CN109888800B - Power distribution area three-phase unbalanced load adjustment method based on load prediction and commutation strategy - Google Patents
Power distribution area three-phase unbalanced load adjustment method based on load prediction and commutation strategy Download PDFInfo
- Publication number
- CN109888800B CN109888800B CN201910174226.6A CN201910174226A CN109888800B CN 109888800 B CN109888800 B CN 109888800B CN 201910174226 A CN201910174226 A CN 201910174226A CN 109888800 B CN109888800 B CN 109888800B
- Authority
- CN
- China
- Prior art keywords
- phase
- load
- clustering
- prediction
- commutation
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
- Y02E40/50—Arrangements for eliminating or reducing asymmetry in polyphase networks
Landscapes
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a distribution station three-phase unbalanced load adjusting method based on load prediction and a commutation strategy, which comprises the steps of firstly clustering historical daily loads by adopting a K-means algorithm, carrying out short-term load prediction on historical data subjected to statistical classification processing by utilizing a support vector machine, calculating the unbalance degree of three-phase load current when a corresponding distribution transformer operates, finally establishing an optimal commutation mathematical model taking the minimum unbalance degree of the three-phase current of a distribution station and the minimum switching times of commutation switches as targets, and obtaining an optimal commutation scheme by using a genetic algorithm. The invention effectively reduces the line loss and the three-phase load unbalance degree and relieves the three-phase load unbalance problem of the distribution station area.
Description
Technical Field
The invention belongs to the technical field of three-phase unbalanced load adjustment, and particularly relates to a power distribution station three-phase unbalanced load adjustment method based on load prediction and a phase conversion strategy.
Background
The degree of automation of a power distribution network in China is low, single-phase loads are large, particularly in rural areas, power consumers are complicated and are not easy to plan, single-phase load distribution is unbalanced, and a power distribution platform area has a serious problem of unbalanced three-phase loads. At present, the method for treating the three-phase load unbalance problem at home and abroad mainly comprises the steps that a reactive compensation device switches a capacitor bank; the inter-phase capacitance transfers active power; the commutation switch device adjusts a load, and the like. The static var generator, the interphase jumper power capacitor and other reactive power compensation devices do not fundamentally solve the problem of three-phase load imbalance; the load automatic phase modulation device is expensive, the communication between the control terminal and the phase change switch is complex, and the load automatic phase modulation device is difficult to be widely applied in a power distribution station area; the manual phase modulation operation has a certain time delay and hysteresis. In summary, the current three-phase unbalanced load adjustment has the problems of high operation and maintenance cost, time lag and the like.
With the development of an intelligent algorithm becoming mature gradually, the load prediction based on the clustering analysis and the support vector machine algorithm can accurately predict the power utilization data of the load in a period of time in the future, and the genetic algorithm is widely applied to the screening of the optimal scheme of the power system. However, these techniques have limited application in three-phase unbalanced load regulation.
Disclosure of Invention
The invention provides a three-phase unbalanced load adjusting method for a power distribution area based on load prediction and a phase change strategy, aiming at the problems that the influence of phase change on load, the service life of a phase change switch, the economical efficiency of the power distribution area and the like are neglected when the traditional power distribution area governs three-phase unbalance.
In order to achieve the technical purpose, the technical scheme adopted by the invention is as follows:
a three-phase unbalanced load adjustment method for a power distribution area based on load prediction and a commutation strategy comprises the following steps:
s1: clustering historical daily loads by adopting a K-means algorithm, and performing short-term load prediction on historical data subjected to statistical classification processing by utilizing a support vector machine;
s2: calculating the unbalance degree of the three-phase load current when the corresponding distribution transformer operates;
s3: establishing an optimal phase change strategy target function which aims at minimizing the unbalance degree of three-phase currents in a power distribution area and minimizing the switching times of a phase change switch;
s4: and obtaining an optimal commutation scheme through a genetic algorithm.
In order to optimize the technical scheme, the specific measures adopted further comprise:
the step S1 includes the steps of:
s11: preprocessing all historical data related to the time to be predicted, namely sequentially inputting the historical data into a sample set, wherein each piece of historical data comprises meteorological factors such as time, load, temperature, pressure, humidity and the like;
s12: setting a clustering number K, and carrying out clustering analysis on the data according to meteorological factors;
s13: screening out a historical data set similar to meteorological factors at the moment to be predicted according to the clustering result to serve as a training sample;
s14: and (4) completing the training and prediction of the load by using a support vector machine algorithm.
Step S12, the cluster analysis method comprises the following steps: the meteorological data with high similarity belong to the same class, and the meteorological data between different classes have larger difference; the clustering effect is in direct proportion to the similarity of the meteorological data in the same class and in inverse proportion to the similarity between the classes.
The step S12 specifically includes the following steps:
(1) Let X = { X 1 ,x 2 ,x 3 …,x n Is a data set of n m-dimensional samples to be sorted, where x i ∈R m Selecting k samples from the data set as initial clustering centers;
(2) Calculating Euclidean distances from all points in the data set to K clustering centers, wherein the smaller the distance is, the higher the similarity is, and clustering samples with higher similarity into one class;
(3) Each cluster C j Represents a class, n j Representing the number of samples contained in the j cluster, and re-determining the central sample omicron in the k cluster samples j :
(4) Calculating out-of-center samples of each cluster to the center samples thereof j Sum of squares of total distances S:
and when the S obtains the minimum value and the clustering center is not changed any more, finishing clustering.
Step S14, the prediction function is:
in the formula, the kernel function meeting the Mercer condition is:
and (3) completing the nonlinear regression prediction of the support vector machine algorithm by controlling the kernel function, the balance coefficient c and the three-phase current unbalance degree epsilon.
The objective function of the commutation strategy in step S4 is expressed as equation (11):
M＝min{(ε,t)} (11)
in the formula, the three-phase current unbalance epsilon is calculated by the feeder current, and the calculation formula is as follows:
wherein the content of the first and second substances,
for the effective value of each phase current at the outlet side of the distribution transformer, I av The average value of three-phase current at the outlet side of the distribution transformer is obtained;
the calculation formula of the adjusting times t of the phase change switch is as follows:
wherein N is the number of users, | Δ S, of phase change switches installed in the distribution area y And | is a module value of the phase sequence vector difference of the states before and after the phase change of the switch.
The invention has the following beneficial effects:
the invention provides a novel three-phase load unbalance treatment method for a power distribution area, which comprises the following steps: the method comprises the steps of collecting historical load data of each user in a distribution substation area, classifying each time of a day to be predicted according to meteorological factors by using a K-means clustering algorithm, then establishing a support vector machine model to predict load electricity consumption, and adjusting load phase in advance to prevent the three-phase imbalance problem when a serious three-phase load imbalance problem occurs in a prediction result. The effect of aggravating the unbalanced degree of the three-phase load due to failure of timely load adjustment is avoided, meanwhile, the line loss is reduced, the service life of the phase change switch is prolonged, and the economical efficiency and the power supply reliability of the power distribution station area are improved.
Drawings
FIG. 1 is a flow chart of load prediction according to the present invention.
FIG. 2 is a genetic algorithm commutation optimization flow chart of the present invention.
Fig. 3 shows the load prediction results of year 2016, 9, month 11.
Fig. 4 is a three-phase current imbalance fitness function convergence curve according to an embodiment of the present invention.
Detailed Description
Embodiments of the present invention are described in further detail below with reference to the accompanying drawings.
1 load prediction principle of the invention
According to the distribution area three-phase unbalanced load adjusting method based on the load prediction and the phase commutation strategy, the load prediction result is used for predicting the three-phase unbalanced degree of the distribution area in advance, and the requirement for quite high load prediction accuracy is met. Since the short-term load prediction is influenced by meteorological factors such as temperature and air pressure, historical data similar to the meteorological factors at the moment to be predicted is used as a training volume book, and the method is a key step for improving the short-term load prediction accuracy.
The invention adopts a method of firstly carrying out K-means clustering and then carrying out SVM modeling to carry out load prediction: firstly, preprocessing all historical data related to a moment to be predicted, namely sequentially including reference values such as moment, load, temperature, pressure, humidity and the like in each piece of historical data; then inputting the data into a sample set, setting a clustering number K, and carrying out clustering analysis on the data according to meteorological factors; and finally, according to the principle of 'big-end-up and small-end-up', screening out a historical data set similar to meteorological factors at the moment to be predicted according to the clustering result as a training sample, and completing the training and prediction of the load by using a support vector machine algorithm. The specific flow is shown in figure 1.
1.1K-means Cluster analysis
The K-means algorithm belongs to a typical dynamic clustering method, roughly divides all sample points into K classes, then corrects the original rough division by calculating the Euclidean distance between two objects, and completes the clustering of the sample points through the hierarchical iteration of the algorithm. The invention clusters the historical data set according to the meteorological factors such as temperature, pressure, humidity and the like: the meteorological data with high similarity belong to the same category, and the meteorological data between different categories have larger difference. The clustering effect is in direct proportion to the similarity of the meteorological data in the same class and in inverse proportion to the similarity between the classes. The specific steps are as follows:
(1) Let X = { X 1 ,x 2 ,x 3 …,x n Is a data set of n m-dimensional samples to be classified, where x i ∈R m . Selecting k samples from the data set as initial clustering centers;
(2) Calculating Euclidean distances from all points in the data set to K clustering centers, wherein the smaller the distance is, the higher the similarity is, and clustering samples with higher similarity into one class;
(3) Each cluster C j Represents a class, n j Representing the number of samples contained in the j cluster, and re-determining the central sample omicron in the k cluster samples j :
(4) Calculating the samples from the center to the center of each cluster j Sum of squares of total distances S:
and when the S obtains the minimum value and the clustering center is not changed any more, finishing clustering.
1.2 support vector machine
A Support Vector Machine (SVM) algorithm is proposed in the last 90 th century, is widely applied under complex practical conditions of small samples, system nonlinearity and the like, and overcomes the limitation of experience risk minimization of a neural network [10-11] . On the basis of sample cluster analysis, the SVM is used for prediction, so that the training speed and the prediction precision can be effectively improved. The SVM principle model is as follows:
suppose { (x) i ,y i ),i＝1,2,…,n,x i ∈R n ,y i e.R } is given training sample set, where x i ∈R n Representing the input value of a variable, y i E.g. the output value of R corresponding input variable, n is the training sampleThis total number. The support vector machine principle is that the original sample input value of nonlinear relation is mapped through nonlinearity
And performing linear regression in the high-dimensional feature space, so that the effect of nonlinear regression in the original space can be achieved.
Let y = f (x) = (ω · x) + b be the regression function from input space to output space, where x is the known quantity representing the variable input value; omega ∈ R n Representing weight, b ∈ R representing threshold, and solving by the SVM by adopting the principle of structure risk minimization, which is specifically as follows:
in the formula (I), the compound is shown in the specification,
indicating a degree of smoothing; c is the equilibrium coefficient; l (y, f (x)) represents the loss function, i.e.:
in the formula, y represents a predicted value, f (x) represents an actual value, and if a penalty function xi is introduced i 、
Formula (1) can be converted into:
because the direct solution of the formula (3) is difficult, the Langcange factor and the kernel function are introduced, and the solution is simplified by using dual skills:
the final regression function expression may be:
in the formula (I), the compound is shown in the specification,
representing a kernel function that meets the Mercer condition. The invention adopts a Gaussian kernel function:
and (3) finishing SVM nonlinear regression prediction by controlling a kernel function and two parameters of c and epsilon.
2 the commutation optimization strategy based on genetic algorithm of the invention
Genetic Algorithm (GA), originally proposed by professor j.h.holland, usa, is an adaptive global optimization search algorithm that is developed when an organism is simulated to inherit and evolve in a natural environment. [13-17] The genetic algorithm has the advantages of strong robustness and self-adaptability, high searching capability and low degree of dependence on a target function, and is very suitable for solving the problem of the optimal solution of commutation. According to the method, under the actual condition that the commutation switch does not cover all users, an optimal commutation mathematical model is built by using the objective function, and then an optimal commutation scheme of the load of the power distribution area is obtained by screening by using an improved genetic algorithm, so that the imbalance degree of three-phase current of the power distribution area is reduced to the maximum extent, the switch adjustment times are reduced as far as possible, the power supply economy and safety of the power distribution area are improved, and the service life of the commutation device is prolonged.
2.1 objective function
The invention establishes a target function according to the standard of minimum unbalance degree of three-phase current and minimum adjustment times of a phase change switch, and the method specifically comprises the following steps:
the first target is: the unbalance of the three-phase current is minimal. The three-phase current unbalance epsilon is an important index for measuring whether the three-phase load of the distribution area is balanced or not, and can be calculated by a feeder line current:
wherein the content of the first and second substances,
representing the effective value of each phase current at the outlet side of the distribution transformer; i is av The average value of three-phase current at the outlet side of the distribution transformer is obtained. To minimize the unbalance of three phases, only the corresponding phase sequence of the user who installs the phase change switch is adjusted to make
The minimum is required.
And a second target: the number of times t of phase change switch adjustment is as small as possible. The service life of the phase-changing switch is limited, and the economical efficiency of the operation of the power distribution area can be effectively improved by reducing the phase-changing times. The third-order square matrix unit column vectors respectively represent phase sequences A, B and C of the commutation switch, the adjustment times of the commutation switch can be obtained through phase sequence state vectors before and after commutation, and the formula (10):
n is the number of users for installing a phase change switch in a power distribution station area; | Δ S y And | is a module value of the phase sequence vector difference of the states before and after the phase change of the switch.
In summary, the objective function of the commutation strategy is expressed as equation (11):
M＝min{(ε,t)} (11)
2.2 implementation scheme
Genetic Algorithm Using Gene coding strategies to achieve genetic coding, i.e., genes [100 ]] T 、[010] T 、 [001] T Respectively generation by generationAnd switching the meter load switch to A, B and C phases. Inter-gene interdependence and mutual exclusion. The N users provided with the phase change switches arrange the genes according to the sequence of the distribution lines to form a chromosome, namely a switch phase sequence state matrix K. The genetic operation comprises cross operation and mutation operation, wherein the cross operation adopts a mode of replacing gene lines with vectors at the same positions, breakpoints are randomly selected at the same positions of two chromosomes, and the left segments of the breakpoints are mutually exchanged to form two new chromosomes. Then, mutation operation is carried out, after genes needing mutation are randomly selected, each gene is enabled to be 100 according to the mutual exclusion dependence relationship among the genes and the mutation rate of the algorithm] T 、[010] T 、[001] T The three vectors are mutated to form different chromosomes. The specific flow is shown in fig. 2.
3 analysis of examples
3.1 load forecast data preparation
According to the embodiment of the invention, load data of a whole point from 2016, 8, 11 and 9 to 10 days in the distribution area of the Anhui, mega, china is taken as historical load data for analysis. Wherein, each historical load data comprises 9 weather-related data such as air temperature, pressure intensity, humidity and the like besides date and time. The predicted time is load data at the point of 2016, 9, 11 days in the future. The clustering analysis sample data takes historical data L (t-i) 2 hours before a time t to be predicted and historical data L (t-24 j) at the time t in the last 30 days, wherein i =1,2; j =1,2,3, \ 8230;, 30. Before prediction, clustering analysis is carried out on sample data according to meteorological factors, the meteorological category to which the time to be predicted belongs is classified according to a clustering result, and the historical load of the meteorological category is used as a training sample of the time to be predicted. Then, the training set and the test set are normalized, see formula (12):
in the formula, x and y are data before and after normalization respectively; x is the number of min 、x max Respectively the minimum and maximum values in the raw data x.
3.2 load prediction results
The experimental computing environment was matlab2016blibsvm3.14. The cluster analysis before prediction and the prediction method using only the support vector machine algorithm are compared and the average relative error value is calculated by equation (13).
It can be calculated from fig. 3 and table 1 that the prediction accuracy of the single SVM algorithm is 4.03%, while the average relative error of the method adopting cluster analysis and support vector machine prediction is only 1.71%, and the prediction accuracy is improved by 2.33%.
TABLE 1 average relative error values obtained by different methods
3.3 commutation optimization
Load data of 2016, 9, 8 and 8 days of a certain power distribution area in Mega Chang is predicted through the prediction method. Table 2 shows the times when the unbalance degree of the three-phase load current is the highest in the prediction result, and it can be seen from the table that the unbalance degree of the three-phase load in the platform area gradually increases along with the time, and the load needs to be adjusted in advance.
TABLE 2 Current values before commutation and corresponding imbalances
And (3) searching an optimal solution of load commutation by improving a genetic algorithm. The parameters are set as follows: the iteration number n is 50, the population number m is 100, and the variation rate is 0.01. Varying the variance ratio results in a fitness function convergence curve as shown in fig. 4. It can be seen from the figure that the convergence effect is better when the variation ratio is 0.01.
Table 3 shows the current values and the imbalance degrees of the phases at the corresponding time after the load adjustment. Comparing table 2, it can be clearly seen that the unbalance of the three-phase current is greatly reduced, the minimum is 6.7%, and the maximum is 14.2%.
TABLE 3 Current values after phase change and their corresponding unbalance degrees
In summary, according to the method for adjusting the three-phase unbalanced load of the power distribution area based on the load prediction and the phase change strategy, the three-phase current unbalance degree at each moment is obtained through the load prediction, the moment when the load adjustment needs to be performed in advance is determined, and then the optimal load adjustment scheme is obtained through the improved genetic algorithm. The analysis result of the embodiment shows that the method can adjust the phase sequence of the load in advance, avoid the result that the unbalance degree of the three-phase load is aggravated because the load cannot be adjusted in time, reduce the line loss, prolong the service life of the phase change switch, and improve the economy and the power supply reliability of the power distribution area.
The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above-mentioned embodiments, and all technical solutions belonging to the idea of the present invention belong to the protection scope of the present invention. It should be noted that modifications and embellishments within the scope of the invention may be made by those skilled in the art without departing from the principle of the invention.
Claims (1)
1. A power distribution station three-phase unbalanced load adjustment method based on load prediction and a phase commutation strategy is characterized by comprising the following steps:
s1: clustering the historical daily load by adopting a K-means algorithm, and performing short-term load prediction on the historical data subjected to statistical classification processing by utilizing a support vector machine;
s2: calculating the unbalance degree of the three-phase load current when the corresponding distribution transformer operates;
s3: establishing an optimal phase change strategy target function which aims at minimizing the unbalance degree of three-phase currents in a power distribution area and minimizing the switching times of a phase change switch;
s4: obtaining an optimal commutation scheme through a genetic algorithm;
the step S1 includes the steps of:
s11: preprocessing all historical data related to the time to be predicted, namely, each piece of historical data sequentially comprises meteorological factors of time, load, temperature, pressure and humidity, and inputting the historical data into a sample set;
s12: setting a clustering number K, and carrying out clustering analysis on the data according to meteorological factors;
s13: screening out a historical data set similar to meteorological factors at the moment to be predicted according to the clustering result to serve as a training sample;
s14: completing the training and prediction of the load by using a support vector machine algorithm;
step S12, the cluster analysis method comprises the following steps: the meteorological data with high similarity belong to the same category, and the meteorological data between different categories have larger difference; the clustering effect is in direct proportion to the similarity of the meteorological data in the same class and in inverse proportion to the similarity between the classes;
step S12 specifically includes the following steps:
(1) Let X = { X = 1 ,x 2 ,x 3 …,x n Is a data set of n m-dimensional samples to be classified, where x i ∈R m Selecting k samples from the data set as initial clustering centers;
(2) Calculating Euclidean distances from all points in the data set to K clustering centers, wherein the smaller the distance is, the higher the similarity is, and clustering samples with higher similarity into one class;
(3) Each cluster C j Represents a group, n j Representing the number of samples contained in the j cluster, and re-determining the central sample o in the k cluster samples j :
(4) Calculate the off-center samples of each cluster to its center sample o j Sum of squares of total distances S:
when the S obtains the minimum value and the clustering center does not change any more, the clustering is finished;
step S14, the prediction function is:
in the formula, the kernel function meeting the Mercer condition is:
K(x i ,x)＝exp(-||x-x i || 2 /2σ 2 ) (8)
the nonlinear regression prediction of the support vector machine algorithm is completed by controlling the kernel function, the balance coefficient c and the three-phase current unbalance epsilon;
the objective function of the commutation strategy in step S3 is expressed as formula (11):
M＝min{(ε,t)} (11)
in the formula, the three-phase current unbalance degree epsilon is calculated by the feeder current, and the calculation formula is as follows:
wherein, the first and the second end of the pipe are connected with each other,
for the effective value of each phase current at the outlet side of the distribution transformer, I av The average value of three-phase current at the outlet side of the distribution transformer is obtained;
the calculation formula of the adjusting times t of the phase change switch is as follows:
wherein N is the number of users, | Δ S, of phase change switches installed in the distribution area y And | is a module value of the phase sequence vector difference of the states before and after the phase change of the switch.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910174226.6A CN109888800B (en) | 2019-03-07 | 2019-03-07 | Power distribution area three-phase unbalanced load adjustment method based on load prediction and commutation strategy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910174226.6A CN109888800B (en) | 2019-03-07 | 2019-03-07 | Power distribution area three-phase unbalanced load adjustment method based on load prediction and commutation strategy |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109888800A CN109888800A (en) | 2019-06-14 |
| CN109888800B true CN109888800B (en) | 2022-11-29 |
Family
ID=66931250
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910174226.6A Active CN109888800B (en) | 2019-03-07 | 2019-03-07 | Power distribution area three-phase unbalanced load adjustment method based on load prediction and commutation strategy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109888800B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110137985B (en) * | 2019-06-21 | 2021-04-23 | 广东电网有限责任公司 | Commutation switch control method and related device |
| CN110533304B (en) * | 2019-08-12 | 2022-12-09 | 湖南大学 | Power system load uncertainty analysis method |
| CN110460072A (en) * | 2019-08-21 | 2019-11-15 | 天津大学 | Consider the real-time Management strategy of low-voltage platform area three-phase imbalance problem of phase index |
| CN110942173B (en) * | 2019-10-15 | 2022-04-19 | 合肥工业大学 | Power distribution station energy-saving loss-reducing method based on load prediction and phase sequence optimization |
| CN111200290B (en) * | 2020-03-16 | 2021-12-31 | 广东电网有限责任公司 | Intelligent control method of phase change switch for three-phase unbalance treatment of transformer area |
| CN111581883B (en) * | 2020-05-09 | 2022-09-23 | 国网上海市电力公司 | Method for calculating and predicting load on automation device |
| CN111525585B (en) * | 2020-07-06 | 2020-10-23 | 深圳华工能源技术有限公司 | Voltage-stabilizing energy-saving and three-phase imbalance treatment energy-saving coordination control method |
| CN113036786A (en) * | 2021-03-05 | 2021-06-25 | 云南电网有限责任公司电力科学研究院 | Low-voltage distribution transformer user phase sequence identification and three-phase imbalance adjustment method |
| CN113541165A (en) * | 2021-07-19 | 2021-10-22 | 安徽大学 | Three-phase imbalance intelligent phase commutation method based on load prediction |
| CN115378001B (en) * | 2022-10-25 | 2023-03-24 | 南昌科晨电力试验研究有限公司 | Low-voltage distribution network artificial phase modulation method and system based on load periodicity |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108173273A (en) * | 2017-12-30 | 2018-06-15 | 国网天津市电力公司电力科学研究院 | A kind of intelligent phase-change switch system and method for adjusting three-phase imbalance |
| CN108921324A (en) * | 2018-06-05 | 2018-11-30 | 国网江苏省电力有限公司南通供电分公司 | Platform area short-term load forecasting method based on distribution transforming cluster |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9041246B2 (en) * | 2010-09-29 | 2015-05-26 | General Electric Company | System and method for phase balancing in a power distribution system |
-
2019
- 2019-03-07 CN CN201910174226.6A patent/CN109888800B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108173273A (en) * | 2017-12-30 | 2018-06-15 | 国网天津市电力公司电力科学研究院 | A kind of intelligent phase-change switch system and method for adjusting three-phase imbalance |
| CN108921324A (en) * | 2018-06-05 | 2018-11-30 | 国网江苏省电力有限公司南通供电分公司 | Platform area short-term load forecasting method based on distribution transforming cluster |
Non-Patent Citations (1)
| Title |
|---|
| 配电台区三相负荷不平衡实时在线治理方法研究;方恒福等;《中国电机工程学报》;20150505;第35卷(第09期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109888800A (en) | 2019-06-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109888800B (en) | Power distribution area three-phase unbalanced load adjustment method based on load prediction and commutation strategy | |
| Yang et al. | Day-ahead wind power forecasting based on the clustering of equivalent power curves | |
| CN107688879B (en) | Active power distribution network distributed power supply planning method considering source-load matching degree | |
| Huang et al. | A clustering based grouping method of nearly zero energy buildings for performance improvements | |
| CN110909912A (en) | Park electric power system net load combination prediction method based on self-adaptive error feedback | |
| CN108173273A (en) | A kind of intelligent phase-change switch system and method for adjusting three-phase imbalance | |
| CN113033917B (en) | Sewage treatment plant prediction planning operation management method based on peripheral data | |
| CN109376950A (en) | A kind of polynary Load Forecasting based on BP neural network | |
| CN104376371B (en) | A kind of distribution based on topology is layered load forecasting method | |
| CN111461921B (en) | Load modeling typical user database updating method based on machine learning | |
| CN114970362A (en) | Power grid load scheduling prediction method and system under multi-energy structure | |
| CN112785119A (en) | Distribution network voltage out-of-limit reason analysis method based on clustering and hierarchical analysis algorithm | |
| CN114611842B (en) | Whole-county roof distributed photovoltaic power prediction method | |
| CN110009385A (en) | A kind of photovoltaic power generation user group division methods based on multifactor mapping | |
| CN115313403A (en) | Real-time voltage regulation and control method based on deep reinforcement learning algorithm | |
| CN110649633B (en) | Power distribution network reactive power optimization method and system | |
| CN110163437A (en) | Day-ahead photovoltaic power generation power prediction method based on DPK-means | |
| Jahan et al. | Intelligent system for power load forecasting in off-grid platform | |
| CN115102195A (en) | Three-phase load unbalance treatment method and device based on rural power grid | |
| Shao et al. | Optimization method based on load forecasting for three-phase imbalance mitigation in low-voltage distribution network | |
| CN107294112B (en) | Dynamic reactive power optimization method based on daily real-time | |
| CN112700069B (en) | Short-term load prediction method for regional power distribution network containing energy storage | |
| CN113410852B (en) | Power distribution network three-phase load unbalance adjustment optimization method and system | |
| Zhang et al. | A three-phase unbalance adjustment method based on improved k-means clustering | |
| Fan et al. | Three-phase imbalance governance method for distribution network based on Big Data analysis |
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 |