CN117595231B - Intelligent power grid distribution management system and method thereof - Google Patents
Intelligent power grid distribution management system and method thereof Download PDFInfo
- Publication number
- CN117595231B CN117595231B CN202311370220.9A CN202311370220A CN117595231B CN 117595231 B CN117595231 B CN 117595231B CN 202311370220 A CN202311370220 A CN 202311370220A CN 117595231 B CN117595231 B CN 117595231B
- Authority
- CN
- China
- Prior art keywords
- power grid
- data
- power
- distribution network
- information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000009826 distribution Methods 0.000 title claims abstract description 82
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000007726 management method Methods 0.000 claims abstract description 46
- 238000004458 analytical method Methods 0.000 claims abstract description 43
- 238000012544 monitoring process Methods 0.000 claims abstract description 23
- 238000007781 pre-processing Methods 0.000 claims abstract description 19
- 238000004364 calculation method Methods 0.000 claims abstract description 16
- 238000003062 neural network model Methods 0.000 claims abstract description 11
- 238000012795 verification Methods 0.000 claims description 65
- 230000008569 process Effects 0.000 claims description 23
- 230000002159 abnormal effect Effects 0.000 claims description 9
- 238000009499 grossing Methods 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 9
- 238000013528 artificial neural network Methods 0.000 claims description 7
- 230000008859 change Effects 0.000 claims description 4
- 238000013459 approach Methods 0.000 claims description 3
- 238000013135 deep learning Methods 0.000 claims description 3
- 230000005611 electricity Effects 0.000 claims description 3
- 239000000284 extract Substances 0.000 claims description 3
- 229940060587 alpha e Drugs 0.000 claims description 2
- 238000004146 energy storage Methods 0.000 claims description 2
- 230000001105 regulatory effect Effects 0.000 claims description 2
- 238000012549 training Methods 0.000 description 15
- 210000002569 neuron Anatomy 0.000 description 10
- 238000010586 diagram Methods 0.000 description 8
- 230000006870 function Effects 0.000 description 6
- 230000005540 biological transmission Effects 0.000 description 5
- 238000004590 computer program Methods 0.000 description 5
- 238000012423 maintenance Methods 0.000 description 4
- 239000002699 waste material Substances 0.000 description 4
- 238000005265 energy consumption Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 description 2
- 230000035772 mutation Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000033228 biological regulation Effects 0.000 description 1
- 210000004027 cell Anatomy 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 239000002737 fuel gas Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000003345 natural gas Substances 0.000 description 1
- 210000004205 output neuron Anatomy 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 230000001932 seasonal effect Effects 0.000 description 1
Abstract
The invention relates to the field of intelligent power grid distribution management, in particular to a system and a method for intelligent power grid distribution management. Monitoring power grid information, collecting key parameter data in a power grid, and preprocessing the obtained power grid data information; according to the acquired preprocessed power grid data information and recorded related factors, a load prediction neural network model is established, and future load demands are accurately predicted; by monitoring and acquiring relevant parameters of the power quality in the power grid, constructing a power quality analysis model, and identifying a power quality problem; and presetting the work efficiency and the energy utilization rate of the power grid distribution network based on the load prediction information and the power quality analysis result, and obtaining the corrected energy utilization rate and the corrected work efficiency of the power grid distribution network through calculation. The intelligent power grid distribution management system can provide intelligent power grid distribution management, reduce energy cost and improve stability and reliability of a power grid.
Description
Technical Field
The invention relates to the field of intelligent power grid distribution management, in particular to a system and a method for intelligent power grid distribution management.
Background
The power grid distribution network is a process of introducing power from a power transmission network to a user power utilization terminal and comprises links of power transmission, power distribution, power utilization terminal access and the like. In the power system, the power grid distribution network is the last ring in the power transmission and distribution process, and is also a key link for transmitting electric energy from a power plant to a user power terminal.
With the development of social economy, the power demand is continuously growing, and the traditional power grid architecture and management method cannot meet the complicated power operation demand. Therefore, the intelligent power grid distribution network is used as a further upgrade of the power system, has the characteristics of centralized control, intelligent energy management and the like, and an effective distribution network management method is needed to realize system optimization, and monitoring, control and management of electric energy flow.
However, the prior art has the following problems: the load prediction cannot be accurately performed, so that energy waste and surplus are caused; the electric energy quality is not monitored in place, so that the work efficiency of a power grid distribution system is low, and the risk of faults and damages of power equipment is accompanied; the efficiency optimization of the power grid distribution network management work cannot be realized, the cost of the whole power system is increased, and the sustainable development concept cannot be met.
Disclosure of Invention
The invention provides a distribution network management system of a smart power grid and a method thereof, which aim to improve the reliability and stability of the distribution network system of the power grid by predicting the load demand and combining the result of power quality analysis, adjust the power supply strategy in real time according to the load condition, reasonably allocate the power resources, optimize the energy utilization to the greatest extent, reduce the energy consumption, discover and correct the power quality problem in time and improve the working efficiency; meanwhile, the intelligent power grid distribution network management system and the intelligent power grid distribution network management method can provide intelligent management and scheduling, and realize efficient utilization of renewable energy sources.
The technical scheme of the invention is as follows:
A smart grid distribution network management system and a method thereof, comprising the following steps:
s1, monitoring power grid information, collecting key parameter data in a power grid, and preprocessing the obtained power grid data information;
S2, building a load prediction neural network model according to the acquired preprocessed power grid data information and recorded related factors, and accurately predicting future load demands;
s3, constructing an electric energy quality analysis model through monitoring and acquiring relevant parameters of the electric energy quality in the power grid, and identifying electric energy quality problems;
S4, presetting the work efficiency and the energy utilization rate of the power grid distribution network based on the load prediction information and the power quality analysis result, and obtaining the corrected energy utilization rate and the corrected work efficiency of the power grid distribution network through calculation so as to manage the power grid distribution network.
Further, the step S1 specifically includes:
when data preprocessing is carried out, a grid data set Di is formed according to the collected grid data information, and a historical grid data matrix is established B represents the record number of the historical power grid data, c represents the number in the record number of each historical power grid data; then checking the power grid data information; the device comprises a first-stage check unit, a second-stage check unit and a third-stage check unit.
Further, in the third-level check unit, the processed power grid data information is obtained according to the first-level check unit and the second-level check unit, the power grid data adjacent to the current check result h are taken out, a data matrix Z adjacent to the moment is obtained,The power grid data which are similar to the current check unit s are taken out to obtain a similar data matrix V,Then obtaining the verification result processed by the three-level verification unit through weighted average calculation Wherein w h represents the weighting coefficients for the grid data at the moment of approach; w s represents the weighting coefficients for the similar grid data; di a EMA represents the verification result processed by the secondary verification unit; h 0 and s 0 respectively represent similarity of the adjacent time power grid data and the adjacent power grid data with the current verification result, wherein the similarity is obtained through cosine similarity; di a ave represents the verification result after the three-stage verification result processing.
Further, the step S2 specifically includes:
Defining a power grid data sample after pretreatment by {{r1,r2,…,rk},{o1,o2,…,ol}},{r1,r2,…,rk} for any sample, wherein k represents the number of elements in a sequence of pretreated power grid data samples, { o 1,o2,…,ol } represents a sequence of preliminary predicted load demand data samples, and l represents the number of elements in a sequence of preliminary predicted load demand data samples, and the specific process is as follows:
R= { r 1,r2,…,rk } is input to the load prediction neural network, and the initial state is set to u 0, specifically as follows:
Wherein ω 0 represents an initial weight value; b 0 denotes an initial bias; and introducing a gating unit in the load prediction neural network, wherein the gating unit comprises an input gate, an update gate, a forget gate and an output gate.
Further, the step S3 specifically includes:
The electric energy quality analysis model is constructed through deep learning to analyze the data, and an electric energy quality state value in the power grid distribution network is obtained, wherein the specific process is as follows:
wherein model PQA represents the constructed power quality analysis model; The preprocessed power grid data set comprises various data including voltage parameters, current parameters, frequency parameters, harmonic parameters, power factors, electric energy fluctuation parameters and electric energy loss parameters; q represents a parameter set of the statistically required electrical energy; re represents the electric energy loss data set recorded in the monitoring process; g represents the obtained electricity consumption data set; y out represents the output result of the power quality analysis model.
Further, the step S4 specifically includes:
Presetting the working efficiency Ework of the power grid and the energy utilization ratio Epower, wherein the initial working efficiency preset in the power grid distribution network is Ew 0, and the initial energy utilization ratio preset in the power grid distribution network is Ep 0; meanwhile, the power grid working efficiency is graded, wherein the grading comprises a high-efficiency grade, a good grade and a low-difference grade; the initial working efficiency is Ew 1.
Further still include:
The obtained load prediction information and the electric energy quality analysis result are respectively represented by Fuy and Dif, and then the work efficiency adjusting parameter Tcs of the power grid distribution network is obtained through calculation; then, correcting the energy utilization ratio Epower according to the load prediction result Fuy; and comparing the working efficiency adjusting parameter Tcs of the power grid distribution network with a preset adjusting parameter, and then combining the comparison result to obtain the final working efficiency Ework of the power grid distribution network.
A smart grid distribution management system comprising:
The system comprises a data acquisition module, a preprocessing module, a load prediction module, an electric energy quality analysis module, an overall management module and a risk early warning module;
the data acquisition module is used for monitoring and acquiring key parameter data in the power grid to obtain initial power grid data information;
The preprocessing module is used for preprocessing the initial power grid data information acquired by the data acquisition module, laying the data for accuracy, and transmitting the data to the load prediction module and the power quality analysis module; the preprocessing module comprises a primary check unit, a secondary check unit and a tertiary check unit;
the primary verification unit is used for deducting according to the acquired initial power grid data information and the information of the working windows of the verification units;
the secondary verification unit is used for performing smoothing processing to reflect the data change trend;
The third-level verification unit performs third-level verification according to the power grid data information obtained after the first-level verification unit and the second-level verification unit are processed;
The load prediction module predicts the load demand according to the preprocessed power grid data information, extracts useful information and transmits the predicted data information to the overall management module;
The power quality analysis module is used for analyzing the power quality according to the preprocessed power grid data information to obtain a power quality state result, and transmitting the analyzed data to the overall management module;
The overall management module is used for managing the power grid distribution network through the obtained power quality analysis data and the load prediction information, and optimizing power dispatching;
The risk early warning module is used for monitoring the running state in the power grid distribution network, finding out abnormal conditions in time and carrying out early warning so as to avoid potential faults and accidents.
The beneficial effects are that:
1. According to the invention, the primary verification unit is established to conduct deduction training, and the original data can be analyzed, so that the accuracy of the data is improved, and the influence of errors and abnormal data on a final result is reduced; the secondary verification unit can further smooth the data, short-term fluctuation and noise are eliminated, and the final data is more reliable and consistent; the three-level verification unit performs weighted average, so that information of a plurality of data sources can be integrated to obtain comprehensive and comprehensive power grid data information. Based on the establishment of the multistage verification unit and the adoption of deduction, smoothing and weighted average methods, the preprocessed power grid data information is obtained, so that fluctuation and mutation of data can be reduced, the final data information is more stable and reliable, the stability and reliability of a power grid system are improved, and the occurrence of abnormal conditions and the influence on the system are reduced.
2. According to the invention, the load prediction is carried out by establishing the load prediction neural network model, the data is divided into the training set and the testing set, the generalization capability of the model, namely the performance on unknown data, can be verified, and the model is further optimized; the similarity between the prediction result and the input preliminary prediction load demand data sample is calculated by utilizing the Euclidean distance, so that the judgment basis of whether the model training is successful is further increased, and the accuracy and the reliability of the model training are improved. According to the prediction result, the load scheduling is optimized, and the coordination between the power supply and the charges can be reasonably arranged, so that the power supply cost is reduced, the energy waste is reduced, and the distribution reliability of the power grid is improved.
3. The invention can identify and locate the electric energy quality problem by establishing the electric energy quality analysis model, can discover and solve the potential electric energy quality problem in time, and ensures the stability and reliability of electric power supply; by analyzing the electric energy quality data, the energy consumption mode and the load characteristic in the power grid can be known, the energy scheduling and the resource distribution of the power grid can be optimized, the energy loss can be reduced, the power transmission efficiency can be optimized, the running cost of the power grid can be reduced, and the overall energy efficiency can be improved.
4. The invention can realize the remote monitoring and management of the power equipment and the distribution network, reduce the workload of manual inspection and maintenance, automatically identify and report the fault and abnormal condition of the power equipment, early warn operation and maintenance personnel in advance, and optimize the distribution scheme to reduce the energy loss. The electric energy quality analysis result can indicate the problems in the power grid, and the power supply quality can be improved and the influence of the electric energy quality problem can be reduced by adjusting the working efficiency parameters of the distribution network system; the load prediction result can provide prediction of future load demands, so that the distribution network work can be planned and adjusted in advance to meet the power demands of all areas, and reasonable distribution and efficient utilization of power resources are realized.
Drawings
Fig. 1 is a flowchart of a smart grid distribution network management method of the present invention;
fig. 2 is a block diagram of a smart grid distribution network management system according to the present invention;
Fig. 3 is a block diagram of a preprocessing module of the present invention.
Detailed Description
In order to better understand the above technical solutions, the following detailed description will refer to the accompanying drawings and specific embodiments. It should also be understood that the particular embodiments described herein are illustrative only and are not intended to limit the invention.
Referring to fig. 1, the embodiment provides a method for managing a smart grid distribution network, including the following steps:
S1, monitoring power grid information through an ammeter, a sensor, monitoring equipment and the like, and collecting key parameters of the power grid in real time, wherein the key parameters comprise information such as current, voltage, power, load, frequency and harmonic waves, so that corresponding power grid data are obtained.
These data may be massive, highly dimensional, and may contain noise and outliers. The data needs to be preprocessed to ensure accuracy and reliability of the data.
According to the collected power grid data information, a power grid data set Di is formed, and a historical power grid data matrix is establishedB represents the record number of the historical power grid data, c represents the number in the record number of each historical power grid data; and then checking the power grid data information. The specific process is as follows: di= { Di 1,Di2,...,DiA }, a represents a check units, di a represents initial grid data information of an a-th check unit, where a e (1, 2., a); based on the initial power grid data information and the working window of each verification unit, obtaining verified power grid data information:
Firstly, a primary check unit is established, deduction is carried out according to the acquired initial power grid data information and the information of each check unit working window, and the specific process is as follows:
Wherein Di a (1) represents the verification result in the primary verification unit Representing characteristic items in the working window, including power, harmonic wave, voltage, current, phase, etc.; d 0,d1,...,dn represents the resulting parameters that need to be trained through the historical dataset.
Then, a secondary check unit is established to carry out smoothing treatment, so that the deduction error is minimized; by using EMA (exponentially weighted moving average) calculation, by giving higher weight to the latest data, the earlier data weight gradually decreases to reflect the trend of the data change, and the specific process is:
Dia EMA＝α*Dia,t (1)+(1-α)*Dia,t-1 (1)
wherein Di a EMA represents the verification result processed by the secondary verification unit; alpha represents a smoothing coefficient, alpha e [0,1]; di a,t (1) represents the verification result at the time point t; di a,t-1 (1) represents the verification result at time t-1.
In general, a larger α represents a higher weight of the most recent grid data information and is more sensitive to short term fluctuations. In the initial state, the exponentially weighted moving average of the initial time is equal to the first verification result.
Then, a third-level check unit is established, the processed power grid data information is obtained according to the first-level check unit and the second-level check unit, the power grid data adjacent to the current check result h are taken out, a data matrix Z adjacent to the moment is obtained,The power grid data which are similar to the current check unit s are taken out to obtain a similar data matrix V,Then obtaining the verification result processed by the three-level verification unit through weighted average calculation Wherein w h represents the weighting coefficients for the grid data at the moment of approach; w s represents the weighting coefficients for the similar grid data; h 0 and s 0 respectively represent similarity of the adjacent time power grid data and the adjacent power grid data with the current verification result, wherein the similarity is obtained through cosine similarity; di a ave represents the verification result after the three-stage verification result processing. Finally, a preprocessed power grid information data set/> isobtained
According to the invention, the primary verification unit is established to conduct deduction training, and the original data can be analyzed, so that the accuracy of the data is improved, and the influence of errors and abnormal data on a final result is reduced; the secondary verification unit can further smooth the data, short-term fluctuation and noise are eliminated, and the final data is more reliable and consistent; the three-level verification unit performs weighted average, so that information of a plurality of data sources can be integrated to obtain comprehensive and comprehensive power grid data information.
The invention is based on establishing the multi-stage verification unit and adopting the methods of deduction, smoothing and weighted average to obtain the preprocessed power grid data information, so that the fluctuation and mutation of the data can be reduced, the final data information is more stable and reliable, the stability and reliability of a power grid system are improved, and the occurrence of abnormal conditions and the influence on the system are reduced.
S2, building a load prediction neural network model according to the acquired preprocessed power grid data information and recorded related factors, such as weather, seasons and the like, and accurately predicting future load demands.
In one embodiment, historical data and historical load data over a period of time are collected, including load demand and related factors, such as: weather data, seasonal factors, holidays, etc.
According to the obtained preprocessed power grid data information, characteristic data extracted from other historical data by utilizing the prior art are included, wherein the time characteristics include hours, dates and days of the week, and the weather characteristics include temperature, humidity and wind speed. Setting a sample selection threshold and a threshold interval (set according to expert experience), and taking data which are higher than the threshold or meet the threshold interval from the extracted characteristic data and the power grid data information as sample data.
And building a load prediction neural network model based on the cyclic neural network, and dividing sample data into a training set and a testing set. Firstly, carrying out data fitting on a load prediction neural network model through a training set, and then carrying out momentum optimization on training errors in the training process.
The training set comprises a preprocessed power grid data set and a corresponding preliminary predicted load demand set which are required to be input into the load prediction neural network model, so as to find out the optimal load prediction information. The pre-processed power grid data sample is represented by {{r1,r2,…,rk},{o1,o2,…,ol}},{r1,r2,…,rk} for any one sample, k represents the number of elements in a sequence of pre-processed power grid data samples, { o 1,o2,…,ol } represents a sequence of preliminary predicted load demand data samples, and l represents the number of elements in a sequence of preliminary predicted load demand data samples, as follows:
R= { r 1,r2,…,rk } is input to the load prediction neural network, and the initial state is set to u 0, specifically as follows:
Wherein ω 0 represents an initial weight value; b 0 denotes an initial bias; and introducing a gating unit in the load prediction neural network, wherein the gating unit comprises an input gate, an update gate, a forget gate and an output gate. The input gate determines how much of the input information for the current time step is stored in the cell state, and the specific calculation process is as follows:
it (1)＝σ(ωir*rt+ωiu*ut-1+bi)
Wherein i t (1) denotes an input portal neuron auxiliary output; r t denotes the state of the current time step of the input value; u t-1 represents the state at the previous time; b i denotes the bias of the input portal neurons; omega ir and omega iu respectively represent corresponding weight values; i t (2) denotes the last output of the input portal neuron; epsilon represents a threshold; u t denotes the current time state. The procedure for updating portal neurons is as follows:
Where v t denotes the update of the output of the portal neuron at time step t; omega v denotes the weight value of the update portal neuron; Representing the regulation parameters; tanh represents an activation function; i represents a constant coefficient; b v denotes the bias of the update portal neuron. The specific calculation process of the amnestic portal neurons is as follows:
Wherein f t denotes the output of the forgetting portal neuron at time step t; omega f,t-1 represents the weight value corresponding to the previous time step; sigmoid represents an activation function; b f denotes the bias of the amnestic portal neurons. And finally, outputting the following output signals in an output gate:
ot＝ft*σ(ωout)+bo
Wherein o t represents an output result; omega o represents an output weight value; b o denotes the bias of the output neuron.
Carrying out similarity calculation on an output result of the load prediction neural network model and an input preliminary prediction load demand data sample by utilizing Euclidean distance; wherein the smaller the Euclidean distance is, the larger the similarity is, which indicates that the result is more accurate; if the Euclidean distance exceeds the preset range, training is needed to be continued, parameters in the load prediction neural network model are continuously optimized, iteration is continuously circulated and compared, and training is completed until the result accords with the preset value range.
According to the invention, the load prediction is carried out by establishing the load prediction neural network model, the data is divided into the training set and the testing set, the generalization capability of the model, namely the performance on unknown data, can be verified, and the model is further optimized; the similarity between the prediction result and the input preliminary prediction load demand data sample is calculated by utilizing the Euclidean distance, so that the judgment basis of whether the model training is successful is further increased, and the accuracy and the reliability of the model training are improved. According to the prediction result, the load scheduling is optimized, and the coordination between the power supply and the charges can be reasonably arranged, so that the power supply cost is reduced, the energy waste is reduced, and the distribution reliability of the power grid is improved.
S3, through monitoring and acquiring relevant parameters of the power quality in the power grid, a power quality analysis model is built, and the power quality problem is identified.
According to the continuous monitoring of the power grid working process, the related parameters of the power quality are obtained, the accuracy of the data is ensured through preprocessing, and the processed power grid information data setThe electric energy quality analysis model is constructed through deep learning to analyze the data, and an electric energy quality state value in the power grid distribution network is obtained, wherein the specific process is as follows:
wherein model PQA represents the constructed power quality analysis model; The preprocessed power grid data set comprises various data including voltage parameters, current parameters, frequency parameters, harmonic parameters, power factors, electric energy fluctuation parameters and electric energy loss parameters; q represents a parameter set of the statistically required electrical energy; re represents the electric energy loss data set recorded in the monitoring process; g represents the obtained electricity consumption data set; y out represents the output result of the power quality analysis model.
The specific calculation process is as follows:
Wherein ω (1)、ω(2)、ω(3)、ω(4)、ω(5)、ω(6)、ω(7) represents the corresponding weight values of the voltage parameter, the current parameter, the frequency parameter, the harmonic parameter, the power factor, the power fluctuation parameter, and the power loss parameter, and ω (1)+ω(2)+ω(3)+ω(4)+ω(5)+ω(6)+ω(7)＝1;γ1 represents the learning factor; t zj represents a termination statistical period when the power quality information is acquired; t zk represents a future statistical period of acquiring power quality information; t z0 represents a start statistical period when power quality information is acquired; The influence value of the power consumption after the average value is taken on the power quality analysis model is represented; inf (Q) represents the influence value of the required electric energy on the electric energy quality analysis model; inf (Re) represents the influence value of the electric energy loss on the electric energy quality analysis model; /(I) Representing the inverse error factor.
The invention can identify and locate the electric energy quality problem by establishing the electric energy quality analysis model, can discover and solve the potential electric energy quality problem in time, and ensures the stability and reliability of electric power supply; by analyzing the electric energy quality data, the energy consumption mode and the load characteristic in the power grid can be known, the energy scheduling and the resource distribution of the power grid can be optimized, the energy loss can be reduced, the power transmission efficiency can be optimized, the running cost of the power grid can be reduced, and the overall energy efficiency can be improved.
S4, monitoring the power equipment in the power grid in real time through a smart grid management method, managing according to the power quality analysis result and the load prediction information, and optimizing power dispatching.
Presetting the working efficiency Ework of the power grid and the energy utilization ratio Epower according to the acquired related data of the power grid distribution network, wherein the initial working efficiency preset in the power grid distribution network is Ew 0, and the initial energy utilization ratio preset in the power grid distribution network is Ep 0; meanwhile, the power grid working efficiency is graded, wherein the grading comprises a high-efficiency grade, a good grade and a low-difference grade;
in the high-efficiency level, the power grid has high reliability and stability, and the corresponding adjustment coefficient is mu 1, so that the initial working efficiency is Ew 1＝Ew0*μ1;
In a good class, the grid performs well in most respects but there may be some room for improvement, and the corresponding adjustment coefficient is μ 2, then the initial operating efficiency is Ew 1＝Ew0*μ2;
In low-level differences, the power grid has low efficiency, needs to be improved and upgraded, and the corresponding adjustment coefficient is mu 3, so that the initial working efficiency is Ew 1＝Ew0*μ3.
The load prediction result and the electric energy quality analysis result are obtained by combining the steps, the load prediction result and the electric energy quality analysis result are respectively represented by Fuy and Dif, and then the work efficiency adjusting parameter Tcs of the power grid distribution network is obtained through calculation, and the specific calculation process is as follows:
Wherein Fuy' represents the actual load result; delta Fuy represents a weight coefficient corresponding to the load prediction result; delta Dif represents a weight coefficient corresponding to the power quality analysis result; ρ c represents an adjustment constant.
The energy utilization rate Epower is corrected according to the load prediction result Fuy, and the specific process is as follows:
The energy supply amount is now defined as Pe; the charge amount in the energy storage system is Ein, and the discharge amount is Eout; wherein time represents time; determining the obtained energy supply amount according to the energy sources such as fuel gas, natural gas, hydraulic power, wind energy, solar energy and the like;
wherein, Representing a load prediction error rate; time represents a defined time. When the load prediction error rate is low, namely the difference between the predicted load and the actual load is small, the power grid distribution network management system can supply and distribute according to accurate load demands, so that the power grid distribution network management system has higher energy utilization rate; on the contrary, when the load prediction error rate is higher, that is, a larger difference exists between the predicted load and the actual load, the distribution work of the power grid distribution network management system cannot be timely and accurately performed, so that the energy utilization rate is reduced.
And comparing the working efficiency adjusting parameter Tcs of the power grid distribution network with a preset adjusting parameter, and then combining the comparison result to obtain the final working efficiency Ework of the power grid distribution network, wherein the specific process is as follows:
defining a preset adjustment parameter as Tcs 1, wherein the preset adjustment parameter is obtained according to historical experimental data, and Tcs 1 is more than 0; wherein ρ 1 represents a regulatory factor;
and finally, the power grid distribution network management system performs management work according to the corrected energy utilization rate Wpower and the power grid distribution network work efficiency Ewprk.
The invention can realize the remote monitoring and management of the power equipment and the distribution network, reduce the workload of manual inspection and maintenance, automatically identify and report the fault and abnormal condition of the power equipment, early warn operation and maintenance personnel in advance, and optimize the distribution scheme to reduce the energy loss. The electric energy quality analysis result can indicate the problems in the power grid, and the power supply quality can be improved and the influence of the electric energy quality problem can be reduced by adjusting the working efficiency parameters of the distribution network system; the load prediction result can provide prediction of future load demands, so that the distribution network work can be planned and adjusted in advance to meet the power demands of all areas, and reasonable distribution and efficient utilization of power resources are realized.
The invention is beneficial to reducing energy waste, reducing energy cost, promoting sustainable energy use and improving the stability and reliability of the power grid.
Referring to fig. 2 and 3, the present embodiment provides a smart grid distribution network management system, which includes the following contents:
The system comprises a data acquisition module, a preprocessing module, a load prediction module, an electric energy quality analysis module, an overall management module and a risk early warning module;
The data acquisition module is used for monitoring and acquiring key parameter data in the power grid to obtain initial power grid data information;
The preprocessing module is used for preprocessing the initial power grid data information acquired by the data acquisition module, laying the data for accuracy, and transmitting the data to the load prediction module and the power quality analysis module; the preprocessing module comprises a primary check unit, a secondary check unit and a tertiary check unit;
The primary verification unit is used for deducting according to the acquired initial power grid data information and the information of the working windows of the verification units;
the second-level verification unit is used for performing smoothing processing to reflect the data change trend;
The third-level verification unit is used for performing third-level verification according to the power grid data information obtained after the first-level verification unit and the second-level verification unit are processed;
The load prediction module predicts the load demand according to the preprocessed power grid data information, extracts useful information and transmits the predicted data information to the overall management module;
The power quality analysis module is used for analyzing the power quality according to the preprocessed power grid data information to obtain a power quality state result, and transmitting the analyzed data to the overall management module;
The overall management module is used for managing the power grid distribution network through the obtained power quality analysis data and the load prediction information, and optimizing power dispatching;
And the risk early warning module is used for monitoring the running state in the power grid distribution network, finding out abnormal conditions in time and carrying out early warning so as to avoid potential faults and accidents.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
The above is only for illustrating the technical idea of the present invention, and the protection scope of the present invention is not limited by this, and any modification made on the basis of the technical scheme according to the technical idea of the present invention falls within the protection scope of the claims of the present invention.
Claims (4)
1. The intelligent power grid distribution network management method is characterized by comprising the following steps of:
s1, monitoring power grid information, collecting key parameter data in a power grid, and preprocessing the obtained power grid data information;
when data preprocessing is carried out, a grid data set Di is formed according to the collected grid data information, and a historical grid data matrix is established B represents the record number of the historical power grid data, c represents the number in the record number of each historical power grid data; then checking the power grid data information; the device comprises a first-stage check unit, a second-stage check unit and a third-stage check unit;
The method comprises the steps of establishing a primary verification unit, and deducing according to the acquired initial power grid data information and the information of the working window of each verification unit, wherein the specific process is as follows:
Wherein Di a (1) represents the verification result in the primary verification unit; representing feature items in a work window; d 0,d1,...,dn denotes the resulting parameters that need to be trained by the historical dataset;
Establishing a secondary verification unit for smoothing treatment, and calculating by using EMA, wherein the specific process is as follows:
Dia EMA＝α*Dia,t (1)+(1-α)*Dia,t-1 (1)
Wherein Di a EMA represents the verification result processed by the secondary verification unit; alpha represents a smoothing coefficient, alpha e [0,1]; di a,t (1) represents the verification result at the time point t; di a,t-1 (1) represents the verification result at time t-1;
in the third-level check unit, the processed power grid data information is obtained according to the first-level check unit and the second-level check unit, the power grid data at the moment h adjacent to the current check result is taken out to obtain a data matrix Z at the moment adjacent to the moment, The power grid data which are similar to the current check unit s are taken out to obtain a similar data matrix V,Then obtaining the verification result processed by the three-level verification unit through weighted average calculation Wherein w h represents the weighting coefficients for the grid data at the moment of approach; w s represents the weighting coefficients for the similar grid data; /(I)Representing the verification result processed by the secondary verification unit; h 0 and s 0 respectively represent similarity of the adjacent time power grid data and the adjacent power grid data with the current verification result, wherein the similarity is obtained through cosine similarity; di a ave represents the verification result after the three-level verification result is processed;
S2, building a load prediction neural network model according to the acquired preprocessed power grid data information and recorded related factors, and accurately predicting future load demands;
s3, constructing an electric energy quality analysis model through monitoring and acquiring relevant parameters of the electric energy quality in the power grid, and identifying electric energy quality problems;
S4, presetting the work efficiency and the energy utilization rate of the power grid distribution network based on the load prediction information and the power quality analysis result, and obtaining the corrected energy utilization rate and the corrected work efficiency of the power grid distribution network through calculation so as to manage the power grid distribution network;
Grading is carried out through the working efficiency of the power grid, wherein the grading comprises a high-efficiency grade, a good grade and a low-difference grade, and the initial working efficiency Ew 1 is obtained; the obtained load prediction information and the obtained electric energy quality analysis result are respectively represented by Fuy and Dif, and the work efficiency adjusting parameter Tcs of the power grid distribution network is obtained through calculation, wherein the specific calculation process is as follows:
Wherein Fuy' represents the actual load result; delta Fuy represents a weight coefficient corresponding to the load prediction result; delta Dif represents a weight coefficient corresponding to the power quality analysis result; ρ c represents an adjustment constant;
Correcting the energy utilization rate Epower according to a load prediction result Fuy, and defining the energy supply quantity as Pe; the charge amount in the energy storage system is Ein, and the discharge amount is Eout; wherein time represents time; wherein/> Representing a load prediction error rate;
Comparing the working efficiency adjusting parameter Tcs of the power grid distribution network with a preset adjusting parameter, and then combining the comparison result to obtain the working efficiency Ework of the power grid distribution network, wherein the specific process is as follows:
Defining a preset adjusting parameter as Tcs 1, wherein the preset adjusting parameter is obtained according to historical experimental data, and Tcs 1 is more than 0; wherein ρ 1 represents a regulatory factor;
2. the smart grid distribution network management method according to claim 1, wherein the step S2 specifically includes:
Defining a power grid data sample after pretreatment by {{r1,r2,...,rk},{o1,o2,...,ol}},{r1,r2,...,rk} for any sample, wherein k represents the number of elements in a sequence of pretreated power grid data samples, { o 1,o2,...,ol } represents a sequence of preliminary predicted load demand data samples, and l represents the number of elements in a sequence of preliminary predicted load demand data samples, and the specific process is as follows:
r= { r 1,r2,...,rk } is input to the load prediction neural network, and the initial state is set to u 0, specifically as follows:
Wherein ω 0 represents an initial weight value; b 0 denotes an initial bias; and introducing a gating unit in the load prediction neural network, wherein the gating unit comprises an input gate, an update gate, a forget gate and an output gate.
3. The smart grid distribution network management method according to claim 1, wherein the step S3 specifically includes:
The electric energy quality analysis model is constructed through deep learning to analyze the data, and an electric energy quality state value in the power grid distribution network is obtained, wherein the specific process is as follows:
wherein model PQA represents the constructed power quality analysis model; The preprocessed power grid data set comprises various data including voltage parameters, current parameters, frequency parameters, harmonic parameters, power factors, electric energy fluctuation parameters and electric energy loss parameters; q represents a parameter set of the statistically required electrical energy; re represents the electric energy loss data set recorded in the monitoring process; g represents the obtained electricity consumption data set; y out represents the output result of the power quality analysis model.
4. A smart grid distribution network management system, applied to the smart grid distribution network management method of claim 1, comprising the following contents:
The system comprises a data acquisition module, a preprocessing module, a load prediction module, an electric energy quality analysis module, an overall management module and a risk early warning module;
the data acquisition module is used for monitoring and acquiring key parameter data in the power grid to obtain initial power grid data information;
The preprocessing module is used for preprocessing the initial power grid data information acquired by the data acquisition module, laying the data for accuracy, and transmitting the data to the load prediction module and the power quality analysis module; the preprocessing module comprises a primary check unit, a secondary check unit and a tertiary check unit;
the primary verification unit is used for deducting according to the acquired initial power grid data information and the information of the working windows of the verification units;
the secondary verification unit is used for performing smoothing processing to reflect the data change trend;
The third-level verification unit performs third-level verification according to the power grid data information obtained after the first-level verification unit and the second-level verification unit are processed;
The load prediction module predicts the load demand according to the preprocessed power grid data information, extracts useful information and transmits the predicted data information to the overall management module;
The power quality analysis module is used for analyzing the power quality according to the preprocessed power grid data information to obtain a power quality state result, and transmitting the analyzed data to the overall management module;
The overall management module is used for managing the power grid distribution network through the obtained power quality analysis data and the load prediction information, and optimizing power dispatching;
The risk early warning module is used for monitoring the running state in the power grid distribution network, finding out abnormal conditions in time and carrying out early warning so as to avoid potential faults and accidents.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311370220.9A CN117595231B (en) | 2023-10-20 | Intelligent power grid distribution management system and method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311370220.9A CN117595231B (en) | 2023-10-20 | Intelligent power grid distribution management system and method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117595231A CN117595231A (en) | 2024-02-23 |
| CN117595231B true CN117595231B (en) | 2024-05-31 |
Family
ID=
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012171147A1 (en) * | 2011-06-17 | 2012-12-20 | 辽宁省电力有限公司 | Coordination and control system for regulated charging and discharging of pure electric vehicle in combination with wind power generation |
| CN110738344A (en) * | 2018-07-20 | 2020-01-31 | 中国农业大学 | Distributed reactive power optimization method and device for load prediction of power system |
| CN110929927A (en) * | 2019-11-18 | 2020-03-27 | 国网甘肃省电力公司 | Artificial intelligence prediction model construction method applying power distribution network universe big data |
| CN112733417A (en) * | 2020-11-16 | 2021-04-30 | 南京邮电大学 | Abnormal load data detection and correction method and system based on model optimization |
| WO2022036778A1 (en) * | 2020-08-21 | 2022-02-24 | 山东大学 | Optimal control method and system for load recovery of transmission and distribution coordination |
| CN115392340A (en) * | 2022-07-11 | 2022-11-25 | 江苏科技大学 | Power load prediction system and prediction method for multi-energy electric propulsion ship |
| CN116187516A (en) * | 2022-12-13 | 2023-05-30 | 郑振富 | Big data-based analysis and correction method |
| CN116388150A (en) * | 2023-02-15 | 2023-07-04 | 国网辽宁省电力有限公司电力科学研究院 | Matching short-term load prediction method based on digital twin and neural network technology |
| WO2023142424A1 (en) * | 2022-01-25 | 2023-08-03 | 国网江苏省电力有限公司南京供电分公司 | Power financial service risk control method and system based on gru-lstm neural network |
| CN116581750A (en) * | 2023-05-24 | 2023-08-11 | 江苏大学扬州(江都)新能源汽车产业研究所 | Intelligent line load charging method based on power grid load level |
| CN116646933A (en) * | 2023-07-24 | 2023-08-25 | 北京中能亿信软件有限公司 | Big data-based power load scheduling method and system |
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012171147A1 (en) * | 2011-06-17 | 2012-12-20 | 辽宁省电力有限公司 | Coordination and control system for regulated charging and discharging of pure electric vehicle in combination with wind power generation |
| CN110738344A (en) * | 2018-07-20 | 2020-01-31 | 中国农业大学 | Distributed reactive power optimization method and device for load prediction of power system |
| CN110929927A (en) * | 2019-11-18 | 2020-03-27 | 国网甘肃省电力公司 | Artificial intelligence prediction model construction method applying power distribution network universe big data |
| WO2022036778A1 (en) * | 2020-08-21 | 2022-02-24 | 山东大学 | Optimal control method and system for load recovery of transmission and distribution coordination |
| CN112733417A (en) * | 2020-11-16 | 2021-04-30 | 南京邮电大学 | Abnormal load data detection and correction method and system based on model optimization |
| WO2023142424A1 (en) * | 2022-01-25 | 2023-08-03 | 国网江苏省电力有限公司南京供电分公司 | Power financial service risk control method and system based on gru-lstm neural network |
| CN115392340A (en) * | 2022-07-11 | 2022-11-25 | 江苏科技大学 | Power load prediction system and prediction method for multi-energy electric propulsion ship |
| CN116187516A (en) * | 2022-12-13 | 2023-05-30 | 郑振富 | Big data-based analysis and correction method |
| CN116388150A (en) * | 2023-02-15 | 2023-07-04 | 国网辽宁省电力有限公司电力科学研究院 | Matching short-term load prediction method based on digital twin and neural network technology |
| CN116581750A (en) * | 2023-05-24 | 2023-08-11 | 江苏大学扬州(江都)新能源汽车产业研究所 | Intelligent line load charging method based on power grid load level |
| CN116646933A (en) * | 2023-07-24 | 2023-08-25 | 北京中能亿信软件有限公司 | Big data-based power load scheduling method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11876374B2 (en) | System and method for optimal control of energy storage system | |
| CN113962364B (en) | Multi-factor power load prediction method based on deep learning | |
| CN105069525B (en) | Round-the-clock 96 Day Load Curve Forecastings and optimization update the system | |
| CN116646933A (en) | Big data-based power load scheduling method and system | |
| CN106408223A (en) | Short-term load prediction based on meteorological similar day and error correction | |
| CN107944594B (en) | Short-term load prediction method based on spearman grade and RKELM microgrid | |
| Kolhe et al. | GA-ANN for short-term wind energy prediction | |
| CN112329990A (en) | User power load prediction method based on LSTM-BP neural network | |
| CN111915092A (en) | Ultra-short-term wind power prediction method based on long-time and short-time memory neural network | |
| CN115965132A (en) | Distributed photovoltaic digital twin system power prediction method based on GA-BP neural network | |
| CN115425680A (en) | Power prediction model construction and prediction method of multi-energy combined power generation system | |
| CN114648176A (en) | Wind-solar power consumption optimization method based on data driving | |
| CN115481918A (en) | Active sensing and predictive analysis system for unit state based on source network load storage | |
| CN114021848A (en) | Generating capacity demand prediction method based on LSTM deep learning | |
| CN113537582A (en) | Photovoltaic power ultra-short-term prediction method based on short-wave radiation correction | |
| CN109858668B (en) | Coordination prediction method for power load region in thunder and lightning climate | |
| CN115238854A (en) | Short-term load prediction method based on TCN-LSTM-AM | |
| CN112734141B (en) | Diversified load interval prediction method and device | |
| Buonanno et al. | Comprehensive method for modeling uncertainties of solar irradiance for PV power generation in smart grids | |
| CN117595231B (en) | Intelligent power grid distribution management system and method thereof | |
| CN117113086A (en) | Energy storage unit load prediction method, system, electronic equipment and medium | |
| CN116128211A (en) | Wind-light-water combined short-term optimization scheduling method based on wind-light uncertainty prediction scene | |
| CN117595231A (en) | Intelligent power grid distribution management system and method thereof | |
| CN114759579A (en) | Power grid active power optimization control system, method and medium based on data driving | |
| CN113112085A (en) | New energy station power generation load prediction method based on BP neural network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |