CN109687456B - Scheduling method and system of electric power natural gas system - Google Patents
Scheduling method and system of electric power natural gas system Download PDFInfo
- Publication number
- CN109687456B CN109687456B CN201910072171.8A CN201910072171A CN109687456B CN 109687456 B CN109687456 B CN 109687456B CN 201910072171 A CN201910072171 A CN 201910072171A CN 109687456 B CN109687456 B CN 109687456B
- Authority
- CN
- China
- Prior art keywords
- time period
- load data
- output power
- power
- granulated
- 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.)
- Active
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
-
- 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
-
- 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)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a scheduling method and system of an electric power natural gas system. The method comprises the following steps: acquiring predicted load data of scheduling days; carrying out information granulation processing on the predicted load data to obtain granulated load data; calculating corresponding optimal output power according to the granulated load data; correcting the optimal output power by using historical load data to obtain corrected output power; and scheduling the output of each generator and the gas well according to the corrected output power. The scheduling method and system of the power and natural gas system can improve scheduling efficiency.
Description
Technical Field
The invention relates to the field of optimal configuration of an integrated energy system, in particular to a scheduling method and system of an electric power and natural gas system.
Background
In recent years, due to the gradual depletion of fossil fuels and the problem of environmental pollution caused by the depletion, the industrial model based on the large-scale utilization of fossil fuels, which is laid down in the second industrial revolution, is moving to the end, and the third industrial revolution represented by new energy technologies and internet technologies is being started. The energy internet tries to combine a renewable energy technology and an internet technology, promotes the mutual fusion of various complex network systems, and achieves the purposes of changing an energy utilization mode and promoting economic and social sustainable development.
To mitigate global warming threats, emission reduction is an irreversible trend in sustainable development of power systems. Among the various low carbon technologies, the use of gas turbines and the utilization of power to natural gas (P2G) technologies play an important role in reducing emissions, and they are increasing the interdependence between power and natural gas systems.
First, modern power systems are certainly one of the most complex industrial systems of human-supplied horsepower, both in scale and in architecture, and as power systems are scaled further, higher demands are placed on system speed and system safety. Secondly, as gas turbines are used and P2G technology matures, the coupling of power and natural gas systems makes the interconnection system more complex and the demand for scheduling efficiency of the power and natural gas coupled system higher.
Disclosure of Invention
The invention aims to provide a scheduling method and a scheduling system of a power and natural gas system, which improve scheduling efficiency.
In order to achieve the purpose, the invention provides the following scheme:
a scheduling method of an electric power natural gas system comprises the following steps:
acquiring predicted load data of scheduling days;
carrying out information granulation processing on the predicted load data to obtain granulated load data;
calculating corresponding optimal output power according to the granulated load data;
correcting the optimal output power by using historical load data to obtain corrected output power;
and scheduling the output of each generator and the gas well according to the corrected output power.
Optionally, the performing information granulation processing on the predicted load data to obtain granulated load data specifically includes:
and performing information granulation processing on the time of the predicted load data according to the similarity of the predicted load data at adjacent moments to obtain the granulated load data of each time period.
Optionally, the correcting the optimal output power by using the historical load data to obtain the corrected output power specifically includes:
acquiring load data of a previous time period of each time period;
inputting the load data of each granulated time period and the load data of the previous time period of each time period into a trained power difference correction model to obtain an output power correction value of each time period;
and adding the output power of each time period under the optimized power flow after granulation to the output power correction value of each time period to obtain the optimized output power of each time period.
Optionally, the training process of the power difference correction model includes:
acquiring historical load data of historical scheduling days;
calculating the output power before granulation at each moment by using an optimal power flow calculation method according to the historical load data;
performing information granulation processing on the time of the historical load data according to the similarity of the historical load data at adjacent moments to obtain the granulated historical load data of each time period;
calculating the output power of each granulated time period by using an optimal power flow calculation method according to the historical load data of each granulated time period;
the output power of each time period after granulation is differed with the output power of the corresponding time period before granulation at each moment to obtain the output power difference value of each time period;
the historical load data of each granulated time period is differenced with the historical load data of the corresponding time period to obtain a load difference value of each time period;
and training the least square support vector machine by taking the load difference value of each time period and the load data of the previous time period of each time period as input and the corresponding output power difference value of each time period as output to obtain a trained power difference value correction model.
The invention also discloses a dispatching system of the power and natural gas system, which comprises the following components:
the scheduling date data acquisition module is used for acquiring predicted load data of scheduling dates;
the granulation processing module is used for carrying out information granulation processing on the predicted load data to obtain granulated load data;
the power calculation module is used for calculating corresponding optimal output power aiming at the granulated load data;
the correction module is used for correcting the optimal output power by utilizing historical load data to obtain corrected output power;
and the scheduling module is used for scheduling the output of each generator and each gas well according to the corrected output power.
Optionally, the granulation treatment module includes:
and the first granulation processing unit is used for carrying out information granulation processing on the time of the predicted load data according to the similarity of the predicted load data at the adjacent time to obtain the granulated load data of each time slot.
Optionally, the correction module includes:
a previous-period data acquiring unit configured to acquire load data of a previous period of each period;
the model prediction unit is used for inputting the load data of each granulated time period and the load data of the previous time period of each time period into the trained power difference correction model to obtain an output power correction value of each time period;
and the correction unit is used for adding the output power of each time period under the optimized power flow after granulation to the output power correction value of each time period to obtain the optimized output power of each time period.
Optionally, the scheduling system further includes a model training module, where the model training module is configured to train the power difference correction model; the model training module comprises:
a historical data acquisition unit for acquiring historical load data of historical scheduling days;
the pre-granulation power calculation unit is used for calculating the output power before granulation at each moment by using an optimal power flow calculation method according to the historical load data;
the second granulation processing unit is used for carrying out information granulation processing on the time of the historical load data according to the similarity of the historical load data at adjacent moments to obtain the granulated historical load data of each time period;
the granulated power calculation unit is used for calculating the output power of each granulated time period by using an optimal power flow calculation method according to the historical load data of each granulated time period;
a power difference making unit, configured to make a difference between the output power of each time segment after the granulation and the output power before the granulation at each time of the corresponding time segment to obtain an output power difference value of each time segment;
the load difference unit is used for making a difference between the historical load data of each granulated time period and the historical load data of the corresponding time period to obtain a load difference value of each time period;
and the training unit is used for training the least square support vector machine by taking the load difference value of each time period and the load data of the previous time period of each time period as input and taking the corresponding output power difference value of each time period as output to obtain a trained power difference value correction model.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects: the invention discloses a dispatching method and a dispatching system of an electric power natural gas system. Through information granulation processing, the calculated amount is greatly reduced, and the scheduling efficiency is improved. Meanwhile, after the output power of the load after granulation is calculated, in order to avoid the accuracy reduction caused by information granulation processing, the historical data is used for correcting the output power, so that the scheduling efficiency is improved, and the scheduling accuracy is ensured.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without inventive exercise.
FIG. 1 is a method flow diagram of an embodiment of a method for scheduling an electric power and natural gas system of the present invention;
FIG. 2 is a flow chart of the pelletization step in an embodiment of the scheduling method of the electric power natural gas system of the present invention;
fig. 3 is a system configuration diagram of an embodiment of a scheduling system of an electric power and natural gas system according to 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.
The invention aims to provide a scheduling method and a scheduling system of a power and natural gas system, which improve scheduling efficiency.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Fig. 1 is a flowchart of a scheduling method of an electric power and natural gas system according to an embodiment of the present invention.
Referring to fig. 1, the scheduling method of the power and natural gas system includes:
step 1: acquiring predicted load data of scheduling days; the load can be accurately predicted by using the conventional load prediction method.
Step 2: and performing information granulation processing on the predicted load data to obtain granulated load data.
The method specifically comprises the following steps:
and performing information granulation processing on the time of the predicted load data according to the similarity of the predicted load data at adjacent moments to obtain the granulated load data of each time period.
And step 3: and calculating corresponding optimal output power according to the granulated load data.
And 4, step 4: and correcting the optimal output power by using historical load data to obtain corrected output power. The method specifically comprises the following steps:
load data of a time period preceding each time period is acquired.
And inputting the load data of each granulated time period and the load data of the previous time period of each time period into a trained power difference correction model to obtain an output power correction value of each time period.
And adding the output power of each time period under the optimized power flow after granulation to the output power correction value of each time period to obtain the optimized output power of each time period.
And 5: and scheduling the output of each generator and the gas well according to the corrected output power.
The training process of the power difference correction model in the step 4 is as follows:
acquiring historical load data of historical scheduling days;
calculating the output power before granulation at each moment by using an optimal power flow calculation method according to the historical load data;
performing information granulation processing on the time of the historical load data according to the similarity of the historical load data at adjacent moments to obtain the granulated historical load data of each time period;
calculating the output power of each granulated time period by using an optimal power flow calculation method according to the historical load data of each granulated time period;
the output power of each time period after granulation is differed with the output power of the corresponding time period before granulation at each moment to obtain the output power difference value of each time period;
the historical load data of each granulated time period is differenced with the historical load data of the corresponding time period to obtain a load difference value of each time period;
and training the least square support vector machine by taking the load difference value of each time period and the load data of the previous time period of each time period as input and the corresponding output power difference value of each time period as output to obtain a trained power difference value correction model.
In the scheduling method, the method for calculating the output power in the steps 2 and 4 is to establish an optimal power flow model of the electric power natural gas system, and then call a CPLEX commercial solver on an MATLAB platform by using a Yalmip toolbox to solve the output of each generator, gas well and the like under the optimal power flow.
The optimal power flow model of the power natural gas system is an economic optimal model, namely the output power conditions of each thermal power generator, each gas turbine and each gas well are reasonably distributed, so that the energy consumption cost of the whole system is minimum, and meanwhile, in order to increase the utilization of new energy, a punishment item of abandoned wind power is added. The constraint conditions comprise the constraint conditions of the power system and the constraint conditions of the natural gas network, and the minimum cost of energy consumption is taken as an objective function. The method specifically comprises the following steps:
constraint conditions of power system
(1) Supply and demand balance constraints
In the formula, Ni，Nw，NDThe number of generators, the number of wind power plants and the number of loads are respectively; pit，Pwt，PdtThe output power of the generator, the actual output power of the wind power plant and the power of the load at the time of the t period are respectively. j. w and d are the serial number of the generator, the serial number of the wind power plant and the serial number of the load respectively.
(2) Rotational back-up restraint
In the formula, PimaxThe maximum output power of the generator; SRtTo rotate the reserve capacity at time t.
(3) Power output constraints for a unit and a wind farm
In the formula (I), the compound is shown in the specification,
is the predicted output power of the wind farm at time t; piminIs the minimum output power of the generator at time t.
(4) Climbing restraint
In the formula, Pit，Pi,t-1The output power of the generator at the time t and the output power of the generator at the time t-1 are respectively; RU (RU)i，RDiAre respectively generatorsThe uphill and downhill rates of the group.
(5) Line transmission power constraint
PLbmin≤PLbt≤PLbmax(5)
In the formula, PLbmin，PLbt，PLbmaxRespectively, the minimum transmission power of the line, the actual transmission power of the line at the time of the t period, and the maximum transmission power of the line.
Second, constraint conditions of natural gas network
(1) Supply and demand constraint balancing
Wherein S (m) is a building set connected to natural gas node m; qωtThe yield of the natural gas well at the moment t; qltThe natural gas load of the node at the moment t; qitIs the natural gas consumption of the gas turbine at time t; qmn,tThe natural gas flow from the node m to the node n at the moment t.
(2) Gas flow model of transmission pipeline
In the formula, CmnIs a natural gas pipeline characteristic constant; p is a radical ofmtIs the air pressure at the node m at the time t; p is a radical ofntThe pressure at node n at time t.
(3) Restriction of gas point pressure limit
In the formula, pm,minMinimum value of gas pressure at node m, pm,maxIs the maximum value of the air pressure at node m, pn,minIs the minimum value of the gas pressure of node n, pn,maxIs the maximum value of the air pressure at node n.
Third, economic optimum model objective function
The objective function of the economic optimum model aiming at minimizing the cost of energy consumption of the whole system comprises three aspects:
minF＝F1+F2+F3(9)
wherein, F1For the objective function of the power system:
in the formula, ai、bi、ciIs the fuel cost factor of the generator.
F2For the objective function of the natural gas system:
in the formula, ggIs the cost coefficient of the g natural gas source; pgtRepresenting the natural gas output, N, over a period of tGThe number of natural gas sources.
F3A penalty function term for the abandoned wind power:
in the formula, ρwIs a penalty factor.
Regarding the CPLEX solver:
CPLEX is a high-performance mathematical programming problem solver of IBM company, and can quickly and stably solve a series of programming problems such as linear programming, mixed integer programming, quadratic programming and the like. And CPLEX can be called under the matlab platform to efficiently solve the mathematical programming problem. However, since the cost function and internal constraints in the mathematical model have non-linear components, the CPLEX solver may be used after linearization is required.
Nonlinear problem linearization process:
(1) carrying out segmentation processing on a cost function method and internal constraint, and averagely dividing independent variables into a plurality of segments (determining the number of the segments according to actual variables);
(2) take the objective function as an example, let f (x) be ax2+ bx + c in each segment, at x0(x) performing a taylor series expansion on f to obtain:
wherein f (x) is a simplified form of the objective function of the power system mentioned above; a, b, c fuel cost coefficient of the generator; x is the output power of the generator; x is the number of0Is the generator output power at the segment node (a constant).
(3) When the increment is (x-x)0) When small, omitting the highest power, then
And if the proportionality coefficient is k:
(4) finally, a linearized cost function can be obtained:
f(x)＝(ax0+b)x(16)
the information graining process involved in step 2 and step 4 is a process of dividing the study object into several sets of particles. A particle refers to a block formed by some individuals (elements, points, etc.) through an unclear relationship, a similar relationship, a proximity relationship, a functional relationship, or the like.
Fig. 2 is a flow chart of the granulation step in the scheduling method embodiment of the power natural gas system of the present invention.
Referring to fig. 2, the granulating step includes:
A. decomposition of payload data into a multiscale dataset X ═ X by wavelet decomposition1,x2,…,xn}。
B. Randomly selecting a sample from the data set as an initial cluster center c1。
C. Firstly, the shortest distance between each sample and the current existing cluster center, namely the distance between each sample and the nearest cluster center is calculated and is represented by D (x):
wherein x is1i,x2iRespectively, the sample data in the data set, and N is the dimension of the data.
D. Calculate the probability that each sample is selected as the next cluster center:
E. selecting the next clustering center by using a wheel disc method until all initial clustering centers C ═ C are selected1,c2,c3,...,ck}。
F. And calculating the distances from each sample in each data set to the K cluster centers and classifying the samples into the class corresponding to the cluster center with the minimum distance.
G. For each class ciRecalculating its cluster center
(i.e., the centroids of all samples belonging to the class).
H. Repeating the step F and the step G until the position of the cluster center is not changed. By the end of this granulation.
I. And B, after granulation, obtaining the load data after granulation, solving the root mean square error of the difference value of the load data before and after granulation, finishing the final granulation if the error meets a preset judgment value, and returning to the step B if the error does not meet the preset judgment value.
The principle of the Least Square Support Vector Machine (LSSVM) in step 4 is as follows:
the LSSVM introduces a least square linear system, has smaller generalization capability on small samples and nonlinear samples, can better solve the nonlinear problem, and can obtain better learning results without a large number of observation data samples when solving the actual problem. The specific calculation process is as follows:
for training sample data
Wherein eiFor the ith input data, piFor the ith output data, the sample set is fitted with a linear function of the high dimensional feature space as follows:
in the formula, eiThe output error of the generator before and after granulation; p is a radical ofiThe output of the generator before and after granulation; w is a feature space weight coefficient vector; c is an offset;
is a non-linear mapping from the input space to the high-dimensional feature space. According to the principle of minimizing structural risk, the following constraint optimization problem can be expressed:
in the formula, J (w, η) is an objective function in the risk minimization principle, χ is a regularization parameter, η is a relaxation factor, in order to solve the optimal problem, the constrained optimization problem needs to be changed into an unconstrained optimization problem, a Lagrangian function needs to be introduced, and the optimization problem in the formula (19) is transformed into a dual space, so that the risk minimization is realized
Wherein L is an augmented Lagrange function; a isiIs a lagrange multiplier.
According to KKT (Karush-Kuhn-tucker) conditions, then:
namely;
and finally solving a result by using a least square method to obtain regression of the LSSVM as follows:
in the formula (I), the compound is shown in the specification,
is piIs a function approximation; k (e, e)i) Is a kernel function. The kernel function is part of the idea of truly embodying the nonlinear variation in LSSVM, and a Gaussian radial basis kernel function is selected herein, which can be compared with eiAnd ejAnd mapping to 0 to 1, i.e. the original features can be mapped to a high dimension. The expression is as follows:
in the formula, ξ is a nuclear parameter.
After the step 4, the scheduling method of the invention also comprises a method of utilizing the nuclear density estimation to calculate the confidence interval of the output condition of each generator and each gas well, and a certain spare capacity is reserved for scheduling. The kernel density estimation is used for estimating an unknown density function in probability theory, and belongs to one of nonparametric inspection methods. The method for calculating the confidence interval of the output condition of each generator and each gas well by using the nuclear density estimation method comprises the following steps:
first, the probability density function of the error is obtained
For a set of generator output power difference values before and after granulation { e1,e2,…,enFor one of the differences edThe probability density function of (a) is:
in the formula, NiThe number of data in the difference set is k, which is the width of the neighborhood window, and k (x) is a kernel function, which is often a gaussian kernel function.
Second, the calculation of output power confidence interval
After the kernel density function is calculated, the corresponding accumulated integral of the kernel density function in the error interval [ a ', b' ], namely
For a certain output power value piWhen the given confidence is α, the confidence interval of its prediction error satisfies:
Fd(elower≤e≤eupper)＝1-α (29)
in the formula, elower，eupperRespectively, a lower limit and an upper limit of the scheduling interval. Therefore, a confidence interval corresponding to an error before and after the granulation of a certain output power value can be obtained, and a final scheduling interval of the output power of the generator can be obtained by adding the confidence interval of the error to the granulated output power.
Fig. 3 is a system configuration diagram of an embodiment of a scheduling system of an electric power and natural gas system according to the present invention.
Referring to fig. 3, the scheduling system of the power and natural gas system includes:
a scheduling day data obtaining module 301, configured to obtain predicted load data of a scheduling day;
a granulation processing module 302, configured to perform information granulation processing on the predicted load data to obtain granulated load data;
a power calculating module 303, configured to calculate corresponding optimal output power for the granulated load data;
a correction module 304, configured to correct the optimal output power by using historical load data to obtain a corrected output power;
and the scheduling module 305 is configured to schedule the output of each generator and each gas well according to the corrected output power.
The granulation processing module 302 includes:
and the first granulation processing unit is used for carrying out information granulation processing on the time of the predicted load data according to the similarity of the predicted load data at the adjacent time to obtain the granulated load data of each time slot.
The correction module 304 includes:
a previous-period data acquiring unit configured to acquire load data of a previous period of each period;
the model prediction unit is used for inputting the load data of each granulated time period and the load data of the previous time period of each time period into the trained power difference correction model to obtain an output power correction value of each time period;
and the correction unit is used for adding the output power of each time period under the optimized power flow after granulation to the output power correction value of each time period to obtain the optimized output power of each time period.
The scheduling system further comprises a model training module 306, wherein the model training module 306 is used for training the power difference correction model; the model training module comprises:
a historical data acquisition unit for acquiring historical load data of historical scheduling days;
the pre-granulation power calculation unit is used for calculating the output power before granulation at each moment by using an optimal power flow calculation method according to the historical load data;
the second granulation processing unit is used for carrying out information granulation processing on the time of the historical load data according to the similarity of the historical load data at adjacent moments to obtain the granulated historical load data of each time period;
the granulated power calculation unit is used for calculating the output power of each granulated time period by using an optimal power flow calculation method according to the historical load data of each granulated time period;
a power difference making unit, configured to make a difference between the output power of each time segment after the granulation and the output power before the granulation at each time of the corresponding time segment to obtain an output power difference value of each time segment;
the load difference unit is used for making a difference between the historical load data of each granulated time period and the historical load data of the corresponding time period to obtain a load difference value of each time period;
and the training unit is used for training the least square support vector machine by taking the load difference value of each time period and the load data of the previous time period of each time period as input and taking the corresponding output power difference value of each time period as output to obtain a trained power difference value correction model.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects: the invention discloses a dispatching method and a dispatching system of an electric power natural gas system. Through information granulation processing, the calculated amount is greatly reduced, and the scheduling efficiency is improved. Meanwhile, after the output power of the load after granulation is calculated, in order to avoid the accuracy reduction caused by information granulation processing, the historical data is used for correcting the output power, so that the scheduling efficiency is improved, and the scheduling accuracy is ensured.
For the system disclosed by the embodiment, the description is relatively simple because the system corresponds to the method disclosed by the embodiment, and the relevant points can be referred to the method part for description.
The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.
Claims (4)
1. A scheduling method of an electric power natural gas system is characterized by comprising the following steps:
acquiring predicted load data of scheduling days;
performing information granulation processing on the predicted load data to obtain granulated load data, which specifically comprises the following steps:
performing information granulation processing on the time of the predicted load data according to the similarity of the predicted load data at adjacent moments to obtain granulated load data of each time period;
calculating corresponding optimal output power according to the granulated load data;
correcting the optimal output power by using historical load data to obtain corrected output power, which specifically comprises the following steps:
acquiring load data of a previous time period of each time period;
inputting the load data of each granulated time period and the load data of the previous time period of each time period into a trained power difference correction model to obtain an output power correction value of each time period;
adding the output power of each time period under the optimized power flow after granulation to the output power correction value of each time period to obtain the optimized output power of each time period;
and scheduling the output of each generator and the gas well according to the corrected output power.
2. The scheduling method of the electric power natural gas system according to claim 1, wherein the training process of the power difference correction model comprises:
acquiring historical load data of historical scheduling days;
calculating the output power before granulation at each moment by using an optimal power flow calculation method according to the historical load data;
performing information granulation processing on the time of the historical load data according to the similarity of the historical load data at adjacent moments to obtain the granulated historical load data of each time period;
calculating the output power of each granulated time period by using an optimal power flow calculation method according to the historical load data of each granulated time period;
the output power of each time period after granulation is differed with the output power of the corresponding time period before granulation at each moment to obtain the output power difference value of each time period;
the historical load data of each granulated time period is differenced with the historical load data of the corresponding time period to obtain a load difference value of each time period;
and training the least square support vector machine by taking the load difference value of each time period and the load data of the previous time period of each time period as input and the corresponding output power difference value of each time period as output to obtain a trained power difference value correction model.
3. A dispatch system for an electric power and natural gas system, comprising:
the scheduling date data acquisition module is used for acquiring predicted load data of scheduling dates;
the granulation processing module is configured to perform information granulation processing on the predicted load data to obtain granulated load data, and specifically includes:
the first granulation processing unit is used for carrying out information granulation processing on the time of the predicted load data according to the similarity of the predicted load data at adjacent moments to obtain granulated load data of each time slot;
the power calculation module is used for calculating corresponding optimal output power aiming at the granulated load data;
the correction module is configured to correct the optimal output power by using historical load data to obtain a corrected output power, and specifically includes:
a previous-period data acquiring unit configured to acquire load data of a previous period of each period;
the model prediction unit is used for inputting the load data of each granulated time period and the load data of the previous time period of each time period into the trained power difference correction model to obtain an output power correction value of each time period;
the correction unit is used for adding the output power of each time period under the granulated optimal power flow to the output power correction value of each time period to obtain the optimal output power of each time period;
and the scheduling module is used for scheduling the output of each generator and each gas well according to the corrected output power.
4. The dispatching system of an electric power natural gas system according to claim 3, further comprising a model training module for training the power difference correction model; the model training module comprises:
a historical data acquisition unit for acquiring historical load data of historical scheduling days;
the pre-granulation power calculation unit is used for calculating the output power before granulation at each moment by using an optimal power flow calculation method according to the historical load data;
the second granulation processing unit is used for carrying out information granulation processing on the time of the historical load data according to the similarity of the historical load data at adjacent moments to obtain the granulated historical load data of each time period;
the granulated power calculation unit is used for calculating the output power of each granulated time period by using an optimal power flow calculation method according to the historical load data of each granulated time period;
a power difference making unit, configured to make a difference between the output power of each time segment after the granulation and the output power before the granulation at each time of the corresponding time segment to obtain an output power difference value of each time segment;
the load difference unit is used for making a difference between the historical load data of each granulated time period and the historical load data of the corresponding time period to obtain a load difference value of each time period;
and the training unit is used for training the least square support vector machine by taking the load difference value of each time period and the load data of the previous time period of each time period as input and taking the corresponding output power difference value of each time period as output to obtain a trained power difference value correction model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910072171.8A CN109687456B (en) | 2019-01-25 | 2019-01-25 | Scheduling method and system of electric power natural gas system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910072171.8A CN109687456B (en) | 2019-01-25 | 2019-01-25 | Scheduling method and system of electric power natural gas system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109687456A CN109687456A (en) | 2019-04-26 |
| CN109687456B true CN109687456B (en) | 2020-07-03 |
Family
ID=66194799
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910072171.8A Active CN109687456B (en) | 2019-01-25 | 2019-01-25 | Scheduling method and system of electric power natural gas system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109687456B (en) |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5550410A (en) * | 1994-08-02 | 1996-08-27 | Titus; Charles H. | Gas turbine electrical power generation scheme utilizing remotely located fuel sites |
| DE20209993U1 (en) * | 2002-06-28 | 2002-10-17 | Klinck Johann Heinrich | Vehicle with fuel cell drive |
| CN107622324A (en) * | 2017-09-01 | 2018-01-23 | 燕山大学 | A kind of robust environmental economy dispatching method for considering more microgrid energy interactions |
| CN108964041B (en) * | 2018-07-23 | 2021-02-23 | 燕山大学 | Control method for maximum load supply capacity of electricity-gas interconnection comprehensive energy system |
| CN109242366B (en) * | 2018-11-06 | 2020-08-07 | 国电南瑞科技股份有限公司 | Multi-period power flow optimization method of electricity-gas interconnection comprehensive energy system |
-
2019
- 2019-01-25 CN CN201910072171.8A patent/CN109687456B/en active Active
Non-Patent Citations (1)
| Title |
|---|
| 独立模式下微网多能存储系统优化配置;卢志刚等;《电力系统自动化》;20180225;第42卷(第4期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109687456A (en) | 2019-04-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | Short-term wind power forecasting based on support vector machine with improved dragonfly algorithm | |
| Ghaderi et al. | Deep forecast: Deep learning-based spatio-temporal forecasting | |
| CN110717688B (en) | Water, wind and light short-term combined optimization scheduling method considering new energy output uncertainty | |
| Qu et al. | Prediction of electricity generation from a combined cycle power plant based on a stacking ensemble and its hyperparameter optimization with a grid-search method | |
| CN110880789B (en) | Economic dispatching method for wind power and photovoltaic combined power generation system | |
| CN107766990B (en) | Method for predicting power generation power of photovoltaic power station | |
| Hong et al. | Optimal sizing of hybrid wind/PV/diesel generation in a stand-alone power system using Markov-based genetic algorithm | |
| Shi et al. | Hybrid forecasting model for very-short term wind power forecasting based on grey relational analysis and wind speed distribution features | |
| Zhang et al. | Wind power prediction based on LS-SVM model with error correction | |
| CN106529719B (en) | Wind power prediction method based on particle swarm optimization algorithm wind speed fusion | |
| CN110729764B (en) | Optimal scheduling method for photovoltaic power generation system | |
| CN112990558B (en) | Meteorological temperature and illumination prediction method based on deep migration learning | |
| CN102479347A (en) | Method and system for forecasting short-term wind speed of wind farm based on data driving | |
| CN105809349A (en) | Scheduling method considering incoming water correlation cascade hydropower stations | |
| Chen et al. | Research on wind power prediction method based on convolutional neural network and genetic algorithm | |
| CN103996087A (en) | Method and system for forecasting new energy power generation power | |
| CN113298407B (en) | Industrial park electricity-gas comprehensive energy system optimization scheduling model establishing method | |
| Li et al. | Research on short-term joint optimization scheduling strategy for hydro-wind-solar hybrid systems considering uncertainty in renewable energy generation | |
| CN114154684A (en) | Short-term photovoltaic power prediction method based on data mining and multi-core support vector machine | |
| CN106208136A (en) | Meter and benefit and the dispatching method a few days ago containing uncertain wind-powered electricity generation of risk | |
| CN116029404A (en) | Multi-source power system coordination optimization scheduling method and system in typhoon environment | |
| CN112288130B (en) | New energy consumption calculation method based on two-stage multi-objective optimization | |
| CN113344283A (en) | Energy internet new energy consumption capacity assessment method based on edge intelligence | |
| CN109687456B (en) | Scheduling method and system of electric power natural gas system | |
| Kader et al. | Active power control strategy for wind farms based on power prediction errors distribution considering regional data |
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 |