CN115800330A - Optimal configuration and scheduling method of automatic three-phase load adjusting device in power distribution area - Google Patents
Optimal configuration and scheduling method of automatic three-phase load adjusting device in power distribution area Download PDFInfo
- Publication number
- CN115800330A CN115800330A CN202211619638.4A CN202211619638A CN115800330A CN 115800330 A CN115800330 A CN 115800330A CN 202211619638 A CN202211619638 A CN 202211619638A CN 115800330 A CN115800330 A CN 115800330A
- Authority
- CN
- China
- Prior art keywords
- phase
- load
- data
- value
- adjusting device
- 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
-
- 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 relates to an optimal configuration and scheduling method of an automatic three-phase load adjusting device in a power distribution area, and belongs to the field of three-phase imbalance of an intelligent power distribution network. Based on the distribution condition of each single-phase load, the three-phase load automatic adjusting device with the least number is designed to be connected, so that the three-phase unbalance is reduced, and the economy is improved; and designing a minimum regulation and control action method, reducing the action times of the switch, taking the minimum switch action times and the minimum three-phase unbalance degree as objective functions and quickly searching for an optimal solution in the objective functions by using an intelligent algorithm in order to prolong the service life of the phase change switch. The beneficial effect is that the economy is improved on the premise of ensuring the three-phase balance. The method can effectively reduce the action times of the three-phase load automatic regulating device, prolong the service life of the phase change switch, effectively solve the optimization problem, improve the economical efficiency while reducing the unbalance of three phases, and can be popularized and applied, and improve the comprehensive income of power supply companies.
Description
Technical Field
The invention belongs to the field of three-phase imbalance of an intelligent power distribution network, and particularly relates to an optimal configuration and scheduling method for an automatic three-phase load adjusting device of a low-voltage distribution area.
Background
A three-phase four-wire system power supply mode is mainly adopted for a power distribution network in China, more than 90% of loads of urban users are single-phase loads, the load of a certain phase is rapidly increased or reduced, three-phase unbalance of the power distribution network is caused, once the three-phase unbalance is intensified, related equipment of the power distribution network is endangered, users and merchant equipment are harmed, and huge economic loss is caused. The three-phase imbalance occurs for several reasons:
(1) Load distribution is uneven, and it is single-phase load mostly to join in marriage net user, and incomplete balance causes the three-phase unbalance in the actual operation. In the planning design of the power distribution network, designers do not fully consider the actual load distribution of the power grid and the development trend of newly added loads. In actual operation, a certain phase or two phases are overloaded, while the other phases are lightly loaded, and an imbalance phenomenon occurs.
(2) The loads have randomness, the proportion of single-phase loads in all the loads in the power distribution network is large, the use conditions of the loads in 24 hours are different when the power distribution network operates daily, and corresponding adjustment is carried out according to the requirements of social production activities.
(3) Other factors include load change caused by abnormal load equipment and various loopholes in the management system of a power supply company; and thirdly, the periodic change of natural seasons, such as summer, the air conditioner load is greatly consumed due to hot weather.
At present, the unbalance phenomenon of the power distribution network is quite common, and according to statistics, the three-phase unbalance degree of nearly one third of the distribution network exceeds 25 percent and far exceeds 15 percent of the national standard. The method can achieve the purpose of reducing three-phase imbalance only by changing three-phase load distribution, and can effectively overcome the defects of low manual operation speed and high danger.
Disclosure of Invention
The invention provides an optimal configuration and scheduling method of an automatic three-phase load adjusting device of a power distribution area, and aims to effectively solve the optimization problem.
The technical scheme adopted by the invention is that the method comprises the following steps:
the method comprises the following steps of (I) optimizing configuration of a three-phase load automatic adjusting device of a power distribution station area based on probability distribution;
based on the distribution condition of each single-phase load, under the condition of meeting the adjustment of load unbalance, the number of the three-phase load automatic adjusting devices is configured as less as possible to improve the economy,
(II) phase change switch optimization scheduling method for minimizing action cost;
when three phases are unbalanced, the corresponding three-phase load automatic adjusting device is required to be used; a target is set to operate as much as possible in a shorter period of time to achieve three-phase equalization.
The step (one) of the invention specifically comprises the following steps:
(1) Analyzing the system data, identifying and correcting abnormal points of the data,
(2) Processing a data missing value;
(3) Data are processed in a standardized way so as to facilitate data analysis;
(4) And carrying out reasonable configuration of the three-phase load automatic adjusting device according to the system data analysis result.
The data anomaly point identification and correction in the step (1) comprises the following steps:
because the power load has a periodic variation trend, the load difference between the adjacent working days and the rest days in the same time is not too large, if the difference between the two is over a specified threshold, abnormal data may be generated, care must be taken when handling outliers, otherwise misjudgment may be caused, and data distortion may be caused, and for the wrong data, there are generally two processing methods: transverse processing and longitudinal processing;
transverse processing: the load data is usually a continuous smooth sequence without abrupt changes, so there is no significant difference between the current load data and the adjacent load data, and if the load data at a certain time exceeds a predetermined threshold, there is a relationship in equation (1):
then, the horizontal processing may be performed by using an averaging method, and the specific processing manner is as shown in formula (2):
wherein, alpha (t), beta (t) are threshold values, x (k, t) is the load value at the time t on the kth day;
longitudinal treatment: the power load has obvious periodic variation, and as known from the data period law, the load values of the same date and type at the same time are relatively close, and if the difference between the load values of the two exceeds a preset threshold value, the relation in the formula (3) exists:
|X(k,t)-M(t)|>γ(t) (3)
the vertical processing method may be used to correct the out-of-range abnormal data, and the correction process is shown in equation (4):
wherein, M (t) is the average value of recent days in the historical load data, M (t) is a prescribed threshold value, and X (k, t)' is the load data after vertical processing;
the data missing value processing in the step (2) of the invention comprises the following steps:
there are three methods for handling data loss: deleting, supplementing and ignoring the data, wherein the power load data is generally processed by a supplementing method, the missing data part is generally filled according to loads of similar dates or adjacent dates, and a specific repair formula is as follows:
X(d,t)＝ω 1 X(d,t-1)+ω 2 X(d,t+1)+ω 3 X(d-1,t)+ω 4 X(d+1,t) (5)
wherein X (d, t) represents the power load value at the time t on the d day; w is a i (i =1, 2, 3, 4) is a weighting coefficient;
the step (3) of data standardization processing of the invention comprises the following steps:
the number of electric loads is often very large, if the electric loads are directly introduced into the system, convergence is reduced, so that it is necessary to normalize the data of the loads to avoid rapid deterioration, when the digital type variables are normalized, a Min-Max method is used, after the data are normalized according to equation (6), the distribution of the data is limited to 0-1, the normalized method can shorten the data amount of the sample and reduce the timeliness of the training mode, and the normalized data can be trained to obtain greater improvements in convergence rate and forecast accuracy, and the conversion formula is expressed in (6):
where x is the unprocessed data sample, x min Is the minimum value, x, in the sample max Is the maximum value in the sample and,
for the normalized sample, after the prediction is completed, the prediction result needs to be subjected to inverse transformation processing to recover the original dimension of the data, and the inverse transformation calculation process is as shown in formula (7):
x * ＝(x max -x min )x pred +x min (7)
in the formula (7), x is a predicted value after being converted into the original dimension, and x pred Is a normalized predicted value;
according to the load prediction result of the power distribution area, the load distribution characteristics of each phase can be known, a long-term three-phase load automatic adjusting device installation plan is made according to the load distribution characteristics, and the power supply reliability and the economy are guaranteed.
The step (II) of the invention specifically comprises the following steps:
(1) Establishing a commutation model for governing three-phase unbalance
For the calculation of the three-phase unbalance degree on the outlet side of the distribution network transformer, corresponding acquired data are obtained firstly, and then the calculation can be carried out by using a formula (9);
in the formula: parameter beta A ，β B ，β C Respectively representing the unbalance degrees of the A phase, the B phase and the C phase; parameter I A ，I B ，I C The current values of A, B and C phase are respectively represented, because the current transformer in the actual distribution network has the limitation of switching times, the minimum switching number is adopted to improve the operation period of the system and reduce the investment cost, Y (i) epsilon {0,1} is used for representing whether the ith switch acts for phase change, wherein the parameter 0 means the openingThe switch is not phase-changed, the meaning of parameter 1 is that the switch needs to be phase-changed, and the model of equation (10) is established:
further, the total number of phase changes of the switches can be calculated by using the formula (11), wherein i represents the number of the switches;
a (Y) is the total action times, and an objective function with the minimum commutation times is constructed, as shown in a formula (11);
δ 2 ＝min(A(Y)) (11)
establishing a multi-objective function model, as shown in formula (13), wherein the number of switching actions is the minimum, and the three-phase unbalance degree is the minimum;
according to the concrete conditions of the actual distribution network, after the phase change is carried out, the current of a single phase cannot exceed the maximum load, and the three-phase imbalance reaches the state specified level, so that corresponding constraint conditions are constructed, as shown in formula (13):
wherein, b max Maximum degree of unbalance for low-voltage distribution areas, b Preset value The critical value given for the system according to the maximum unbalance that the station can bear, I Current carrying capacity The maximum ampacity is given to the system;
in addition to the above multiple targets and multiple constraints, a new genetic algorithm, i.e., a "cultural gene" method, is selected, and the number of switching operations is limited for each solution in consideration of whether the switch can operate continuously, so that a weight function is introduced:
F＝[n,β MAX ,K Cmax ,w] (14)
in the weighting function, the parameter n represents the number of phase change switches needing to be actuated; parameter(s) MAX Representing the maximum unbalance degree of three phases; parameter K Cmax The maximum action times of a single commutation switch in a representative commutation period; the preset weight represented by the parameter w is introduced into the weight function, the content of a double-objective function is converted into a triple-objective function, wherein the value of the newly added objective parameter is K Cmax ∈[0，1，2]The maximum action frequency of a single phase change switch in a phase change period is 2;
(2) Encoding
The cultural gene algorithm firstly needs to carry out coding operation on specific problems, abstracts the specific problems into chromosome variables, and mainly aims at arranging and combining the specific problems according to a certain rule and then carrying out gene evolution, the algorithm is generally abstract and does not need to carry out calculation in a solution space, if a coding rule is not properly selected, the following calculation result can be greatly influenced, the three-phase imbalance management of a power distribution area is carried out, the phase sequence of the terminal connection is coded by the automatic load regulation device according to the rule of A phase [1, 0], B phase [0, 1, 0] and C phase [0, 1] to obtain the chromosome variables, and an initial matrix, namely a single individual, namely a chromosome sequence is generated according to the number of execution terminals, and each column represents the phase sequence of the current number execution terminal;
(3) Algorithm initialization
Randomly generating a population related to a chromosome sequence, namely a first generation population, continuously carrying out iterative evolution according to an algorithm until the evolution cannot be continued, stopping the evolution at the moment, supposing that the number of individuals in an initialized population is N, 20-50 for the individuals, obtaining the optimal individual through a hill climbing algorithm and called an intelligent agent, and when the number of the individuals is more, the convergence speed of the algorithm is higher, generally 5-8, and setting N ag Number of agents, N pub The number of common individuals is as follows:
N＝N ag +N pub (15)
(4) Computing
The fitness function value is mainly used for evaluating whether a solution is optimal or not and whether an individual value is optimal or not after local search and global search, different fitness functions can be selected for different problems, when the fitness function is selected, the characteristics of a problem to be optimized must be known, in a cultural genetic algorithm, the selection of the fitness function is a difficult point, if the optimization problem is the solution of an extreme value or a minimum value, a target function is directly used as the fitness function for searching, so that the optimal solution is found, and the cultural genetic algorithm has the maximum characteristic that different fitness functions are selected for different problems;
(5) Evolution operations
In the cultural gene algorithm, the evolution operation mainly utilizes the evolution operator to ensure that offspring can inherit the advantages of the previous generation of gene, so that each iteration can obtain a better value.
The step (5) of the present invention specifically comprises;
1) The crossover operation is that the random chromosomes of the previous generation are combined together, then a part of genes are exchanged to form a new individual, the crossover operation plays an important role in a cultural gene algorithm, and more individuals can be obtained, so that the probability of obtaining the best solution is greatly improved;
2) The method comprises the following steps of performing mutation operation, wherein the mutation operation mainly means that an optimal solution cannot be obtained through crossing, or local loop iteration or local optimization causes the algorithm to stop, and mutation operation is performed in order to obtain the optimal cultural genetic algorithm, wherein the mutation operation mainly causes each individual to be mutated to generate new individuals, and more new individuals appear in the whole group along with the increase of the number of the individuals;
3) Selecting operation, wherein the selecting operation is mainly used for simulating the evolution of organisms, a large number of new individuals appear in each operation, better individuals can be reserved through the selecting operation, and the two operations are continuously executed, so that more individuals are obtained, and the optimal solution is achieved.
The method has the advantages that short-term load prediction is carried out on the low-voltage distribution network by considering the distribution condition of single-phase loads, the number of the three-phase load automatic adjusting devices which are connected is determined according to the prediction result, and the economy is improved on the premise of ensuring three-phase balance. In addition, the minimum number of switching actions and the minimum three-phase unbalance degree are taken as objective functions, an intelligent algorithm is utilized to quickly find an optimal solution in the objective functions, the action times of the three-phase load automatic adjusting device can be effectively reduced, the service life of the phase change switch is prolonged, the optimization problem can be effectively solved, the three-phase unbalance is reduced, the economy is improved, the scheme can be popularized and applied, and the comprehensive benefits of power supply companies are improved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the invention and do not constitute a limitation of the invention.
FIG. 1 is a schematic view of a low pressure station of the present invention;
FIG. 2 is a flow chart of the optimized scheduling of the zones of the present invention;
fig. 3 is a flow chart of the optimal configuration of the minimum number of switching actions of the station area.
Detailed Description
The details of the present invention and its embodiments are further described below with reference to the accompanying drawings.
Fig. 1 is a schematic diagram of a power distribution station area regulation and control system according to the present invention, wherein the three-phase load automatic regulation device should satisfy the following contents:
in a first aspect: according to the load size of the distribution area, the number of three-phase load automatic adjusting devices installed in the distribution area is determined, installation cost and subsequent maintenance cost are considered as much as possible, the load types and characteristics are analyzed, and reasonable installation is achieved.
In a second aspect: the three-phase load automatic adjusting device has limited action times, and the action times need to be considered in order to prolong the service life of the three-phase load automatic adjusting device.
Comprises the following steps:
the method comprises the following steps of (I) optimizing configuration of a three-phase load automatic adjusting device of a power distribution station area based on probability distribution;
based on the distribution condition of each single-phase load, under the condition of meeting the adjustment of load unbalance, the number of the three-phase load automatic adjusting devices is configured as few as possible so as to improve the economy, and the method specifically comprises the following steps:
(1) Analyzing the system data, identifying and correcting abnormal points of the data,
(2) Processing a data missing value;
(3) The data is subjected to standardized processing to facilitate data analysis;
(4) And carrying out reasonable configuration of the three-phase load automatic adjusting device according to the system data analysis result.
The data anomaly point identification and correction in the step (1) comprises the following steps:
because the power load has a periodic variation trend, the difference between the loads of the working day and the rest day in the same time is not too large, if the difference between the two is over a specified threshold, the data may be abnormal data, care should be taken when handling the outlier, otherwise misjudgment may be caused, and data distortion may be caused, and for the wrong data, there are generally two processing methods: transverse processing and longitudinal processing;
transverse processing: the load data is usually a continuous smooth sequence without abrupt changes, so that there is no significant difference between the current load data and the adjacent load data, and if the load data at a certain moment exceeds a predetermined threshold, there is a relationship in equation (1):
then, the horizontal processing may be performed by using an averaging method, and the specific processing manner is as shown in formula (2):
wherein, alpha (t), beta (t) are threshold values, x (k, t) is the load value at the time t on the kth day;
longitudinal treatment: the power load has obvious periodic variation, and as known from the data period law, the load values of the same date and type at the same time are relatively close, and if the difference between the load values of the two exceeds a preset threshold value, the relation in the formula (3) exists:
|X(k,t)-M(t)|>γ(t) (3)
the vertical processing method may be used to correct the out-of-range abnormal data, and the correction process is shown in equation (4):
wherein, M (t) is the average value of recent days in the historical load data, M (t) is a prescribed threshold value, and X (k, t)' is the load data after vertical processing;
the data missing value processing in the step (2) of the invention comprises the following steps:
there are three methods for handling data loss: deleting, supplementing and ignoring the power load data, wherein the power load data are generally processed by a supplementing method, the missing data part is generally filled according to loads of similar dates or adjacent dates, and a specific repair formula is as follows:
X(d,t)＝ω 1 X(d,t-1)+ω 2 X(d,t+1)+ω 3 X(d-1,t)+ω 4 X(d+1,t) (5)
wherein X (d, t) represents the electric load value at the time t on the d day; w is a i (i =1, 2, 3, 4) is a weighting coefficient;
the step (3) of data standardization processing of the invention comprises the following steps:
the number of electric loads is often very large, if the electric loads are directly introduced into the system, the convergence is reduced, so that the data of the loads need to be normalized to avoid rapid deterioration, when the digital type variables are normalized, a Min-Max method is used, after the data are normalized according to an equation (6), the distribution of the data is limited to 0-1, the normalized method can shorten the data quantity of a sample and reduce the timeliness of a training mode, the normalized data can be trained to obtain greater improvement in the convergence rate and the forecast accuracy, and the conversion formula is expressed in (6):
where x is the unprocessed data sample, x min Is the minimum value, x, in the sample max Is the maximum value in the sample or samples,
for the normalized sample, after the prediction is completed, the prediction result needs to be subjected to inverse transformation processing to recover the original dimension of the data, and the inverse transformation calculation process is as shown in formula (7):
x * ＝(x max -x min )x pred +x min (7)
in the formula (7), x is a predicted value after being converted into the original dimension, and x pred Is a normalized predicted value;
according to the load prediction result of the power distribution area, the load distribution characteristics of each phase can be known, a long-term three-phase load automatic adjusting device installation plan is made according to the load distribution characteristics, the power supply reliability is guaranteed, meanwhile, the economy is guaranteed, and a basic flow chart is shown in fig. 2.
(II) minimizing the optimal scheduling method of the commutation switch of the action cost;
when three phases are unbalanced, the corresponding three-phase load automatic adjusting device is required to be used; setting a target, operating as much as possible in a short period of time to achieve three-phase balance, specifically comprising:
(1) Establishing a commutation model for governing three-phase unbalance
For the calculation of the three-phase unbalance degree at the outlet side of the distribution network transformer, corresponding acquired data are obtained firstly, and then the calculation can be carried out by using a formula (9);
in the formula: parameter beta A ，β B ，β C Respectively representing the unbalance degrees of the A phase, the B phase and the C phase; parameter I A ，I B ，I C Respectively representing phase current values of A, B and C, adopting the minimum switching number to improve the operation period of a system and reduce investment cost because a current transformer in an actual power distribution network has the limitation of the switching times, and utilizing Y (i) epsilon {0,1} to represent whether an ith switch acts for phase commutation or not, wherein the meaning of a parameter 0 is that the switch does not perform the phase commutation, the meaning of a parameter 1 is that the switch needs to perform the phase commutation, and establishing a model of a formula (10):
further, the formula (11) can be used to calculate the total phase change times of the switches, i represents the number of the switches;
a (Y) is the total action times, and an objective function with the minimum commutation times is constructed, as shown in a formula (11);
δ 2 ＝min(A(Y)) (11)
establishing a multi-objective function model, as shown in formula (13), wherein the number of switching actions is the minimum, and the three-phase unbalance degree is the minimum;
according to the concrete conditions of the actual distribution network, after the phase change is carried out, the current of a single phase cannot exceed the maximum load, and the three-phase imbalance reaches the state specified level, so that corresponding constraint conditions are constructed, as shown in formula (13):
wherein, b max Maximum unbalance of the low-voltage distribution area, b Preset value The critical value given for the system according to the maximum unbalance that the station can bear, I Current carrying capacity The maximum ampacity is given to the system;
the invention selects a new genetic algorithm, namely a 'cultural gene' method, and in addition to the multiple targets and the multiple limits, the invention considers whether the switcher can operate continuously or not and limits the number of switching operations each time the switcher is solved, so that a weight function is introduced:
F＝[n,β MAX ,K Cmax ,w] (14)
in the weighting function, the parameter n represents the number of phase change switches needing to be actuated; parameter(s) MAX Representing the maximum unbalance degree of three phases; parameter K Cmax The maximum action times of a single commutation switch in a representative commutation period; the preset weight represented by the parameter w is introduced into the weight function, the content of a double-objective function is converted into a triple-objective function, wherein the value of the newly added objective parameter is K Cmax ∈[0，1，2]The maximum action times of a single phase change switch in a phase change period is 2;
(2) Coding
The culture gene algorithm firstly needs to carry out coding operation on specific problems, abstracts the specific problems into chromosome variables, and mainly aims at arranging and combining the specific problems according to a certain rule and then carrying out gene evolution, the algorithm is generally abstract and does not need to carry out calculation in a solution space, if a coding rule is not properly selected, the subsequent calculation result is greatly influenced, the three-phase imbalance of a power distribution area is treated, a three-phase load automatic adjusting device is coded according to the rules of A phases [1, 0], B phases [0, 1, 0] and C phases [0, 0 and 1] to execute the phase sequence of terminal connection, so that the chromosome variables are obtained, and an initial matrix, namely a single individual, namely a chromosome sequence, is generated according to the number of execution terminals in sequence, and each column represents the phase sequence of the current number execution terminal;
(3) Algorithm initialization
Randomly generating a population related to a chromosome sequence, namely a first generation population, continuously carrying out iterative evolution according to an algorithm until the evolution cannot be continued, stopping the evolution at the moment, assuming that the number of individuals in an initialized population is N, 20-50 for the individuals, obtaining the optimal individual through a hill climbing algorithm and called an agent, and when the number of the individuals is more, the convergence speed of the algorithm is faster, generally 5-8, and setting N ag Number of agents, N pub The number of common individuals is as follows:
N＝N ag +N pub (15)
(4) Calculating out
The fitness function value is mainly used for evaluating whether a solution is optimal or not and whether an individual value is optimal or not after local search and global search, different fitness functions can be selected for different problems, when the fitness function is selected, the characteristics of a problem to be optimized must be known, in a cultural genetic algorithm, the selection of the fitness function is a difficult point, if the optimization problem is the solution of an extreme value or a minimum value, a target function is directly used as the fitness function for searching, so that the optimal solution is found, and the cultural genetic algorithm has the maximum characteristic that different fitness functions are selected for different problems;
(5) Evolution operations
In the cultural genetic algorithm, the evolution operation mainly utilizes an evolution operator to ensure that descendants can inherit the advantages of genes of the previous generation, so that better values can be obtained in each iteration, and the operations of crossing, mutation and selection are described in detail below;
1) The crossover operation is that the random chromosomes of the previous generation are combined together, then a part of genes are exchanged to form a new individual, the crossover operation plays an important role in a cultural gene algorithm, and more individuals can be obtained, so that the probability of obtaining the best solution is greatly improved;
2) The method comprises the following steps of performing mutation operation, wherein the mutation operation mainly means that an optimal solution cannot be obtained through crossing, or local loop iteration or local optimization causes the algorithm to stop, and mutation operation is performed in order to obtain the optimal cultural genetic algorithm, wherein the mutation operation mainly causes each individual to be mutated to generate new individuals, and more new individuals appear in the whole group along with the increase of the number of the individuals;
3) Selecting operation, wherein the selecting operation is mainly used for simulating the evolution of organisms, a large number of new individuals appear in each operation, better individuals can be reserved through the selecting operation, and the two operations are continuously executed, so that more individuals are obtained, and the optimal solution is achieved.
A cultural gene algorithm is introduced into the multi-objective optimization problem, two objective functions and a weight value required by the switching action times are added as fitness functions to evaluate whether the obtained individual is an optimal solution, and a basic flow chart is shown in FIG. 3.
Claims (8)
1. An optimal configuration and scheduling method for a three-phase load automatic adjusting device in a power distribution area is characterized by comprising the following steps:
the method comprises the following steps of (I) optimizing configuration of a three-phase load automatic adjusting device of a power distribution station area based on probability distribution;
based on the distribution condition of each single-phase load, under the condition of meeting the adjustment of load unbalance, the number of the three-phase load automatic adjusting devices is configured as less as possible to improve the economy,
(II) minimizing the optimal scheduling method of the commutation switch of the action cost;
when three phases are unbalanced, the corresponding three-phase load automatic adjusting device is required to be used; a target is set to operate as much as possible in a short period of time to achieve three-phase equalization.
2. The method according to claim 1, wherein the step (one) includes:
(1) Analyzing the system data, identifying and correcting abnormal points of the data,
(2) Processing a data missing value;
(3) Data are processed in a standardized way so as to facilitate data analysis;
(4) And carrying out reasonable configuration of the three-phase load automatic adjusting device according to the system data analysis result.
3. The optimal configuration and dispatching method for the automatic adjusting device of three-phase load in the distribution substation according to claim 2, wherein the identification and correction of the abnormal point of data in the step (1) comprises:
because the power load has a periodic variation trend, the load difference between the adjacent working days and the rest days in the same time is not too large, if the difference between the two is over a specified threshold, abnormal data may be generated, care must be taken when handling outliers, otherwise misjudgment may be caused, and data distortion may be caused, and for the wrong data, there are generally two processing methods: transverse processing and longitudinal processing;
transverse processing: the load data is usually a continuous smooth sequence without abrupt changes, so that there is no significant difference between the current load data and the adjacent load data, and if the load data at a certain moment exceeds a predetermined threshold, there is a relationship in equation (1):
then, the horizontal processing may be performed by using an averaging method, and the specific processing manner is as shown in formula (2):
in the formula, alpha (t) and beta (t) are threshold values, and x (k, t) is a load value at the time t on the k day;
longitudinal treatment: the power load has obvious periodic variation, and as can be known from the data period law, the load values of the same date type at the same time are relatively close, and if the difference between the load values of the two exceeds a preset threshold value, the relation in the formula (3) exists:
|X(k,t)-M(t)|>γ(t) (3)
the out-of-range anomaly data can be corrected using a vertical processing method, as shown in equation (4):
in the formula, M (t) is an average value of recent days in the historical load data, M (t) is a predetermined threshold, and X (k, t)' is load data after vertical processing.
4. The optimal configuration and dispatching method for the automatic adjusting device of three-phase load in the distribution substation according to claim 2, wherein the data missing value processing in the step (2) comprises:
there are three methods for handling data loss: deleting, supplementing and ignoring the power load data, wherein the power load data are generally processed by a supplementing method, the missing data part is generally filled according to loads of similar dates or adjacent dates, and a specific repair formula is as follows:
X(d,t)＝ω 1 X(d,t-1)+ω 2 X(d,t+1)+ω 3 X(d-1,t)+ω 4 X(d+1,t) (5)
wherein X (d, t) represents the power load value at the time t on the d day; w is a i (i =1, 2, 3, 4) is a weighting coefficient.
5. The optimal configuration and dispatching method for the automatic three-phase load adjusting device of the power distribution station area according to claim 2, wherein the step (3) of data standardization processing comprises:
the number of electric loads is often very large, if the electric loads are directly introduced into the system, the convergence is reduced, so that the data of the loads need to be normalized to avoid rapid deterioration, when the digital type variables are normalized, a Min-Max method is used, after the data are normalized according to an equation (6), the distribution of the data is limited to 0-1, the normalized method can shorten the data quantity of a sample and reduce the timeliness of a training mode, the normalized data can be trained to obtain greater improvement in the convergence rate and the forecast accuracy, and the conversion formula is expressed in (6):
where x is the unprocessed data sample, x min Is the minimum value, x, in the sample max Is the maximum value in the sample or samples,
for the normalized sample, after the prediction is completed, the prediction result needs to be subjected to inverse transformation processing to recover the original dimension of the data, and the inverse transformation calculation process is as shown in formula (7):
x * ＝(x max -x min )x pred +x min (7)
in formula (7), x * For the prediction value, x, after conversion to the original dimension pred Is a normalized prediction value.
6. The optimal configuration and dispatching method for the automatic three-phase load adjusting device of the power distribution area as claimed in claim 2, wherein in the step (4), load distribution characteristics of each phase can be known according to the load prediction result of the power distribution area, and a long-term three-phase load automatic adjusting device installation plan is made according to the load distribution characteristics, so that power supply reliability is guaranteed, and meanwhile economy is guaranteed.
7. The optimal configuration and scheduling method of the automatic three-phase load adjusting device of the power distribution area according to claim 1, wherein the step (two) specifically comprises:
(1) Establishing a commutation model for governing three-phase unbalance
For the calculation of the three-phase unbalance degree on the outlet side of the distribution network transformer, corresponding acquired data are obtained firstly, and then the calculation can be carried out by using a formula (9);
in the formula: parameter beta A ，β B ，β C Respectively representing the unbalance degrees of the A phase, the B phase and the C phase; parameter I A ，I B ，I C Respectively representing phase current values of A, B and C, adopting the minimum switching number to improve the operation period of a system and reduce investment cost because a current transformer in an actual power distribution network has the limitation of the switching times, and utilizing Y (i) epsilon {0,1} to represent whether an ith switch acts for phase commutation or not, wherein the meaning of a parameter 0 is that the switch does not perform the phase commutation, the meaning of a parameter 1 is that the switch needs to perform the phase commutation, and establishing a model of a formula (10):
further, the total number of phase changes of the switches can be calculated by using the formula (11), wherein i represents the number of the switches;
a (Y) is the total action times, and an objective function with the minimum commutation times is constructed, as shown in a formula (11);
δ 2 ＝min(A(Y)) (11)
establishing a multi-objective function model, as shown in formula (13), wherein the number of switching actions is the minimum, and the three-phase unbalance degree is the minimum;
according to the concrete conditions of the actual distribution network, after the phase change is carried out, the current of a single phase cannot exceed the maximum load, and the three-phase imbalance reaches the state specified level, so that corresponding constraint conditions are constructed, as shown in formula (13):
wherein, b max Maximum degree of unbalance for low-voltage distribution areas, b Preset value The critical value given for the system according to the maximum unbalance that the station can bear, I Current carrying capacity A maximum ampacity is given for the system;
in addition to the above multiple objectives and multiple constraints, a new genetic algorithm, namely the "cultural gene" method, is selected, and the number of switching operations is limited each time the switcher is solved, so that a weight function is introduced:
F＝[n,β MAX ,K Cmax ,w] (14)
in the weighting function, the parameter n represents the number of phase change switches needing action; parameter(s) MAX Representing the maximum unbalance degree of three phases; parameter K Cmax The maximum action times of a single commutation switch in a representative commutation period; the preset weight represented by the parameter w is introduced into the weight function, the content of a double-objective function is converted into a triple-objective function, wherein the value of the newly added objective parameter is K Cmax ∈[0，1，2]The maximum action times of a single phase change switch in a phase change period is 2;
(2) Encoding
The culture gene algorithm firstly needs to carry out coding operation on specific problems, abstracts the specific problems into chromosome variables, and mainly aims at arranging and combining the specific problems according to a certain rule and then carrying out gene evolution, the algorithm is generally abstract and does not need to carry out calculation in a solution space, if a coding rule is not properly selected, the subsequent calculation result is greatly influenced, the three-phase imbalance of a power distribution area is treated, a three-phase load automatic adjusting device is coded according to the rules of A phases [1, 0], B phases [0, 1, 0] and C phases [0, 0 and 1] to execute the phase sequence of terminal connection, so that the chromosome variables are obtained, and an initial matrix, namely a single individual, namely a chromosome sequence, is generated according to the number of execution terminals in sequence, and each column represents the phase sequence of the current number execution terminal;
(3) Algorithm initialization
Randomly generating a population related to a chromosome sequence, namely a first generation population, continuously carrying out iterative evolution according to an algorithm until the evolution cannot be continued, stopping the evolution at the moment, assuming that the number of individuals in an initialized population is N, 20-50 for the individuals, obtaining the optimal individual through a hill climbing algorithm and called an agent, and when the number of the individuals is more, the convergence speed of the algorithm is faster, generally 5-8, and setting N ag Number of agents, N pub The number of common individuals is as follows:
N＝N ag +N pub (15)
(4) Computing
The fitness function value is mainly used for evaluating whether a solution is optimal or not and whether an individual value is optimal or not after local search and global search, different fitness functions can be selected for different problems, when the fitness function is selected, the characteristics of a problem to be optimized must be known, in a cultural genetic algorithm, the selection of the fitness function is a difficult point, if the optimization problem is the solution of an extreme value or a minimum value, a target function is directly used as the fitness function for searching, so that the optimal solution is found, and the cultural genetic algorithm has the maximum characteristic that different fitness functions are selected for different problems;
(5) Evolutionary operations
In the cultural gene algorithm, the evolution operation mainly utilizes the evolution operator to ensure that offspring can inherit the advantages of the previous generation of gene, so that each iteration can obtain a better value.
8. The optimal configuration and scheduling method of the automatic three-phase load adjusting device of the power distribution substation according to claim 7, wherein the step (5) specifically comprises;
1) The crossover operation is that the random chromosomes of the previous generation are combined together, then a part of genes are exchanged to form a new individual, the crossover operation plays an important role in a cultural gene algorithm, and more individuals can be obtained, so that the probability of obtaining the best solution is greatly improved;
2) The method comprises the following steps of performing mutation operation, wherein the mutation operation mainly means that an optimal solution cannot be obtained through crossing, or local loop iteration or local optimization causes the algorithm to stop, and mutation operation is performed in order to obtain the optimal cultural genetic algorithm, wherein the mutation operation mainly causes each individual to be mutated to generate new individuals, and more new individuals appear in the whole group along with the increase of the number of the individuals;
3) Selecting operation, wherein the selecting operation is mainly used for simulating the evolution of organisms, a large number of new individuals appear in each operation, better individuals can be reserved through the selecting operation, and the two operations are continuously executed, so that more individuals are obtained, and the optimal solution is achieved.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211619638.4A CN115800330A (en) | 2022-12-19 | 2022-12-19 | Optimal configuration and scheduling method of automatic three-phase load adjusting device in power distribution area |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211619638.4A CN115800330A (en) | 2022-12-19 | 2022-12-19 | Optimal configuration and scheduling method of automatic three-phase load adjusting device in power distribution area |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115800330A true CN115800330A (en) | 2023-03-14 |
Family
ID=85425268
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211619638.4A Pending CN115800330A (en) | 2022-12-19 | 2022-12-19 | Optimal configuration and scheduling method of automatic three-phase load adjusting device in power distribution area |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115800330A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117519054A (en) * | 2023-12-11 | 2024-02-06 | 广州智业节能科技有限公司 | High-efficient cold station control system |
-
2022
- 2022-12-19 CN CN202211619638.4A patent/CN115800330A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117519054A (en) * | 2023-12-11 | 2024-02-06 | 广州智业节能科技有限公司 | High-efficient cold station control system |
| CN117519054B (en) * | 2023-12-11 | 2024-06-11 | 广州智业节能科技有限公司 | High-efficient cold station control system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108776869B (en) | Transformer three-phase unbalance management method for transformer in transformer area based on big data of intelligent electric meter | |
| CN111030188A (en) | Hierarchical control strategy containing distributed and energy storage | |
| CN109742776B (en) | Three-phase unbalanced transformer area user adjustment method based on sorting algorithm | |
| CN110198042B (en) | Dynamic optimization method for power grid energy storage and storage medium | |
| CN106803130B (en) | Planning method for distributed power supply to be connected into power distribution network | |
| CN111146821A (en) | DSTATCOM optimal configuration method considering photovoltaic uncertainty | |
| CN115800330A (en) | Optimal configuration and scheduling method of automatic three-phase load adjusting device in power distribution area | |
| Xu et al. | Stochastic multi-objective optimization of photovoltaics integrated three-phase distribution network based on dynamic scenarios | |
| CN116826752A (en) | Multi-objective low-carbon loss reduction optimization scheduling strategy method for energy consumption of transformer area | |
| CN111159619A (en) | Power distribution network planning method based on distributed power supply coordination mechanism | |
| Baghipour et al. | A hybrid algorithm for optimal location and sizing of capacitors in the presence of different load models in distribution network | |
| CN113888202A (en) | Training method and application method of electricity price prediction model | |
| Vítor et al. | Optimal Volt/Var control applied to modern distribution systems | |
| CN112202168A (en) | Multi-element power grid advanced control power supply method and system based on multi-objective coordination optimization | |
| CN112488367A (en) | User phase sequence loss reduction method and system based on quantum inheritance | |
| CN112001474A (en) | Power distribution terminal equipment optimal configuration method for power distribution network | |
| CN112701700B (en) | Multi-objective optimization-based three-phase imbalance treatment method and system for transformer area | |
| CN114759572A (en) | Reactive power compensation loss reduction method and device for user station area | |
| Hu et al. | Two-stage energy scheduling optimization model for complex industrial process and its industrial verification | |
| CN109345411B (en) | Quantitative control method applied to improvement of power supply capacity of power distribution network | |
| CN112241604A (en) | Probability matching series-parallel coupling multi-model power grid rainstorm disaster forecast correction method | |
| CN112584386A (en) | 5G C-RAN resource prediction and allocation method and system | |
| Wang et al. | Site selection and capacity determination of multi-types of dg in distribution network planning | |
| Shao et al. | Optimization method based on load forecasting for three-phase imbalance mitigation in low-voltage distribution network | |
| CN113949108B (en) | Power distribution network power regulation and control method with intelligent soft switch based on two-person zero-sum game |
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 |