thermal power generation based on online rolling type optimization is designed, the active power of the photo-thermal power generation unit and the active power of the photo-thermal power generation unit are distributed optimally, and the active power is adjusted accurately. The invention solves the problems of poor regulation precision, low regulation speed and the like in the power regulation process of the conventional power distribution network containing photo-thermal power generation.
Description
Technical Field
The invention belongs to the technical field of power distribution network power coordination control, and particularly relates to a distributed power coordination control method and device for a power distribution network containing photo-thermal power generation.
Background
In 2020, the national energy strategic targets of 'carbon peak reaching and carbon neutralization' are clearly provided in China, a large amount of clean renewable energy is inevitably accessed into a power grid by low-carbon development supporting the energy field, and randomness, heterogeneity and volatility of the energy bring great challenges to the stable operation of a power system. However, active power adjustment of the current power distribution network mainly takes response to an Automatic Generation Control (AGC) instruction, and mainly issues an adjustment signal once, so that the adjustment precision is poor, the response speed is slow, the influence of fluctuation of new energy on the adjustment precision is not considered, the requirement of fine adjustment is difficult to complete, adverse effects are brought to fine Control of a power system, and how to accurately Control a power Generation unit in the power distribution network provides stable power support for the power system is an unavoidable research problem in the development process of the power system under a "dual-carbon" target.
Disclosure of Invention
The invention aims to provide a distributed power coordination control method and device for a power distribution network containing photo-thermal power generation, and aims to solve the problems of poor regulation precision, low regulation speed and the like in the power regulation process of the existing power distribution network containing photo-thermal power generation.
In order to solve the technical problems, the technical scheme adopted by the invention is as follows:
the invention provides a distributed power coordination control method for a power distribution network containing photo-thermal power generation, which comprises the following steps:
according to historical data of power generation units in the power distribution network containing photo-thermal power generation, the power generation power of the power generation units is predicted in real time, and power prediction values of the power generation units are obtained; the power generation unit comprises a photovoltaic power generation unit and a thermal power generation unit;
correcting the obtained power predicted value of the power generation unit to obtain an active power feedback correction value of the power generation unit;
clustering the obtained active power feedback correction value of the power generation unit;
determining a power generation unit participating in AGC according to the clustering result;
and performing distributed optimal coordination distribution on the active power of the power generation units participating in the AGC.
Further, the predicting the power generation power of the power generation unit in real time according to the historical data of the power generation unit in the power distribution network containing the photo-thermal power generation comprises:
acquiring historical active power data of the power generation unit for at least 7 days and corresponding historical weather data, and performing normalization processing on the acquired historical active power data;
establishing a power prediction model for each power generation unit based on an LSTM network, and training the power prediction model of the power generation unit by adopting the normalized historical active power data of the power generation unit to obtain a trained power prediction model corresponding to the power generation unit;
and inputting the active power of the power generation unit at the current moment into the corresponding trained power prediction model to obtain the power prediction value of the power generation unit at the next moment.
Further, the correcting the obtained power predicted value of the power generation unit to obtain an active power feedback correction value of the power generation unit includes:
constructing an error compensation model based on a deep neural network to obtain a corrected active power error value;
and superposing the active power predicted value of the power generation unit at the current moment and the obtained active power error value corrected by the power generation unit to obtain an active power feedback correction value of the power generation unit.
Further, the constructing the error compensation model based on the deep neural network includes:
training a deep neural network by using historical weather data and an active power prediction error value of the power generation unit as input to obtain an error compensation model; the active power prediction error value is the difference between historical active power data and an active power prediction value.
Further, the clustering the obtained active power feedback correction value of the power generation unit includes:
initializing a membership matrix, wherein matrix elements meet constraint conditions:
wherein, mu pj Representing the membership degree of the jth sample in the pth cluster, wherein c is the total number of clusters; the sample refers to an active power feedback correction value of the power generation unit;
calculating a clustering center:
wherein m ∈ [1, + ∞]As weighting coefficient, x j Represents the jth sample;
performing iterative computation according to an objective function:
wherein d is pj ＝||v p -x j I is the firstDistance of j samples to the class p center;
if the change quantity of the target function in two adjacent times is smaller than a preset iteration stop threshold epsilon, the iteration is finished, classification is carried out according to the following rule, otherwise, the membership degree matrix is updated, and the iteration is continued;
the classification rule is as follows: mu.s of pj ≤max μ 1j ,...,μ cj Then the sample j is determined to belong to the pth class.
Further, the membership matrix is updated as follows:
further, the determining the power generation units participating in the AGC according to the clustering result includes:
if cluster center v of class p p And if the value is less than the preset minimum regulation capacity threshold II, the corresponding power generation unit does not participate in the instruction response of the AGC, otherwise, the corresponding power generation unit participates in the instruction response of the AGC.
Further, the performing distributed optimal coordination distribution on the active power of the power generation units participating in the AGC includes:
for the power generation units participating in AGC, in each AGC scheduling period, the following active power regulation objective function is established, and active power output is regulated through online rolling type optimization:
wherein N is p For the number of scheduling periods, a is the total number of generating units actually participating in AGC power adjustment, phi i (t + l | t) is a cost function of the ith power generation unit;
the constraint conditions to be met by the active power regulation objective function are as follows:
Φ i (t+l|t)＝λ i (t+l|t)P i (t+l|t)；
wherein, P i (t + l | t) is the active power value at time t + l of the ith power generation unit predicted at time t,
the active power feedback correction value, Δ P, at time t + l for the ith power generation unit predicted for time t AGC Is the total active power regulating quantity, lambda, issued by AGC to the distribution network containing photo-thermal i (t + l | t) is a power distribution weight at the t + l time of the ith power generation unit predicted at the t time;
and continuously optimizing and solving the active power regulation objective function at each moment to obtain the optimal active power output value of the power generation unit, and issuing an optimal instruction to the power generation unit.
The invention provides a distributed power coordination control device for a power distribution network containing photo-thermal power generation, which comprises:
the prediction module is used for predicting the power generation power of the power generation unit in real time according to the historical data of the power generation unit in the power distribution network containing photo-thermal power generation to obtain the power prediction value of the power generation unit; the power generation unit comprises a photovoltaic power generation unit and a thermal power generation unit;
the correction module is used for correcting the acquired power predicted value of the power generation unit to obtain an active power feedback correction value of the power generation unit;
the clustering module is used for clustering the obtained active power feedback correction value of the power generation unit;
the screening module is used for determining the power generation units participating in AGC according to the clustering result;
and the distribution module is used for carrying out distributed optimal coordination distribution on the active power of the power generation units participating in the AGC.
Further, the correction module is specifically configured to,
training a deep neural network by using historical weather data and an active power prediction error value of the power generation unit as input to obtain an error compensation model; the active power prediction error value is the difference between historical active power data and an active power prediction value;
and inputting the weather data and the active power prediction error value of the power generation unit at the current moment into the trained error compensation model to obtain a corrected active power error value.
And superposing the active power predicted value of the power generation unit at the current moment and the obtained corrected active power error value of the power generation unit to obtain an active power feedback correction value of the power generation unit.
Further, the clustering module is specifically configured to,
initializing a membership matrix, wherein matrix elements meet constraint conditions:
wherein, mu pj Representing the membership degree of the jth sample in the pth cluster, wherein c is the total number of clusters; the sample refers to an active power feedback correction value of the power generation unit;
calculating a clustering center:
wherein m is ∈ [1, ∞ ]]As weighting coefficient, x j Represents the jth sample;
performing iterative computation according to an objective function:
wherein, d pj ＝||v p -x j The | | is the distance from the jth sample to the class p center;
if the change quantity of the target function in two adjacent times is smaller than a preset iteration stop threshold epsilon, the iteration is finished, classification is carried out according to the following rule, otherwise, the membership degree matrix is updated, and the iteration is continued;
the classification rule is as follows: mu.s of pj ≤max μ 1j ,...,μ cj If yes, judging the sample j to belong to the pth class;
the membership matrix is updated as follows:
furthermore, the screening module is specifically configured to,
if the cluster center v of the pth class p And if the regulation capacity is smaller than the preset minimum regulation capacity threshold pi, the corresponding power generation unit does not participate in the instruction response of the AGC, otherwise, the corresponding power generation unit participates in the instruction response of the AGC.
Furthermore, the distribution module is specifically configured to,
for the power generation units participating in AGC, in each AGC scheduling period, the following active power regulation objective function is established, and active power output is regulated through online rolling type optimization:
wherein N is p For the number of scheduling periods, a is the total number of generating units actually participating in AGC power adjustment, phi i (t + l | t) is a cost function of the ith power generation unit;
the constraint conditions to be met by the active power regulation objective function are as follows:
Φ i (t+l|t)＝λ i (t+l|t)P i (t+l|t)；
wherein, P i (t + l | t) is the active power value at time t + l of the ith power generation unit predicted at time t,
the active power feedback correction value, Δ P, at time t + l for the ith power generation unit predicted for time t AGC Is the total active power regulating quantity lambda issued by AGC to the distribution network containing photo-thermal i (t + l | t) is a power distribution weight at the t + l time of the ith power generation unit predicted at the t time;
and continuously optimizing and solving the active power regulation objective function at each moment to obtain the optimal active power output value of the power generation unit, and issuing an optimal instruction to the power generation unit.
The invention achieves the following beneficial technical effects:
the invention provides a distributed power coordination control method for a power distribution network containing photo-thermal power generation, which is used for estimating the power regulation potential of the power distribution network containing photo-thermal power generation based on a short-term real-time prediction algorithm and correcting errors of a predicted value by using a deep neural network, so that the problem of poor power regulation accuracy caused by uncertainty and volatility of new energy power generation is solved. In addition, in the real-time power distribution of the power distribution network containing photo-thermal power generation responding to the AGC instruction, the actual adjusting capacity of the light and heat power generation unit in the power distribution network containing photo-thermal power generation is considered to design a corresponding optimization objective function, and the accuracy and the economy of power adjustment of the power distribution network containing photo-thermal power generation are guaranteed.
Drawings
Fig. 1 is a flowchart of a distributed power coordination control method for a power distribution network including photo-thermal power generation according to an embodiment of the present invention.
Detailed Description
The invention is further described with reference to specific examples. The following examples are only for illustrating the technical solutions of the present invention more clearly, and the protection scope of the present invention is not limited thereby.
As mentioned above, the problems of poor regulation precision, low regulation speed and the like exist in the power regulation process of the existing power distribution network containing photo-thermal power generation.
In order to solve the technical problem, the invention provides a distributed power coordination control method for a power distribution network containing photo-thermal power generation, which comprises the following steps:
according to historical data of a power generation unit in a power distribution network containing photo-thermal power generation, the power generation power of the power generation unit is predicted in real time, and a power prediction value of the power generation unit is obtained; the power generation unit comprises a photovoltaic power generation unit and a thermal power generation unit;
correcting the obtained power predicted value of the power generation unit to obtain an active power feedback correction value of the power generation unit;
clustering the obtained active power feedback correction value of the power generation unit;
determining power generation units participating in AGC according to the clustering result;
and performing distributed optimal coordination distribution on the active power of the power generation units participating in the AGC.
An embodiment of the invention provides a distributed power coordination control method for a power distribution network including photo-thermal power generation, as shown in fig. 1, which is implemented as follows:
the method comprises the following steps: according to historical data of light and heat power generation units in the power distribution network containing photo-thermal power generation, the generated power of the light and heat power generation units is predicted in real time by using an LSTM network, and power predicted values of the light and heat power generation units are obtained;
step two: based on the power predicted values of the light and heat power generation units obtained in the first step, carrying out error correction on the power predicted values by using a deep neural network to obtain active power feedback correction values of the light and heat power generation units, so as to improve the power prediction accuracy of the light and heat power generation units;
step three: clustering the active power feedback correction values of the light and heat power generation units obtained in the second step by adopting fuzzy C-means clustering;
step four: and determining the light and heat power generation units participating in the AGC according to the clustering result, and performing distributed optimal coordination distribution on the active power of the light and heat power generation units participating in the AGC.
In this embodiment, the LSTM network is used to predict the generated power in real time to obtain the predicted power value of the optical and thermal power generation units, and the specific implementation process is as follows:
step 1-1: respectively acquiring historical active power data and corresponding historical weather data of the light and heat power generation units for at least 7 days;
step 1-2: carrying out normalization processing on the acquired historical active power data, as shown in formula (1):
in the formula (1), x i Historical active power data for the ith light or thermal power generation unit, x min Is the minimum value, x, of the historical active power data of the ith light or thermal power generation unit max Is the maximum value, x, of the historical active power data of the ith light or thermal power generation unit ni Representing normalized historical active power data for the ith light or thermal power generation unit.
Step 1-3: based on the historical active power data normalized in the step 1-2, establishing a generated power prediction model of the ith light or heat generating unit based on an LSTM network, wherein functions involved in the training process of the LSTM network are as follows:
f t ＝sigmoid(W f ·h t-1 +W f ·x t +b f ) (2)
i t ＝sigmoid(W i ·h t-1 +W i ·x t +b i ) (3)
o t ＝sigmoid(W o ·h t-1 +W o ·x t +b o ) (6)
h t ＝o t ⊙tanh(C t ) (7)
wherein f is t 、i t 、o t Output signals of a forgetting gate, an input gate and an output gate of the neuron are respectively represented;
representing neuron candidate state information; c t Representing neuron state information; h is a total of t Representing a hidden state of a neuron; sigmoid () is a type S activation function; tanh () represents a hyperbolic tangent activation function; w f 、W i 、W C 、W o Respectively representing the weights of the forgetting gate, the input gate, the memory unit and the output gate; b f 、b i 、b C 、b o Respectively representing the offset of the forgetting gate, the input gate, the memory unit and the output gate; x is the number of t Representing a neuron input sequence value; an indication of a point-by-point multiplication operation.
The training loss function of the LSTM network is shown as follows:
in the formula (8), n represents the number of predicted results, y i Representing the normalized active power of the ith light or heat generating unit,
and represents the active power predicted value of the ith light or heat generating unit.
And (3) training a generated power prediction model of the ith light or heat generating unit according to the formulas (2) to (8) to obtain a power prediction model without feedback correction.
And inputting the active power of the light or heat power generation unit at the current moment into the corresponding trained prediction model to obtain the power prediction value at the next moment.
In this embodiment, the error correction is performed on the power prediction value by using the deep neural network, and the specific implementation process is as follows:
step 2-1: and constructing an error compensation model based on the deep neural network based on the historical active power data and the historical weather data acquired in the step 1-1 and the active power predicted value of the light or heat power generation unit in the step 1-3.
Specifically, the input data of the deep neural network is: historical weather data of the ith light or heat generating unit and active power actual prediction error data of the ith light or heat generating unit. Wherein the actual prediction error data of the active power of the ith light or thermal power generation unit is obtained by:
in the formula (9), e i Actual prediction error data, y, for the active power of the ith light or thermal power generation unit i A real value of active power representing the ith light or heat generating unit;
and the predicted value of the active power of the ith light or heat generating unit obtained by the prediction model is shown.
Step 2-2: the weight values of the deep neural network are corrected according to equation (10):
in formula (10), E i (k)、w i (k)、Δw i (k) The k-th weight of the deep neural networkThe error value, the weight matrix and the weight updating amount are calculated in a new time, and the error value is calculated as follows:
E i (k)＝Y i (k)-Y i '(k)
(11)
in formula (11), Y i (k)、Y i ' (k) are respectively an active power error value and an actual prediction error value output when the k-th weight of the deep neural network is updated.
And training by adopting historical data to obtain an error compensation model based on the deep neural network.
And inputting weather data and an actual active power prediction error value of the light or heat power generation unit at the current moment into a corresponding trained error compensation model to obtain a corrected active power error value.
Step 2-3: superposing the generated power predicted value of the ith optical or thermal power generation unit obtained based on the step 1-3 and the active power error value of the ith optical or thermal power generation unit obtained based on the step 2-2 to obtain an active power feedback correction value of the ith optical or thermal power generation unit:
in formula (12), Y i The correction value is fed back for the active power of the ith light or heat generating unit.
In this embodiment, the fuzzy C-means clustering is adopted to cluster the obtained active power feedback correction values of the light and the thermal power generation unit, and the specific implementation process is as follows:
step 3-1: initializing a membership matrix S, wherein matrix elements meet constraint conditions:
wherein, mu pj Representing the membership degree of the jth sample in the pth cluster, wherein c is the total number of the clusters; the sample refers to an active power feedback correction value of the light or heat power generation unit;
step 3-2: calculating a clustering center:
wherein m belongs to [1, infinity ] is a weighting coefficient;
step 3-3: performing iterative computation according to an objective function:
wherein d is pj ＝||v p -x j And | is the distance from the jth sample to the class p center.
Step 3-4: setting an iteration stop threshold epsilon, if the change quantity of the target function of two adjacent times is smaller than epsilon, ending the iteration and turning to 3-6, otherwise, turning to the step 3-5;
step 3-5: updating the matrix S and returning to the step 3-2;
step 3-6: classifying according to a judgment rule: mu.s of pj ≤max μ 1j ,...,μ cj Then the sample j is determined to belong to the pth class.
In this embodiment, determining the light and heat power generation units participating in AGC according to the clustering result includes:
screening the clustering samples obtained in the third step according to a preset minimum adjusting capacity threshold value pi, and if the clustering center v of the pth class p And if the threshold pi is less than the threshold pi, the corresponding light or heat power generation unit does not participate in the instruction response process of the AGC, otherwise, the corresponding light or heat power generation unit participates in the AGC. And the light or heat power generation unit with strong regulation capability and large regulation capacity is utilized to the maximum extent to carry out power regulation control.
In this embodiment, the distributed optimal coordination and allocation of the active power of the optical and thermal power generation units participating in AGC includes:
step 4-1: for the light and heat power generation units participating in AGC, an active power regulation objective function is established in each AGC scheduling period, and the active power output is optimally regulated in an online rolling type mode:
in the formula (18), N p For the number of dispatching cycles, a is the total number of light or heat generating units actually participating in AGC power regulation, phi i (t + l | t) is a cost function of the ith light or heat generating unit.
The constraints of the objective function (18) are:
Φ i (t+l|t)＝λ i (t+l|t)P i (t+l|t) (19)
wherein, P i (t + l | t) is the real power value at the t + l moment of the ith light or heat generating unit predicted by the prediction model at the t moment,
feedback correction value of active power at time t + l of the ith light or heat generating unit obtained for time t i min (t + l | t) is the minimum value of the active power at the moment t + l of the ith light or heat generating unit, Δ P AGC Is the total active power regulating quantity, lambda, issued by AGC to the distribution network containing photo-thermal i (t + l | t) is the power distribution weight at time t + l of the ith light or thermal power generation unit predicted at time t.
λ i (t + l | t) is determined by:
λ i the (t + l | t) is related to the actual adjusting capability of the light and heat power generation power supply, namely, the larger the residual adjustable capability of the power generation unit is, the larger the coefficient is, the more the power generation unit is prone to participate in the power adjustment of the AGC.
Step 4-2: the active power of the light or heat power generation unit is accurately adjusted by continuously optimizing and solving the objective function (18) at each moment and sending an optimal instruction to the light or heat power generation unit.
According to the distributed power coordination control method for the power distribution network with the photo-thermal power generation, the power regulation potential of the power distribution network with the photo-thermal power generation can be estimated based on a short-term real-time prediction algorithm, the error correction is carried out on the predicted value by using the deep neural network, and the problem of poor power regulation accuracy caused by uncertainty and volatility of new energy power generation is solved. In addition, in the real-time power distribution of the power distribution network containing photo-thermal power generation responding to the AGC instruction, the actual adjusting capacity of the photo-thermal power generation unit and the actual adjusting capacity of the thermal power generation unit are considered to design a corresponding optimization objective function, and the accuracy and the economy of power adjustment of the power distribution network containing photo-thermal power generation are guaranteed.
Another embodiment of the present invention provides a distributed power coordination control apparatus for a power distribution network including a photo-thermal power generation system, including:
the prediction module is used for predicting the power generation power of the power generation unit in real time according to the historical data of the power generation unit in the power distribution network containing photo-thermal power generation to obtain the power prediction value of the power generation unit; the power generation unit comprises a photovoltaic power generation unit and a thermal power generation unit;
the correction module is used for correcting the obtained power predicted value of the power generation unit to obtain an active power feedback correction value of the power generation unit;
the clustering module is used for clustering the obtained active power feedback correction value of the power generation unit;
the screening module is used for determining the power generation units participating in AGC according to the clustering result;
and the distribution module is used for carrying out distributed optimal coordination distribution on the active power of the power generation units participating in the AGC.
In this embodiment, the calibration module is specifically configured to,
training a deep neural network by using historical weather data and an active power prediction error value of the power generation unit as input to obtain an error compensation model; the active power prediction error value is the difference between historical active power data and an active power prediction value;
and inputting the weather data and the active power prediction error value of the power generation unit at the current moment into the trained error compensation model to obtain a corrected active power error value.
And superposing the active power predicted value of the power generation unit at the current moment and the obtained corrected active power error value of the power generation unit to obtain an active power feedback correction value of the power generation unit.
In this embodiment, the clustering module is specifically configured to,
initializing a membership matrix, wherein matrix elements meet constraint conditions:
wherein, mu pj Representing the membership degree of the jth sample in the pth cluster, wherein c is the total number of clusters; the sample refers to an active power feedback correction value of the power generation unit;
calculating a clustering center:
wherein m ∈ [1, + ∞]As weighting coefficient, x j Represents the jth sample;
performing iterative computation according to an objective function:
wherein d is pj ＝||v p -x j The | | is the distance from the jth sample to the class p center;
if the change quantity of the target function in two adjacent times is smaller than a preset iteration stop threshold epsilon, the iteration is finished, classification is carried out according to the following rule, otherwise, the membership degree matrix is updated, and the iteration is continued;
the classification rule is as follows: mu.s of pj ≤max μ 1j ,...,μ cj If yes, judging the sample j to belong to the pth class;
the membership matrix is updated as follows:
in this embodiment, the screening module is specifically configured to,
if the cluster center v of the pth class p And if the value is less than the preset minimum regulation capacity threshold II, the corresponding power generation unit does not participate in the instruction response of the AGC, otherwise, the corresponding power generation unit participates in the instruction response of the AGC.
In this embodiment, the allocation module is specifically configured to,
for the power generation units participating in AGC, in each AGC scheduling period, the following active power regulation objective function is established, and active power output is regulated through online rolling type optimization:
wherein, N p For the number of scheduling periods, a is the total number of generating units actually participating in AGC power adjustment, phi i (t + l | t) is a cost function of the ith power generation unit;
the constraint conditions to be met by the active power regulation objective function are as follows:
Φ i (t+l|t)＝λ i (t+l|t)P i (t+l|t)；
wherein, P i (t + l | t) is the active power value at time t + l of the ith power generation unit predicted at time t,
the active power feedback correction value, Δ P, at time t + l for the ith power generation unit predicted for time t AGC Is the total active power regulating quantity lambda issued by AGC to the distribution network containing photo-thermal i (t + l | t) is a power distribution weight at the t + l time of the ith power generation unit predicted at the t time;
and continuously optimizing and solving the active power regulation objective function at each moment to obtain the optimal active power output value of the power generation unit, and issuing an optimal instruction to the power generation unit.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting the same, and although the present invention is described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications and equivalents may be made to the embodiments of the invention without departing from the spirit and scope of the invention, which is to be covered by the claims.
Claims (13)
1. A distributed power coordination control method for a power distribution network containing photo-thermal power generation is characterized by comprising the following steps:
according to historical data of power generation units in the power distribution network containing photo-thermal power generation, the power generation power of the power generation units is predicted in real time, and power prediction values of the power generation units are obtained; the power generation unit comprises a photovoltaic power generation unit and a thermal power generation unit;
correcting the obtained power predicted value of the power generation unit to obtain an active power feedback correction value of the power generation unit;
clustering the obtained active power feedback correction value of the power generation unit;
determining power generation units participating in AGC according to the clustering result;
and performing distributed optimal coordination distribution on the active power of the power generation units participating in the AGC.
2. The distributed power coordination control method for the power distribution network containing the photo-thermal power generation system as claimed in claim 1, wherein the step of predicting the power generation power of the power generation unit in real time according to the historical data of the power generation unit in the power distribution network containing the photo-thermal power generation system comprises:
acquiring historical active power data of the power generation unit for at least 7 days and corresponding historical weather data, and performing normalization processing on the acquired historical active power data;
establishing a power prediction model for each power generation unit based on an LSTM network, and training the power prediction model of the power generation unit by adopting the normalized historical active power data of the power generation unit to obtain a trained power prediction model corresponding to the power generation unit;
and inputting the active power of the power generation unit at the current moment into the corresponding trained power prediction model to obtain the power prediction value of the power generation unit at the next moment.
3. The distributed power coordination control method for the power distribution network with the photo-thermal power generation system according to claim 2, wherein the step of correcting the obtained predicted power value of the power generation unit to obtain the feedback corrected active power value of the power generation unit includes:
constructing an error compensation model based on a deep neural network to obtain a corrected active power error value;
and superposing the active power predicted value of the power generation unit at the current moment and the obtained active power error value corrected by the power generation unit to obtain an active power feedback correction value of the power generation unit.
4. The distributed power coordination control method for the power distribution network with the photo-thermal power generation system according to claim 3, wherein the building of the error compensation model based on the deep neural network comprises:
training a deep neural network by using historical weather data and an active power prediction error value of the power generation unit as input to obtain an error compensation model; the active power prediction error value is the difference between historical active power data and an active power prediction value.
5. The distributed power coordination control method for the power distribution network with the photo-thermal power generation system according to claim 1, wherein the clustering the obtained active power feedback correction values of the power generation units comprises:
initializing a membership matrix, wherein matrix elements meet constraint conditions:
wherein, mu pj Representing the membership degree of the jth sample in the pth cluster, wherein c is the total number of clusters; the sample refers to an active power feedback correction value of the power generation unit;
calculating a clustering center:
wherein m ∈ [1, + ∞]As weighting coefficients, x j Represents the jth sample;
performing iterative computation according to an objective function:
wherein d is pj ＝||v p -x j The | | is the distance from the jth sample to the center of the p-type class;
if the variation of the target functions of two adjacent times is smaller than a preset iteration stop threshold epsilon, the iteration is finished, the classification is carried out according to the following rule, otherwise, the membership matrix is updated, and the iteration is continued;
the classification rule is as follows: mu.s of pj ≤maxμ 1j ,...,μ cj Then the sample j is determined to belong to the pth class.
6. The distributed power coordination control method for the power distribution network containing the photo-thermal power generation system as claimed in claim 5, wherein the membership matrix is updated as follows:
7. the distributed power coordination control method for the power distribution network with the photo-thermal power generation function according to claim 5, wherein the determining of the power generation units participating in the AGC according to the clustering result comprises:
if the cluster center v of the pth class p And if the regulation capacity is smaller than the preset minimum regulation capacity threshold pi, the corresponding power generation unit does not participate in the instruction response of the AGC, otherwise, the corresponding power generation unit participates in the instruction response of the AGC.
8. The distributed power coordination control method for the power distribution network containing the photo-thermal power generation system according to claim 7, wherein the performing distributed optimal coordination distribution on the active power of the power generation units participating in the AGC comprises:
for the power generation units participating in AGC, in each AGC scheduling period, the following active power regulation objective function is established, and active power output is regulated through online rolling type optimization:
wherein N is p For the number of scheduling periods, a is the total number of generating units actually participating in AGC power adjustment, phi i (t + l | t) is a cost function of the ith power generation unit;
the constraint conditions to be met by the active power regulation objective function are as follows:
Φ i (t+l|t)＝λ i (t+l|t)P i (t+l|t)；
wherein, P i (t + l | t) is the active power value at time t + l of the ith power generation unit predicted at time t,
the active power feedback correction value, Δ P, at time t + l for the ith power generation unit predicted for time t AGC Is the total active power regulating quantity, lambda, issued by AGC to the distribution network containing photo-thermal i (t + l | t) is the predicted power distribution weight of the ith power generation unit at the t moment and at the t + l moment;
and continuously optimizing and solving the active power regulation objective function at each moment to obtain the optimal active power output value of the power generation unit, and issuing an optimal instruction to the power generation unit.
9. The utility model provides a distribution network distributed power coordinated control device who contains solar-thermal power generation which characterized in that includes:
the prediction module is used for predicting the power generation power of the power generation unit in real time according to the historical data of the power generation unit in the power distribution network containing photo-thermal power generation to obtain the power prediction value of the power generation unit; the power generation unit comprises a photovoltaic power generation unit and a thermal power generation unit;
the correction module is used for correcting the acquired power predicted value of the power generation unit to obtain an active power feedback correction value of the power generation unit;
the clustering module is used for clustering the obtained active power feedback correction value of the power generation unit;
the screening module is used for determining the power generation units participating in AGC according to the clustering result;
and the distribution module is used for carrying out distributed optimal coordination distribution on the active power of the power generation units participating in AGC.
10. The distributed power coordination control method for distribution network including photo-thermal power generation according to claim 9, wherein said calibration module is specifically configured to,
training a deep neural network by using historical weather data and an active power prediction error value of the power generation unit as input to obtain an error compensation model; the active power prediction error value is the difference between historical active power data and an active power prediction value;
and inputting the weather data and the active power prediction error value of the power generation unit at the current moment into the trained error compensation model to obtain a corrected active power error value.
And superposing the active power predicted value of the power generation unit at the current moment and the obtained corrected active power error value of the power generation unit to obtain an active power feedback correction value of the power generation unit.
11. The distributed power coordination control device for distribution network including photo-thermal power generation according to claim 9, wherein said clustering module is specifically configured to,
initializing a membership matrix, wherein matrix elements meet constraint conditions:
wherein, mu pj Representing the membership degree of the jth sample in the pth cluster, wherein c is the total number of clusters; the sample refers to an active power feedback correction value of the power generation unit;
calculating a clustering center:
wherein m ∈ [1, + ∞]As weighting coefficient, x j Represents the jth sample;
performing iterative computation according to an objective function:
wherein d is pj ＝||v p -x j The | | is the distance from the jth sample to the class p center;
if the change quantity of the target function in two adjacent times is smaller than a preset iteration stop threshold epsilon, the iteration is finished, classification is carried out according to the following rule, otherwise, the membership degree matrix is updated, and the iteration is continued;
the classification rule is as follows: mu.f pj ≤maxμ 1j ,...,μ cj Then judging that the sample j belongs to the p-th class;
the membership matrix is updated as follows:
12. the distributed power coordination control device for distribution network including solar thermal power generation according to claim 11, wherein said screening module is specifically configured to,
if the cluster center v of the pth class p And if the value is less than the preset minimum regulation capacity threshold II, the corresponding power generation unit does not participate in the instruction response of the AGC, otherwise, the corresponding power generation unit participates in the instruction response of the AGC.
13. The distributed power coordination control device for distribution network including solar thermal power generation according to claim 12, wherein said distribution module is specifically configured to,
for the power generation units participating in AGC, in each AGC scheduling period, the following active power regulation objective function is established, and active power output is regulated through online rolling type optimization:
wherein, N p For the number of scheduling cycles, a is the total number of generating units actually participating in AGC power adjustment, phi i (t + l | t) is a cost function of the ith power generation unit;
the constraint conditions to be met by the active power regulation objective function are as follows:
Φ i (t+l|t)＝λ i (t+l|t)P i (t+l|t)；
wherein, P i (t + l | t) is the real power value at the t + l time of the ith power generation unit predicted at the t time,
the active power feedback correction value, Δ P, at time t + l for the ith power generation unit predicted for time t AGC Is issued by AGCTotal active power regulation, lambda, including photo-thermal power distribution network i (t + l | t) is a power distribution weight at the t + l time of the ith power generation unit predicted at the t time;
and continuously optimizing and solving the active power regulation objective function at each moment to obtain the optimal active power output value of the power generation unit, and issuing an optimal instruction to the power generation unit.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210739336.4A CN115173491A (en) | 2022-06-28 | 2022-06-28 | Distributed power coordination control method and device for power distribution network containing photo-thermal power generation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210739336.4A CN115173491A (en) | 2022-06-28 | 2022-06-28 | Distributed power coordination control method and device for power distribution network containing photo-thermal power generation |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115173491A true CN115173491A (en) | 2022-10-11 |
Family
ID=83487672
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210739336.4A Pending CN115173491A (en) | 2022-06-28 | 2022-06-28 | Distributed power coordination control method and device for power distribution network containing photo-thermal power generation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115173491A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110867902A (en) * | 2019-10-15 | 2020-03-06 | 东北大学 | Power generation prediction-based micro-grid distributed power supply de-centering optimized operation method |
| CN112217232A (en) * | 2020-09-29 | 2021-01-12 | 浙江中光新能源科技有限公司 | Photovoltaic and photo-thermal coupling power generation system and power generation control method |
| CN113131532A (en) * | 2021-04-16 | 2021-07-16 | 南京邮电大学 | New energy power station primary frequency modulation networked control method |
| CN113675855A (en) * | 2021-08-23 | 2021-11-19 | 南京理工大学 | Dynamic voltage model prediction distributed control method under double-layer architecture |
| CN114336592A (en) * | 2021-11-24 | 2022-04-12 | 华能新能源股份有限公司 | Wind power plant AGC control method based on model predictive control |
| CN114429233A (en) * | 2020-10-14 | 2022-05-03 | 许昌许继风电科技有限公司 | Cluster algorithm-based station-level optimal power control method and system |
-
2022
- 2022-06-28 CN CN202210739336.4A patent/CN115173491A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110867902A (en) * | 2019-10-15 | 2020-03-06 | 东北大学 | Power generation prediction-based micro-grid distributed power supply de-centering optimized operation method |
| CN112217232A (en) * | 2020-09-29 | 2021-01-12 | 浙江中光新能源科技有限公司 | Photovoltaic and photo-thermal coupling power generation system and power generation control method |
| CN114429233A (en) * | 2020-10-14 | 2022-05-03 | 许昌许继风电科技有限公司 | Cluster algorithm-based station-level optimal power control method and system |
| CN113131532A (en) * | 2021-04-16 | 2021-07-16 | 南京邮电大学 | New energy power station primary frequency modulation networked control method |
| CN113675855A (en) * | 2021-08-23 | 2021-11-19 | 南京理工大学 | Dynamic voltage model prediction distributed control method under double-layer architecture |
| CN114336592A (en) * | 2021-11-24 | 2022-04-12 | 华能新能源股份有限公司 | Wind power plant AGC control method based on model predictive control |
Non-Patent Citations (1)
| Title |
|---|
| "基于MPC的风电场有功分配策略研究", 中国优秀硕士学位论文全文数据库 工程科技II辑, no. 6, pages 042 - 788 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11581740B2 (en) | Method, system and storage medium for load dispatch optimization for residential microgrid | |
| CN111222674A (en) | Short-term photovoltaic power generation capacity prediction method based on long-term and short-term memory neural network | |
| CN116187601B (en) | Comprehensive energy system operation optimization method based on load prediction | |
| CN105678404A (en) | Micro-grid load prediction system and method based on electricity purchased on-line and dynamic correlation factor | |
| CN109919658A (en) | A kind of duty control method and system based on game theory | |
| CN111598225B (en) | Air conditioner cold load prediction method based on self-adaptive deep confidence network | |
| CN112182709A (en) | Rapid prediction method for let-down water temperature of large-scale reservoir stop log door layered water taking facility | |
| CN115759415A (en) | Power consumption demand prediction method based on LSTM-SVR | |
| Fan et al. | Multi-objective LSTM ensemble model for household short-term load forecasting | |
| CN115470862A (en) | Dynamic self-adaptive load prediction model combination method | |
| CN113487064A (en) | Photovoltaic power prediction method and system based on principal component analysis and improved LSTM | |
| CN115795992A (en) | Park energy Internet online scheduling method based on virtual deduction of operation situation | |
| CN111815039A (en) | Weekly scale wind power probability prediction method and system based on weather classification | |
| CN114240687A (en) | Energy hosting efficiency analysis method suitable for comprehensive energy system | |
| CN116865343B (en) | Model-free self-adaptive control method, device and medium for distributed photovoltaic power distribution network | |
| Sun et al. | Research on prediction of housing prices based on GA-PSO-BP neural network model: evidence from Chongqing, China | |
| CN116880201A (en) | Water network channel state control system based on fuzzy neural network | |
| CN117291069A (en) | LSTM sewage water quality prediction method based on improved DE and attention mechanism | |
| CN115173491A (en) | Distributed power coordination control method and device for power distribution network containing photo-thermal power generation | |
| CN116402209A (en) | Three gorges reservoir dam front drift prediction method based on improved neural network model | |
| CN116681154A (en) | Photovoltaic power calculation method based on EMD-AO-DELM | |
| CN115689001A (en) | Short-term load prediction method based on pattern matching | |
| CN115097886A (en) | Method, system, equipment and medium for controlling concentration of dissolved oxygen in sewage treatment | |
| CN114861555A (en) | Regional comprehensive energy system short-term load prediction method based on Copula theory | |
| CN115374998A (en) | Load prediction method based on attention mechanism and CNN-BiGRU |
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 |