CN117113086A - Energy storage unit load prediction method, system, electronic equipment and medium - Google Patents
Energy storage unit load prediction method, system, electronic equipment and medium Download PDFInfo
- Publication number
- CN117113086A CN117113086A CN202311066977.9A CN202311066977A CN117113086A CN 117113086 A CN117113086 A CN 117113086A CN 202311066977 A CN202311066977 A CN 202311066977A CN 117113086 A CN117113086 A CN 117113086A
- Authority
- CN
- China
- Prior art keywords
- energy storage
- load prediction
- storage unit
- sample
- data set
- 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
- 238000004146 energy storage Methods 0.000 title claims abstract description 100
- 238000000034 method Methods 0.000 title claims abstract description 36
- 238000012549 training Methods 0.000 claims abstract description 73
- 238000001514 detection method Methods 0.000 claims abstract description 18
- 230000002159 abnormal effect Effects 0.000 claims abstract description 15
- 230000006403 short-term memory Effects 0.000 claims abstract description 9
- 230000007787 long-term memory Effects 0.000 claims abstract description 8
- 230000015654 memory Effects 0.000 claims description 15
- 238000000926 separation method Methods 0.000 claims description 12
- 230000005856 abnormality Effects 0.000 claims description 10
- 238000004422 calculation algorithm Methods 0.000 claims description 10
- 238000004138 cluster model Methods 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 10
- 238000010606 normalization Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 4
- 230000001105 regulatory effect Effects 0.000 claims description 4
- 230000008859 change Effects 0.000 abstract description 4
- 238000012360 testing method Methods 0.000 description 11
- 239000011159 matrix material Substances 0.000 description 8
- 230000006870 function Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000010977 unit operation Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/10—Pre-processing; Data cleansing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/09—Supervised learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/003—Load forecast, e.g. methods or systems for forecasting future load demand
-
- 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
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Abstract
The invention provides a load prediction method, a system, electronic equipment and a medium of an energy storage unit, belonging to the field of energy storage, wherein the method comprises the following steps: acquiring a load prediction training data set; the load prediction training dataset includes a plurality of sample points over a first historical period; each sample point comprises power grid load demand data, energy storage unit load data and influence factor data; the influence factor data comprise temperature, time and date marks corresponding to the sample points; performing anomaly detection on sample points in the load prediction training data set, and removing the abnormal sample points to obtain a sample data set; performing iterative training on the long-term and short-term memory network based on the sample data set to obtain an energy storage unit load prediction model; and predicting the load data of the energy storage unit in a future set period by adopting an energy storage unit load prediction model according to the load prediction data set. The method improves the prediction precision of the load of the energy storage unit, and enables the energy storage unit to respond to the change of the power grid requirement better.
Description
Technical Field
The invention relates to the field of energy storage, in particular to an energy storage unit load prediction method, an energy storage unit load prediction system, electronic equipment and a medium based on time sequence characteristics and power grid requirements.
Background
With the rapid development of economy, the whole society has put higher and higher requirements on power supply reliability. Under the current situation of the strong development of new energy, the output power of the unit containing the energy storage unit can be more conveniently regulated within a certain range, so that the unit has an important role in balancing the power change of the new energy and relieving the new energy consumption.
In recent years, as big data and artificial intelligence technologies are mature, machine learning and intelligent algorithms are widely applied in power plants. The method for predicting the unit load is also developed continuously, and the common unit modeling is mainly divided into two types: and performing mechanism modeling on a unit operation system through a unit operation mechanism, and establishing mapping relation between each input and each output by using unit operation data as a main research object through a data mining analysis method and using an intelligent algorithm. However, the scheme only selects main operation parameters of the unit for modeling the unit to construct a regression model of the unit load and the parameters, so that the operation state data of the real-time computer unit is realized. Along with the continuous popularization of the energy storage unit and the acceleration of the intelligent progress of the power grid, the fluctuation and the randomness of data also enable the energy storage unit to operate by considering the source load and the power grid demand, and the power grid demand has stronger time sequence, such as the early and late load demand, and the power grid demand of the working day and the holiday have great difference, so that the time sequence characteristic is also required to be considered for the operation of the energy storage unit.
Therefore, the traditional unit regression model modeling has the problems that the accuracy is poor and the power grid requirement change cannot be adapted under the condition that the influence of the factors is not considered.
Disclosure of Invention
The invention aims to provide a method, a system, electronic equipment and a medium for predicting the load of an energy storage unit, which can improve the prediction precision of the load of the energy storage unit and enable the load to better respond to the change of the power grid demand.
In order to achieve the above object, the present invention provides the following solutions:
a load prediction method of an energy storage unit comprises the following steps:
acquiring a load prediction training data set; the load prediction training dataset includes a plurality of sample points over a first historical period; each sample point comprises power grid load demand data, energy storage unit load data and influence factor data; the influence factor data comprise temperature, time and date marks corresponding to the sample points;
performing anomaly detection on sample points in the load prediction training data set, and removing the abnormal sample points to obtain a sample data set;
performing iterative training on the long-term and short-term memory network based on the sample data set to obtain an energy storage unit load prediction model;
according to the load prediction data set, predicting the load data of the energy storage unit in a future set period by adopting the load prediction model of the energy storage unit; the load prediction dataset comprising a plurality of sample points of a second historical period; and the ending time in the second history period is the current time.
Optionally, performing anomaly detection on the sample points in the load prediction training data set, and removing the abnormal sample points to obtain a sample data set, which specifically includes:
performing density clustering on sample points in the load prediction training data set by adopting a density-based clustering algorithm, and determining the type of each sample point; the type of the sample point is a core point, a boundary point or a noise point;
removing sample points with noise points as the types in the load prediction training data set to obtain a preliminary sample data set;
and carrying out normalization processing on the preliminary sample data set to obtain a sample data set.
Optionally, a density-based clustering algorithm is adopted to perform density clustering on sample points in the load prediction training data set, and the determining of the types of the sample points specifically comprises:
combining the neighborhood radius and the value of the least sample point randomly in a set value range to obtain a plurality of clustering models;
for any clustering model, clustering sample points in the load prediction training data set by adopting the clustering model, and determining the cohesion and separation degree of each sample point;
determining the contour coefficient of the clustering model according to the cohesion and separation of each sample point;
selecting an optimal clustering model from a plurality of clustering models according to the contour coefficients of the clustering models;
and clustering the sample points in the load prediction training data set by adopting the optimal clustering model, and determining the types of the sample points.
Optionally, the cohesion and separation of the sample points i are determined using the following formula:
wherein a (i) is the cohesive degree of the sample point i, b (i) is the separation degree of the sample point i, d (i, j) is the distance between the sample point i and the sample point j, |C i I is the number of sample points in the cluster to which the sample point i belongs, C i For the sample point set in the cluster to which sample point i belongs, |C k I is the number of sample points in the cluster to which the sample point k belongs, C k Is the set of sample points within the cluster to which sample point k belongs.
Optionally, the following formula is used to determine the profile coefficients of the cluster model:
s is the contour coefficient of the clustering model, n is the number of sample points in the load prediction training data set, and C is the load prediction training data set.
Optionally, the energy storage unit load prediction method further includes:
acquiring power grid load demand data in a future set period;
and determining the charge and discharge states and charge and discharge power of the energy storage units in the energy storage unit according to the difference value between the load data of the energy storage unit and the corresponding load demand data of the power grid in a set time period in the future so as to determine the regulating capacity of the energy storage unit.
In order to achieve the above purpose, the present invention also provides the following solutions:
an energy storage unit load prediction system, comprising:
the data acquisition unit is used for acquiring a load prediction training data set; the load prediction training dataset includes a plurality of sample points over a first historical period; each sample point comprises power grid load demand data, energy storage unit load data and influence factor data; the influence factor data comprise temperature, time and date marks corresponding to the sample points;
the abnormality detection unit is connected with the data acquisition unit and is used for detecting abnormality of sample points in the load prediction training data set and removing the abnormal sample points to obtain a sample data set;
the training unit is connected with the abnormality detection unit and is used for carrying out iterative training on the long-term and short-term memory network based on the sample data set so as to obtain an energy storage unit load prediction model;
the prediction unit is connected with the training unit and is used for predicting the load data of the energy storage unit in a future set period by adopting the load prediction model of the energy storage unit according to the load prediction data set; the load prediction dataset comprising a plurality of sample points of a second historical period; and the ending time in the second history period is the current time.
In order to achieve the above purpose, the present invention also provides the following solutions:
an electronic device comprising a memory and a processor, the memory being configured to store a computer program, the processor being configured to run the computer program to cause the electronic device to perform the energy storage unit load prediction method described above.
In order to achieve the above purpose, the present invention also provides the following solutions:
a computer readable storage medium storing a computer program which when executed by a processor implements the energy storage unit load prediction method described above.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
according to the method, in the prediction of the load data of the energy storage unit, the power grid load demand data is combined, the influences of temperature, time and date marks are added, the abnormal detection is carried out on the sample points in the load prediction training data set, the abnormal sample points are removed, the influence of the abnormal data on the training of the load prediction model of the energy storage unit is reduced, the prediction precision of the load prediction model of the energy storage unit is improved, and the capacity of the energy storage unit for responding to the power grid demand in different time sections at different times is further improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the drawings that are needed in the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a load prediction method of an energy storage unit provided by the invention;
FIG. 2 is a flow chart of a process for load prediction training data sets;
FIG. 3 is a schematic diagram of a long and short term memory network memory module;
FIG. 4 is a schematic diagram of a construction process of a load prediction model of an energy storage unit;
fig. 5 is a schematic diagram of the load prediction system of the energy storage unit provided by the invention.
Symbol description:
1-data acquisition unit, 2-anomaly detection unit, 3-training unit, 4-prediction unit.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention aims to provide a load prediction method, a system, electronic equipment and a medium for an energy storage unit, which are used for rapidly and accurately predicting the adjustment capacity of the energy storage unit by fully mining historical information and utilizing a big data analysis method.
In order that the above-recited objects, features and advantages of the present invention will become more readily apparent, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description.
Example 1
As shown in fig. 1, the present embodiment provides a load prediction method of an energy storage unit, including:
step 100: a load prediction training dataset is obtained. The load prediction training dataset includes a plurality of sample points over a first historical period of the power system. Each sample point comprises power grid load demand data, energy storage unit load data and influence factor data. The influence factor data comprises temperature, time and date marks of corresponding sample points. The date mark comprises month, date, number of weeks, whether the date mark is a working day or not, and the like.
Step 200: and carrying out anomaly detection on the sample points in the load prediction training data set, and removing the abnormal sample points to obtain a sample data set.
Specifically, since partial error values exist in the load prediction training data set to cause data set abnormality, the present invention uses Density-based clustering (Density-Based Spatial Clustering of Applications withNoise, DBSCAN) algorithm to perform abnormality detection on sample points in the load prediction training data set, remove abnormal data, and perform preprocessing. Step 200 comprises:
(1) And carrying out density clustering on the sample points in the load prediction training data set by adopting a density-based clustering algorithm, and determining the types of the sample points. The sample points are of the type core points, boundary points or noise points.
Further, as shown in fig. 2, in the set value range, values of the neighborhood radius and the least sample point are randomly combined to obtain a plurality of cluster models. In this embodiment, the iteration range of the neighborhood radius is set to be [1,30], the step length is 2, the iteration range of the neighborhood radius contains the iteration of the least sample point, the range is set to be [10,100], the step length is set to be 5, and the values of the neighborhood radius and the least sample point are combined to obtain a plurality of cluster models.
And clustering the sample points in the load prediction training data set by adopting the clustering model aiming at any clustering model, and determining the cohesion and separation of each sample point. In this embodiment, the cohesive degree and the separation degree of the sample point i are determined using the following formulas:
wherein a (i) is the cohesive degree of the sample point i, namely the average distance from the sample point i to the rest points in the cluster, b (i) is the separation degree of the sample point i, namely the minimum value of the average distance from the sample point i to the other sample points in each cluster, d (i, j) is the distance from the sample point i to the sample point j, and C i I is the number of sample points in the cluster to which the sample point i belongs, C i For the sample point set in the cluster to which sample point i belongs, |C k I is the number of sample points in the cluster to which the sample point k belongs, C k Is the set of sample points within the cluster to which sample point k belongs.
According to the cohesion degree and the separation degree of each sample point, determining the contour coefficient of the clustering model:
s is the contour coefficient of the clustering model, n is the number of sample points in the load prediction training data set, and C is the load prediction training data set.
And selecting an optimal cluster model from the plurality of cluster models according to the contour coefficients of the cluster models. Specifically, the cluster model with the largest contour coefficient is the optimal cluster model. The neighborhood radius of the optimal cluster model in this embodiment is 10, and the minimum sample point is 45.
And clustering the sample points in the load prediction training data set by adopting the optimal clustering model, and determining the types of the sample points.
The core points are points that contain more than the minimum number of sample points within the neighborhood radius. Boundary points are points within the neighborhood radius for which the number of points is less than the minimum sample point, but which fall within the neighborhood radius of the core point. The noise point is a point that is neither a core point nor a boundary point.
(2) And removing the sample points with the noise points as the types in the load prediction training data set to obtain a preliminary sample data set. Further, preprocessing the data set with noise points removed, and combining the influence factor data with the power grid load demand data and the energy storage unit load data according to the corresponding relation in time to form a preliminary sample data set containing time sequence variables, power grid demands and energy storage unit loads.
(3) Because the primary sample data set has larger difference in magnitude of the load data of the energy storage unit, the load demand data of the power grid and other relevant influence factor data, if the primary sample data set is not processed, the influence of parameters of smaller magnitude is ignored, and the training efficiency of the load prediction model of the energy storage unit is also influenced. In addition, the convergence speed of the algorithm can be improved by adding data normalization, and the training efficiency of the model is improved. Therefore, the preliminary sample data set is normalized to obtain a sample data set.
The formula of normalization processing is:
wherein y is t ′ est_i The normalized sample points; y is test_i For sample points in the preliminary sample dataset, y test_max For maximum value in preliminary sample dataset, y test_min Is the minimum in the preliminary sample dataset.
Step 300: and carrying out iterative training on the long-term and short-term memory network based on the sample data set to obtain the load prediction model of the energy storage unit.
Specifically, 70% of the sample data set was taken as the training data set and 30% was taken as the test data set. And (3) building a load prediction model of the energy storage unit by using a training data set and adopting a Long Short-Term Memory (LSTM), and adjusting internal parameters of the network.
And taking the power grid load demand data, the energy storage unit load data and the influence factor data before the t moment in the training data set as input of the long-period memory network, taking the energy storage unit load prediction data at the t moment as output of the long-period memory network, taking the energy storage unit load data at the t moment as a label, and carrying out iterative training on the long-period memory network.
The memory module of the LSTM is shown in fig. 3 and includes a forget gate, an input gate, and an output gate. Input x in forgetting gate at time t t Output S of state memory module at time t-1 t-1 Intermediate variable h at time t-1 t-1 Together determine the forgotten portion of the state memory module. Input x in input gate at time t t And respectively processing by a sigmoid function and a tanh function, and then jointly calculating a retention vector in the memory module. Intermediate variable h at time t t From output S of state memory module at time t t Output o from output gate at time t t And (5) jointly determining. The corresponding calculation process is as follows:
i t ＝σ(W ix x t +W ih h t-1 +b i )；
g t ＝tanh(W gx x t +W gh h t-1 +b g )；
f t ＝σ(W fx x t +W fh h t-1 +b f )；
o t ＝σ(W ox x t +W oh h t-1 +b o )；
S t ＝g t ⊙i t +S t-1 ⊙f t-1 ；
h t ＝tanh(S t )⊙o t ；
wherein x is t For input at time t, i t For inputting the first output of the gate at time t, g t A second output of the input gate at time t, f t For t moment forget the output of the gate, o t For outputting the output of the gate at the time t, S t For the output of the state memory module at time t, b i A first bias vector of the input gate at the moment t, b g A second bias vector of the input gate at the moment t, b f Bias vector of forgetting gate at t moment, b o Outputting the bias vector of the gate at the time t, W ix For inputting information to the first weight matrix of the input gate, W gx A second weight matrix for inputting information to the input gate, W fx For inputting information into the weight matrix of the forgetting gate, W ox For inputting information to the weight matrix of the output gate, W ih A first weight matrix for inputting gate to intermediate variable, W gh A second weight matrix for inputting gate to intermediate variable, W fh Weight matrix for forgetting gate to intermediate variable, W oh To output a weight matrix of gate to intermediate variables, h t-1 For the intermediate variable at time t-1, σ (. Cndot.) represents the sigmoid function, tanh (. Cndot.) represents the hyperbolic tangent function, and ". Cndot.) represents the bitwise multiplication of vector elements.
And evaluating the error precision of the load prediction model of the energy storage unit by using the test data set, and testing the prediction capability of the load prediction model of the energy storage unit. Test set data X with normalized parameters of training set test Normalizing, inputting the normalized data into an energy storage unit load prediction model for calculation, and performing inverse normalization to obtain a prediction result Y test_pre . Will predict result Y test_pre Tag Y in a test dataset test By comparison with an absolute error MAE, a root mean square error RMSE and a correlation coefficient R 2 Evaluating, wherein the corresponding formula is as follows;
wherein M is the number of sample points in the test data set, Y test,m Label being the mth sample point, Y test_pre,m As a result of the prediction of the mth sample point,is the average of all sample points in the test dataset.
Fig. 4 is a schematic diagram of a construction process of the load prediction model of the energy storage unit.
Further, the invention selects the 2018 4 month history data of a certain power plant for testing, and calculates MAE=0.7022, RMSE=0.9867 and R 2 The accuracy of the model is higher, the prediction value within 24 hours almost coincides with the actual value, the prediction effect of the model is better, and the validity of the invention is verified.
Step 400: and predicting the load data of the energy storage unit in a future set period by adopting the load prediction model according to the load prediction data set. The load prediction dataset includes a plurality of sample points of a second historical period. And the ending time in the second history period is the current time.
Further, the energy storage unit load prediction method further comprises the following steps:
step 500: and acquiring power grid load demand data in a future set period.
Step 600: and determining the charge and discharge states and charge and discharge power of the energy storage units in the energy storage unit according to the difference value between the load data of the energy storage unit and the corresponding load demand data of the power grid in a set time period in the future so as to determine the regulating capacity of the energy storage unit.
In the prediction of the load data of the energy storage unit, the power grid load demand data is combined, the influence of time parameters such as month, week and moment is added, in addition, the DBSCAN algorithm is adopted to perform abnormal detection on the load prediction training data set, iterative optimization is performed on the parameters of the clustering model, the influence of the abnormal data on the training of the load prediction model of the energy storage unit is reduced, the prediction precision of the load prediction model of the energy storage unit is improved, and the capacity of the energy storage unit for responding to the power grid demand in different time sections is further improved.
Example two
In order to execute the corresponding method of the above embodiment to achieve the corresponding functions and technical effects, an energy storage unit load prediction system is provided below.
As shown in fig. 5, the load prediction system of the energy storage unit provided in this embodiment includes: a data acquisition unit 1, an abnormality detection unit 2, a training unit 3 and a prediction unit 4.
Wherein the data acquisition unit 1 is used for acquiring a load prediction training data set. The load prediction training dataset includes a plurality of sample points over a first historical period. Each sample point comprises power grid load demand data, energy storage unit load data and influence factor data. The influence factor data comprises temperature, time and date marks of corresponding sample points.
The anomaly detection unit 2 is connected with the data acquisition unit 1, and the anomaly detection unit 2 is used for performing anomaly detection on sample points in the load prediction training data set and removing the abnormal sample points to obtain a sample data set.
The training unit 3 is connected with the anomaly detection unit 2, and the training unit 3 is used for performing iterative training on the long-term and short-term memory network based on the sample data set so as to obtain an energy storage unit load prediction model.
The prediction unit 4 is connected with the training unit 3, and the prediction unit 4 is used for predicting the load data of the energy storage unit in a future set period by adopting the load prediction model of the energy storage unit according to the load prediction data set. The load prediction dataset comprising a plurality of sample points of a second historical period; and the ending time in the second history period is the current time.
Compared with the prior art, the energy storage unit load prediction system provided by the embodiment has the same beneficial effects as the energy storage unit load prediction method provided by the first embodiment, and is not described herein.
Example III
The embodiment provides an electronic device, including a memory and a processor, where the memory is configured to store a computer program, and the processor runs the computer program to enable the electronic device to execute the energy storage unit load prediction method of the first embodiment.
Alternatively, the electronic device may be a server.
In addition, the embodiment of the invention also provides a computer readable storage medium, which stores a computer program, and the computer program realizes the load prediction method of the energy storage unit of the first embodiment when being executed by a processor.
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other.
The principles and embodiments of the present invention have been described herein with reference to specific examples, the description of which is intended only to assist in understanding the methods of the present invention and the core ideas thereof; also, it is within the scope of the present invention to be modified by those of ordinary skill in the art in light of the present teachings. In view of the foregoing, this description should not be construed as limiting the invention.
Claims (9)
1. The energy storage unit load prediction method is characterized by comprising the following steps of:
acquiring a load prediction training data set; the load prediction training dataset includes a plurality of sample points over a first historical period; each sample point comprises power grid load demand data, energy storage unit load data and influence factor data; the influence factor data comprise temperature, time and date marks corresponding to the sample points;
performing anomaly detection on sample points in the load prediction training data set, and removing the abnormal sample points to obtain a sample data set;
performing iterative training on the long-term and short-term memory network based on the sample data set to obtain an energy storage unit load prediction model;
according to the load prediction data set, predicting the load data of the energy storage unit in a future set period by adopting the load prediction model of the energy storage unit; the load prediction dataset comprising a plurality of sample points of a second historical period; and the ending time in the second history period is the current time.
2. The method for predicting load of an energy storage unit according to claim 1, wherein the method for detecting abnormality of sample points in the load prediction training data set and removing the abnormal sample points to obtain a sample data set comprises the following steps:
performing density clustering on sample points in the load prediction training data set by adopting a density-based clustering algorithm, and determining the type of each sample point; the type of the sample point is a core point, a boundary point or a noise point;
removing sample points with noise points as the types in the load prediction training data set to obtain a preliminary sample data set;
and carrying out normalization processing on the preliminary sample data set to obtain a sample data set.
3. The method for predicting the load of the energy storage unit according to claim 2, wherein the density clustering is performed on the sample points in the load prediction training data set by adopting a density-based clustering algorithm, and the determining of the type of each sample point specifically comprises:
combining the neighborhood radius and the value of the least sample point randomly in a set value range to obtain a plurality of clustering models;
for any clustering model, clustering sample points in the load prediction training data set by adopting the clustering model, and determining the cohesion and separation degree of each sample point;
determining the contour coefficient of the clustering model according to the cohesion and separation of each sample point;
selecting an optimal clustering model from a plurality of clustering models according to the contour coefficients of the clustering models;
and clustering the sample points in the load prediction training data set by adopting the optimal clustering model, and determining the types of the sample points.
4. The method of claim 3, wherein the cohesion and separation of the sample points i are determined using the following formula:
wherein a (i) is the cohesive degree of the sample point i, b (i) is the separation degree of the sample point i, d (i, j) is the distance between the sample point i and the sample point j, |C i I is the number of sample points in the cluster to which the sample point i belongs, C i For the sample point set in the cluster to which sample point i belongs, |C k I is the number of sample points in the cluster to which the sample point k belongs, C k Is the set of sample points within the cluster to which sample point k belongs.
5. The method of claim 4, wherein the profile coefficients of the cluster model are determined using the following formula:
s is the contour coefficient of the clustering model, n is the number of sample points in the load prediction training data set, and C is the load prediction training data set.
6. The energy storage unit load prediction method according to claim 1, further comprising:
acquiring power grid load demand data in a future set period;
and determining the charge and discharge states and charge and discharge power of the energy storage units in the energy storage unit according to the difference value between the load data of the energy storage unit and the corresponding load demand data of the power grid in a set time period in the future so as to determine the regulating capacity of the energy storage unit.
7. An energy storage unit load prediction system, characterized in that the energy storage unit load prediction system comprises:
the data acquisition unit is used for acquiring a load prediction training data set; the load prediction training dataset includes a plurality of sample points over a first historical period; each sample point comprises power grid load demand data, energy storage unit load data and influence factor data; the influence factor data comprise temperature, time and date marks corresponding to the sample points;
the abnormality detection unit is connected with the data acquisition unit and is used for detecting abnormality of sample points in the load prediction training data set and removing the abnormal sample points to obtain a sample data set;
the training unit is connected with the abnormality detection unit and is used for carrying out iterative training on the long-term and short-term memory network based on the sample data set so as to obtain an energy storage unit load prediction model;
the prediction unit is connected with the training unit and is used for predicting the load data of the energy storage unit in a future set period by adopting the load prediction model of the energy storage unit according to the load prediction data set; the load prediction dataset comprising a plurality of sample points of a second historical period; and the ending time in the second history period is the current time.
8. An electronic device comprising a memory for storing a computer program and a processor that runs the computer program to cause the electronic device to perform the energy storage unit load prediction method of any one of claims 1 to 6.
9. A computer readable storage medium, characterized in that it stores a computer program, which when executed by a processor implements the energy storage unit load prediction method according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311066977.9A CN117113086A (en) | 2023-08-23 | 2023-08-23 | Energy storage unit load prediction method, system, electronic equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311066977.9A CN117113086A (en) | 2023-08-23 | 2023-08-23 | Energy storage unit load prediction method, system, electronic equipment and medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117113086A true CN117113086A (en) | 2023-11-24 |
Family
ID=88794129
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311066977.9A Pending CN117113086A (en) | 2023-08-23 | 2023-08-23 | Energy storage unit load prediction method, system, electronic equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117113086A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117353359A (en) * | 2023-12-05 | 2024-01-05 | 国网浙江省电力有限公司宁波供电公司 | Battery combined energy storage and power supply method and system |
-
2023
- 2023-08-23 CN CN202311066977.9A patent/CN117113086A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117353359A (en) * | 2023-12-05 | 2024-01-05 | 国网浙江省电力有限公司宁波供电公司 | Battery combined energy storage and power supply method and system |
| CN117353359B (en) * | 2023-12-05 | 2024-04-12 | 国网浙江省电力有限公司宁波供电公司 | Battery combined energy storage and power supply method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113962364B (en) | Multi-factor power load prediction method based on deep learning | |
| CN110414788B (en) | Electric energy quality prediction method based on similar days and improved LSTM | |
| CN107967542B (en) | Long-short term memory network-based electricity sales amount prediction method | |
| CN110163429B (en) | Short-term load prediction method based on similarity day optimization screening | |
| CN111814956B (en) | Multi-task learning air quality prediction method based on multi-dimensional secondary feature extraction | |
| CN111915092B (en) | Ultra-short-term wind power prediction method based on long-short-term memory neural network | |
| CN111027772A (en) | Multi-factor short-term load prediction method based on PCA-DBILSTM | |
| CN111241755A (en) | Power load prediction method | |
| CN116307652A (en) | Artificial intelligent resource allocation method for intelligent power grid | |
| CN117113086A (en) | Energy storage unit load prediction method, system, electronic equipment and medium | |
| CN112288157A (en) | Wind power plant power prediction method based on fuzzy clustering and deep reinforcement learning | |
| CN114330934A (en) | Model parameter self-adaptive GRU new energy short-term power generation power prediction method | |
| CN115470862A (en) | Dynamic self-adaptive load prediction model combination method | |
| CN113449919B (en) | Power consumption prediction method and system based on feature and trend perception | |
| CN114240687A (en) | Energy hosting efficiency analysis method suitable for comprehensive energy system | |
| CN113033898A (en) | Electrical load prediction method and system based on K-means clustering and BI-LSTM neural network | |
| CN112232570A (en) | Forward active total electric quantity prediction method and device and readable storage medium | |
| CN116826745A (en) | Layered and partitioned short-term load prediction method and system in power system background | |
| CN112149896A (en) | Attention mechanism-based mechanical equipment multi-working-condition fault prediction method | |
| CN113762591B (en) | Short-term electric quantity prediction method and system based on GRU and multi-core SVM countermeasure learning | |
| CN114862032B (en) | XGBoost-LSTM-based power grid load prediction method and device | |
| CN116307250A (en) | Short-term load prediction method and system based on typical daily feature selection | |
| CN116542701A (en) | Carbon price prediction method and system based on CNN-LSTM combination model | |
| CN116011655A (en) | Load ultra-short-term prediction method and system based on two-stage intelligent feature engineering | |
| CN114254828A (en) | Power load prediction method based on hybrid convolution feature extractor and GRU |
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 |