CN110796307B - Distributed load prediction method and system for comprehensive energy system - Google Patents
Distributed load prediction method and system for comprehensive energy system Download PDFInfo
- Publication number
- CN110796307B CN110796307B CN201911035294.0A CN201911035294A CN110796307B CN 110796307 B CN110796307 B CN 110796307B CN 201911035294 A CN201911035294 A CN 201911035294A CN 110796307 B CN110796307 B CN 110796307B
- Authority
- CN
- China
- Prior art keywords
- load
- daily
- data
- curve
- time sequence
- 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
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000005265 energy consumption Methods 0.000 claims description 32
- 230000005611 electricity Effects 0.000 claims description 21
- 238000004364 calculation method Methods 0.000 claims description 10
- 238000013277 forecasting method Methods 0.000 claims description 3
- 239000007789 gas Substances 0.000 description 28
- 230000006870 function Effects 0.000 description 6
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 description 6
- 230000008878 coupling Effects 0.000 description 5
- 238000010168 coupling process Methods 0.000 description 5
- 238000005859 coupling reaction Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 238000012937 correction Methods 0.000 description 4
- 238000010606 normalization Methods 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 239000003345 natural gas Substances 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- YCKRFDGAMUMZLT-UHFFFAOYSA-N Fluorine atom Chemical compound [F] YCKRFDGAMUMZLT-UHFFFAOYSA-N 0.000 description 1
- 238000009412 basement excavation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004146 energy storage Methods 0.000 description 1
- 239000003344 environmental pollutant Substances 0.000 description 1
- 238000003912 environmental pollution Methods 0.000 description 1
- 229910052731 fluorine Inorganic materials 0.000 description 1
- 239000011737 fluorine Substances 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 231100000719 pollutant Toxicity 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000002195 synergetic effect Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- 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—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Primary Health Care (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a distributed load prediction method and a distributed load prediction system for a comprehensive energy system. The method comprises the following steps: acquiring an electric load, a thermal load and an air load of the comprehensive energy system, an electric load time sequence curve, a thermal load time sequence curve and an air load time sequence curve, and acquiring external factor data; calculating a load characteristic index according to the electric load, the thermal load and the air load; carrying out load clustering according to the load characteristic index, the electrical load time sequence curve, the thermal load time sequence curve and the air load time sequence curve; establishing an offline load prediction model for each type of load according to external factor data; performing online load prediction by adopting an offline load prediction model according to the current daily load data and the external factor data of the day to be predicted to obtain the daily load to be predicted; and summing the daily loads to be predicted of each type of load to obtain the total load of the daily comprehensive energy system to be predicted. By adopting the method and the system, the load prediction precision of the comprehensive energy system can be improved.
Description
Technical Field
The invention relates to the technical field of load prediction, in particular to a distributed load prediction method and system of a comprehensive energy system.
Background
In the face of the challenges of resource shortage, environmental pollution and the like, the comprehensive energy system becomes an important energy utilization mode in the energy transformation process, and has obvious effects of improving the energy utilization efficiency and reducing pollutant emission. Compared with the traditional power system, thermodynamic system and natural gas system which are independently planned, designed and operated, the comprehensive energy system integrates various forms of energy supply, energy conversion and energy storage equipment, realizes the coupling of different types of energy sources in different links such as source, network, load and the like, and ensures that the connection among the power system, the thermodynamic system and the natural gas system is tighter, and the comprehensive energy load prediction becomes an important part in the economic dispatching and optimized operation of the comprehensive energy system.
Most of the existing load prediction methods are respectively predicting electric, thermal and gas loads, and the mutual influence among different types of loads cannot be considered in a synergistic manner. Meanwhile, along with the popularization of terminal equipment such as an intelligent electric meter and the like, on one hand, the electricity utilization data of a user is continuously updated along with time, rapidly increases and is in a massive situation; on the other hand, the data acquisition points are distributed on the user side, and the data acquisition points have extremely strong dispersibility. The existing prediction method is difficult to process large-scale distributed data. Therefore, intensive research needs to be carried out on a distributed load prediction method of the comprehensive energy system based on data driving, massive distributed data information is fully utilized, and the electricity-heat-gas load coupling characteristics are analyzed to realize high-precision prediction of the load of the comprehensive energy system.
Disclosure of Invention
The invention aims to provide a distributed load forecasting method and a distributed load forecasting system for a comprehensive energy system, which consider the mutual influence among different types of loads, analyze the electricity-heat-gas load coupling characteristics by utilizing distributed data information and improve the load forecasting precision of the comprehensive energy system.
In order to achieve the purpose, the invention provides the following scheme:
a distributed load prediction method for an integrated energy system comprises the following steps:
acquiring an electric load and an air load of the comprehensive energy system, acquiring an electric load time sequence curve, a heat load time sequence curve and an air load time sequence curve, and acquiring external factor data; the external factor data comprises electricity price, gas price and meteorological condition data;
calculating a load characteristic index from the electrical load, the thermal load, and the air load;
carrying out load clustering according to the load characteristic index, the electric load time sequence curve, the heat load time sequence curve and the gas load time sequence curve to obtain various loads;
establishing an offline load prediction model for each type of load according to the external factor data;
acquiring load data of the current day and external factor data of a day to be predicted aiming at each type of load, and performing online load prediction by adopting the offline load prediction model according to the load data of the current day and the external factor data of the day to be predicted to obtain a day load to be predicted;
and summing the daily loads to be predicted of each type of load to obtain the total load of the daily comprehensive energy system to be predicted.
Optionally, before calculating the load characteristic index according to the electrical load, the thermal load and the air load, the method further includes:
correcting deviation data of the electrical load, the thermal load and the gas load by adopting a data curve fitting method;
and performing normalization processing on the corrected electric load, the corrected heat load and the corrected gas load to obtain the normalized electric load, the normalized heat load and the normalized gas load.
Optionally, the calculating a load characteristic index according to the electrical load, the thermal load, and the air load specifically includes:
the daily average load is calculated according to the following formula:
the daily load rate is calculated according to the following formula:
the peak energy consumption rate was calculated according to the following formula:
calculating the valley time energy consumption rate according to the following formula:
in the formula, KavDenotes the average daily load, EallExpressing daily energy, wherein the daily energy is one of daily electricity, daily heat and daily gas, T expresses the number of daily time segments, T is 24, KdRepresenting the daily load rate, PavDenotes the daily average load, PmaxDenotes the daily maximum load, ρpDenotes the peak hour specific energy consumption, EpRepresenting energy used during peak hours, pvIndicating the energy consumption rate at valley time, EvIndicating energy usage during the daily trough period.
Optionally, the performing load clustering according to the load characteristic index, the electrical load timing sequence curve, the thermal load timing sequence curve, and the air load timing sequence curve to obtain multiple types of loads specifically includes:
load clustering is performed according to the following formula:
wherein F represents a clustering objective function, K represents the number of clusters, M represents the number of load time sequence curves, and hk,mRepresenting an identification variable, h when the load curve m belongs to the load class kk,mThe value is 1, when the load curve m does not belong to the load class k, hk,mThe value is 0, alpha represents a weight coefficient,
representing the load characteristic distance between the load curve m and the load class k,
representing the load time-series curve distance, λ, between the load curve m and the load class k1Denotes a first coefficient, λ2Denotes a second coefficient, λ3Denotes the third coefficient, λ4Represents a fourth coefficient and satisfies λ1+λ2+λ3+λ4＝1；Kav,mThe average daily load factor, K, of the load curve mav,kDenotes the daily average load factor, K, of the load class Kd,mRepresenting the daily load factor, K, of the load curve md,kDenotes the daily load rate, ρ, of the load class kp,mRepresents the peak-time energy consumption rate, ρ, of the load curve mp,kRepresenting the peak-time energy consumption rate, ρ, of the load class kv,mThe energy consumption rate at valley time, ρ, expressed as the load curve mv,kRepresenting the valley time energy consumption rate of the load curve k; p ism,tRepresenting the load value, P, of the load curve m to be clustered during the t periodk,tAnd representing the load value conforming to the class k to be clustered in the period t.
Optionally, the establishing an offline load prediction model for each type of load according to the external factor data specifically includes:
acquiring one type of data in the external factor data; the data of one type is one of electricity price, gas price and meteorological condition data;
calculating the correlation coefficient of the class data and each class of load;
comparing the correlation coefficient with a preset threshold value; if the correlation coefficient is larger than the preset threshold value, determining the class of data as a correlation factor corresponding to the load type; otherwise, determining whether other types of data in the external factor data are relevant factors corresponding to the load types;
and establishing an offline load prediction model according to the correlation factors and the load data corresponding to the correlation factors.
Optionally, before performing online load prediction, the method further includes:
acquiring a current daily prediction error of each type of load, a next-day predicted daily load of each type of load and correlation factors corresponding to the load types;
and establishing a load prediction error model according to the prediction error of the current day, the next day prediction day load and the correlation factor corresponding to the load type.
Optionally, the obtaining of load data of a current day and external factor data of a day to be predicted for each type of load, and performing online load prediction by using the offline load prediction model according to the load data of the current day and the external factor data of the day to be predicted to obtain a day load to be predicted specifically includes:
generating a comprehensive energy system load prediction model according to the offline load prediction model and the load prediction error model;
and acquiring a current day prediction error, current day load data and external factor data of a day to be predicted aiming at each type of load, and performing online load prediction by adopting the comprehensive energy system load prediction model according to the current day prediction error, the current day load data and the external factor data of the day to be predicted to obtain the day load to be predicted.
The invention also provides a distributed load prediction system of the comprehensive energy system, which comprises:
the data acquisition module is used for acquiring the electric load, the heat load and the air load of the comprehensive energy system, acquiring an electric load time sequence curve, a heat load time sequence curve and an air load time sequence curve, and acquiring external factor data; the external factor data comprises electricity price, gas price and meteorological condition data;
a load characteristic index calculation module for calculating a load characteristic index from the electrical load, the thermal load and the air load;
the load clustering module is used for carrying out load clustering according to the load characteristic index, the electric load time sequence curve, the heat load time sequence curve and the gas load time sequence curve to obtain various loads;
the offline load prediction model establishing module is used for establishing an offline load prediction model for each type of load according to the external factor data;
the online load prediction module is used for acquiring current daily load data and external factor data of a day to be predicted aiming at each type of load, and performing online load prediction by adopting the offline load prediction model according to the current daily load data and the external factor data of the day to be predicted to obtain a daily load to be predicted;
and the comprehensive energy system total load calculation module is used for summing the daily loads to be predicted of each type of load to obtain the daily comprehensive energy system total load to be predicted.
Compared with the prior art, the invention has the beneficial effects that:
the invention provides a distributed load prediction method and a distributed load prediction system for a comprehensive energy system, which consider external factor data of electricity price, gas price and meteorological condition data; calculating a load characteristic index according to the electric load, the heat load and the air load; carrying out load clustering according to the load characteristic index, the electrical load time sequence curve, the thermal load time sequence curve and the air load time sequence curve; establishing an offline load prediction model for each type of load according to external factor data, and performing online load prediction according to the prediction model to obtain a daily load to be predicted; and finally, summing the daily loads to be predicted of each type of load to obtain the total load of the daily comprehensive energy system to be predicted, considering the mutual influence among different types of loads, analyzing the electric-heat-gas load coupling characteristics by using distributed data information, and improving the load prediction precision of the comprehensive energy system.
In addition, deviation data correction is carried out on the electric load, the heat load and the gas load by adopting a data curve fitting method, and normalization processing is carried out on the corrected electric load, the corrected heat load and the corrected gas load, so that the load prediction precision of the comprehensive energy system is further improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to these drawings without creative efforts.
Fig. 1 is a flowchart of a distributed load prediction method for an integrated energy system according to an embodiment of the present invention;
fig. 2 is a structural diagram of a distributed load prediction system of an integrated energy system according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.
The invention aims to provide a distributed load prediction method and a distributed load prediction system for a comprehensive energy system, which consider the mutual influence among different types of loads, analyze the electric-heat-gas load coupling characteristics by utilizing distributed data information and improve the load prediction precision of the comprehensive energy system.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Examples
Fig. 1 is a flowchart of a distributed load prediction method of an integrated energy system according to an embodiment of the present invention, and as shown in fig. 1, the embodiment provides a distributed load prediction method of an integrated energy system, including:
step 101: acquiring an electric load, a thermal load and an air load of the comprehensive energy system, acquiring an electric load time sequence curve, a thermal load time sequence curve and an air load time sequence curve, and acquiring external factor data; the external factor data includes electricity price, natural gas price, and weather condition data.
For the fact that poor data may exist in the acquired original data of the comprehensive energy system, deviation data correction needs to be carried out on the electric load, the heat load and the air load by adopting a data curve fitting method.
The data curve fitting method specifically comprises the following steps: acquiring deviation data; selecting front and back normal data of the deviation data point according to a time sequence by taking the deviation data point as a center; performing curve fitting on the selected normal data to obtain a data fitting function; and calculating a function value at the deviation data point according to the data fitting function, and taking the function value at the deviation data point as a correction value of the deviation data.
And (3) considering dimensional differences of data such as electricity, heat and gas loads, and performing normalization processing on the corrected electricity load, the corrected heat load and the corrected gas load by adopting a min-max method to obtain the normalized electricity load, the normalized heat load and the normalized gas load.
The normalization formula is as follows:
in the formula, x,
The original value and the normalized value of the electrical load (or the thermal load and the gas load) are respectively; x is a radical of a fluorine atommax、xminRespectively, the maximum value and the minimum value in the raw data of the electric load (or the heat load and the air load).
Step 102: and calculating the load characteristic index according to the electric load, the heat load and the air load.
The daily average load is calculated according to the following formula:
the daily load rate is calculated according to the following formula:
the peak time energy consumption rate is calculated according to the following formula:
calculating the valley time energy consumption rate according to the following formula:
in the formula, KavDenotes the average daily load, EallThe daily energy is one of daily electricity, daily heat and daily gas, T represents the number of daily time periods, T is 24, KdRepresenting the daily load rate, PavDenotes the daily average load, PmaxDenotes the daily maximum load, ρpDenotes the peak hour specific energy consumption, EpRepresenting energy used during peak hours, pvIndicating the energy consumption rate at valley time, EvIndicating energy usage during the daily trough period.
Step 103: and carrying out load clustering according to the load characteristic indexes, the electric load time sequence curve, the thermal load time sequence curve and the air load time sequence curve to obtain various loads.
Load clustering is performed according to the following formula:
wherein F represents a clustering objective function, K represents the number of clusters, M represents the number of load time sequence curves, and hk,mRepresenting an identification variable, h when the load curve m belongs to the load class kk,mThe value is 1, h when the load curve m does not belong to the load class kk,mThe value is 0, alpha represents the weight coefficient,
representing the load characteristic distance between the load curve m and the load class k,
representing the load time-series curve distance, λ, between the load curve m and the load class k1Denotes a first coefficient, λ2Denotes a second coefficient, λ3Denotes the third coefficient, λ4Represents a fourth coefficient and satisfies λ1+λ2+λ3+λ4＝1；Kav,mThe average daily load factor, K, of the load curve mav,kDenotes the daily average load factor, K, of the load class Kd,mDenotes the daily load factor, K, of the load curve md,kDenotes the daily load rate, ρ, of the load class kp,mRepresents the peak-time energy consumption rate, ρ, of the load curve mp,kRepresents the peak specific energy consumption, ρ, of the load class kv,mThe energy consumption rate at valley time, ρ, expressed as the load curve mv,kThe valley time energy consumption rate of the load category k is represented; p ism,tRepresenting the load value, P, of the load curve m to be clustered during the t periodk,tAnd representing the load value of the load category k to be clustered in the period t.
Step 104: and establishing an offline load prediction model for each type of load according to the external factor data.
Acquiring one type of data in the external factor data; one type of data is one of electricity price, gas price, and meteorological condition (temperature, humidity) data; calendar information (weekdays and holidays) may also be obtained.
And calculating the correlation coefficient of one type of data and each type of load. The correlation between the electrical load and the meteorological condition is analyzed by taking the electrical load and the meteorological condition as an example, and the correlation coefficient calculation formula is as follows:
wherein X, Y are electrical load and meteorological condition input data, respectively; cov (X, Y) is the covariance between X and Y, and D (X), D (Y) are the mean square deviations of X and Y, respectively.
Comparing the correlation coefficient with a preset threshold value; if the correlation coefficient is larger than a preset threshold (for example, 0.6), determining a type of data as a correlation factor corresponding to the load type; otherwise, determining whether other types of data in the external factor data are the correlation factors corresponding to the load types or not, wherein the determination comprises the step of respectively determining whether the electricity price and the gas price are the correlation factors corresponding to the load types or not.
And establishing an offline load prediction model according to the correlation factors and the load data corresponding to the correlation factors.
For the kth load, a load classification prediction model based on a long-short term memory network (LSTM) is constructed, and the LSTM is one of deep learning algorithms. Obtaining J having correlation with the kth class load according to the calculation result of the correlation coefficientkFactor(s) J to be obtainedkThe related data of the factors and the kth class load history data are used as the input of the LSTM, and the kth class load history is used as the output of the LSTM; and training the LSTM by using a large amount of input and output data to obtain a prediction model of the kth class load. The offline load prediction model is as follows:
wherein, the first and the second end of the pipe are connected with each other,
as input data of the LSTM model, yk,d+1Outputting a kth class offline load prediction model; in the LSTM model training process, input and output data are historical data, and yk，d、yk,d+1Historical load data of the kth class load on the d-th day and the d + 1-th day respectively,
d +1 calendar history data for the kth load, jth factor, J1, 2, …, Jk。
After step 104, further comprising:
and acquiring a current day prediction error of each type of load, a next day prediction day load of each type of load and correlation factors corresponding to the load types.
And establishing a load prediction error model according to the prediction error of the current day, the next day prediction day load and the correlation factor corresponding to the load type.
The load prediction error model is built by using LSTM as follows:
the training process of the load prediction error model is similar to the construction method of the off-line load prediction model. In the training process of the model f (-), the input data and the output data of the model are all historical data. Wherein the load prediction error epsilon of the kth load on day d +1k,d+1Load prediction error ε at day d of kth class loadk,dThe predicted value y of the offline load on day d +1 of the kth class loadk,d+1D +1 st day JkThe predicted data of each factor is used as the input of the model; load prediction error epsilon on day d +1k,d+1As output of the model.
Step 105: and acquiring load data of the current day and external factor data of the day to be predicted aiming at each type of load, and performing online load prediction by adopting an offline load prediction model according to the load data of the current day and the external factor data of the day to be predicted to obtain the load of the day to be predicted.
and generating a comprehensive energy system load prediction model according to the offline load prediction model and the load prediction error model. The comprehensive energy system load prediction model comprises the following steps:
wherein, gk() And fk(. cndot.) are the kth class load offline load prediction model and the load prediction error model respectively. In making an on-line load prediction, yk,dAnd epsilonk,dLoad data and load prediction error data of the current day of the kth class load are respectively,
predicted values of day d +1 of the k-th class load are taken into consideration for prediction error correction.
And acquiring a current day prediction error, current day load data and external factor data of a day to be predicted aiming at each type of load, and performing online load prediction by adopting a comprehensive energy system load prediction model according to the current day prediction error, the current day load data and the external factor data of the day to be predicted to obtain the daily load to be predicted.
Step 106: and summing the daily loads to be predicted of each type of load to obtain the total load of the daily comprehensive energy system to be predicted.
The total load calculation formula is as follows:
wherein G is the total load of the daily comprehensive energy system to be predicted;
for the predicted load of the kth class, K represents the total number of clusters.
Fig. 2 is a structural diagram of a distributed load prediction system of an integrated energy system according to an embodiment of the present invention, and as shown in fig. 2, the embodiment provides a distributed load prediction system of an integrated energy system, including:
the data acquisition module 201 is configured to acquire an electrical load, a thermal load, and an air load of the integrated energy system, acquire an electrical load time sequence curve, a thermal load time sequence curve, and an air load time sequence curve, and acquire external factor data; the external factor data includes electricity price, gas price, and weather condition data.
And a load characteristic index calculation module 202, configured to calculate a load characteristic index according to the electrical load, the thermal load, and the air load.
The load clustering module 203 is used for clustering loads according to the load characteristic indexes, the electric load time sequence curve, the heat load time sequence curve and the air load time sequence curve to obtain various loads;
and an offline load prediction model establishing module 204, configured to establish an offline load prediction model for each type of load according to the external factor data.
And the online load prediction module 205 is configured to obtain load data of the current day and external factor data of a day to be predicted for each type of load, and perform online load prediction by using an offline load prediction model according to the load data of the current day and the external factor data of the day to be predicted to obtain a daily load to be predicted.
And the comprehensive energy system total load calculating module 206 is configured to sum the daily loads to be predicted of each type of load to obtain the daily comprehensive energy system total load to be predicted.
For the system disclosed by the embodiment, the description is relatively simple because the system corresponds to the method disclosed by the embodiment, and the relevant points can be referred to the method part for description.
By adopting the distributed load prediction method and system of the comprehensive energy system, the load prediction method and system can be applied to the load prediction of the comprehensive energy system in an actual region. The basic load data based on the load forecasting method comprises historical data of electricity, heat and gas loads, electricity price, gas price and meteorological condition (temperature and humidity) data, and accords with the actual situation of a regional comprehensive energy system; the popularization of terminal equipment such as an intelligent electric meter and the like is considered, and the load is classified by adopting a clustering method, so that the distributed prediction of the load of the comprehensive energy system is facilitated; by off-line calculation of excavation load characteristics, fine modeling of electric, thermal and gas loads is realized; the integration of the accuracy and the timeliness of the load prediction of the comprehensive energy system is realized by combining the offline calculation with the online prediction; the load prediction result is corrected by constructing a load prediction error model of the comprehensive energy system, so that the load prediction precision of the comprehensive energy system is improved, massive data information can be fully utilized, the high-precision prediction of the load of the comprehensive energy system is realized, and the safe and reliable operation of the comprehensive energy system is facilitated.
The principle and the embodiment of the present invention are explained by applying specific examples, and the above description of the embodiments is only used to help understanding the method and the core idea of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, this summary should not be construed as limiting the invention.
Claims (5)
1. A distributed load prediction method for an integrated energy system is characterized by comprising the following steps:
acquiring an electric load, a thermal load and an air load of the comprehensive energy system, acquiring an electric load time sequence curve, a thermal load time sequence curve and an air load time sequence curve, and acquiring external factor data; the external factor data comprises electricity price, gas price and meteorological condition data;
calculating a load characteristic index from the electrical load, the thermal load, and the air load;
carrying out load clustering according to the load characteristic index, the electric load time sequence curve, the heat load time sequence curve and the gas load time sequence curve to obtain various loads;
establishing an offline load prediction model for each type of load according to the external factor data;
acquiring load data of the current day and external factor data of a day to be predicted aiming at each type of load, and performing online load prediction by adopting the offline load prediction model according to the load data of the current day and the external factor data of the day to be predicted to obtain a day load to be predicted;
summing the daily loads to be predicted of each type of load to obtain the total load of the daily comprehensive energy system to be predicted;
before calculating a load characteristic index from the electrical load, the thermal load, and the air load, further comprising:
correcting deviation data of the electric load, the heat load and the gas load by adopting a data curve fitting method;
normalizing the corrected electrical load, the corrected thermal load and the corrected gas load to obtain the normalized electrical load, the normalized thermal load and the normalized gas load;
the calculating a load characteristic index according to the electrical load, the thermal load and the air load specifically includes:
the daily average load is calculated according to the following formula:
the daily load rate is calculated according to the following formula:
the peak time energy consumption rate is calculated according to the following formula:
the valley time energy consumption rate is calculated according to the following formula:
in the formula, KavDenotes the average daily load, EallExpressing daily energy, wherein the daily energy is one of daily electricity, daily heat and daily gas, T expresses the number of daily time segments, T is 24, KdDenotes the daily load rate, PmaxDenotes the daily maximum load, ρpRepresents the peak specific energy consumption, EpRepresenting energy used during peak hours, pvIndicating the energy consumption rate at valley time, EvRepresents energy usage during the daily trough period;
the load clustering is performed according to the load characteristic index, the electrical load time sequence curve, the thermal load time sequence curve and the gas load time sequence curve to obtain multiple types of loads, and the method specifically comprises the following steps:
load clustering was performed according to the following formula:
wherein F represents a clustering objective function, K represents the number of clusters, M represents the number of load time sequence curves, and hk,mRepresenting an identification variable, h when the load curve m belongs to the load class kk,mThe value is 1, h when the load curve m does not belong to the load class kk,mThe value is 0, alpha represents the weight coefficient,
representing the load characteristic distance between the load curve m and the load class k,
representing the load time-series curve distance, λ, between the load curve m and the load class k1Denotes a first coefficient, λ2Denotes the second coefficient, λ3Denotes a third coefficient, λ4Represents a fourth coefficient and satisfies λ1+λ2+λ3+λ4＝1；Kav,mDenotes the daily average load factor, K, of the load curve mav,kDenotes the daily average load factor, K, of the load class Kd,mDenotes the daily load factor, K, of the load curve md,kDenotes the daily load rate, ρ, of the load class kp,mRepresents the peak-time energy consumption rate, ρ, of the load curve mp,kRepresenting the peak-time energy consumption rate, ρ, of the load class kv,mThe energy consumption rate at valley time, ρ, expressed as the load curve mv,kRepresenting the valley time energy consumption rate of the load category k; p ism,tRepresenting the load value, P, of the load curve m to be clustered during the t periodk,tAnd representing the load value of the load category k to be clustered in the period t.
2. The distributed load prediction method for the integrated energy system according to claim 1, wherein the establishing of the offline load prediction model for each type of load according to the external factor data specifically comprises:
acquiring one type of data in the external factor data; the data of one type is one of electricity price, gas price and meteorological condition data;
calculating the correlation coefficient of the class data and each class of load;
comparing the correlation coefficient with a preset threshold value; if the correlation coefficient is larger than the preset threshold value, determining the class of data as a correlation factor corresponding to the load type; otherwise, determining whether other types of data in the external factor data are relevant factors corresponding to the load types;
and establishing an offline load prediction model according to the correlation factors and the load data corresponding to the correlation factors.
3. The distributed load forecasting method of the integrated energy system according to claim 2, further comprising, before performing the online load forecasting:
acquiring a current day prediction error of each type of load, a next day prediction day load of each type of load and correlation factors corresponding to the load types;
and establishing a load prediction error model according to the prediction error of the current day, the next day prediction day load and the correlation factor corresponding to the load type.
4. The distributed load prediction method of the integrated energy system according to claim 3, wherein the acquiring load data of a current day and external factor data of a day to be predicted for each type of load, and performing online load prediction by using the offline load prediction model according to the load data of the current day and the external factor data of the day to be predicted to obtain a daily load to be predicted specifically comprises:
generating a comprehensive energy system load prediction model according to the offline load prediction model and the load prediction error model;
and acquiring a current day prediction error, current day load data and external factor data of a day to be predicted aiming at each type of load, and performing online load prediction by adopting the comprehensive energy system load prediction model according to the current day prediction error, the current day load data and the external factor data of the day to be predicted to obtain the day load to be predicted.
5. An integrated energy system distributed load prediction system, comprising:
the data acquisition module is used for acquiring the electric load, the heat load and the air load of the comprehensive energy system, acquiring an electric load time sequence curve, a heat load time sequence curve and an air load time sequence curve and acquiring external factor data; the external factor data comprises electricity price, gas price and meteorological condition data;
a load characteristic index calculation module for calculating a load characteristic index from the electrical load, the thermal load and the air load;
the load clustering module is used for carrying out load clustering according to the load characteristic index, the electric load time sequence curve, the heat load time sequence curve and the gas load time sequence curve to obtain various loads;
the offline load prediction model establishing module is used for establishing an offline load prediction model for each type of load according to the external factor data;
the online load prediction module is used for acquiring load data of the current day and external factor data of a day to be predicted aiming at each type of load, and performing online load prediction by adopting the offline load prediction model according to the load data of the current day and the external factor data of the day to be predicted to obtain a daily load to be predicted;
the comprehensive energy system total load calculation module is used for summing the daily loads to be predicted of each type of load to obtain the daily comprehensive energy system total load to be predicted;
before calculating the load characteristic index according to the electric load, the heat load and the air load, the method further comprises the following steps:
correcting deviation data of the electrical load, the thermal load and the gas load by adopting a data curve fitting method;
normalizing the corrected electrical load, the corrected thermal load and the corrected gas load to obtain the normalized electrical load, the normalized thermal load and the normalized gas load;
the calculating a load characteristic index according to the electrical load, the thermal load and the air load specifically includes:
the daily average load is calculated according to the following formula:
the daily load rate is calculated according to the following formula:
the peak energy consumption rate was calculated according to the following formula:
the valley time energy consumption rate is calculated according to the following formula:
in the formula, KavDenotes the daily average load, EallExpressing daily energy, wherein the daily energy is one of daily electricity consumption, daily heat and daily gas, T represents the number of daily time segments, T is 24, KdRepresenting the daily load rate, PmaxRepresents the daily maximum load, ρpRepresents the peak specific energy consumption, EpRepresenting energy used during peak hours, pvIndicating the energy consumption rate at valley time, EvRepresents energy usage during the daily trough period;
the load clustering is performed according to the load characteristic index, the electrical load time sequence curve, the thermal load time sequence curve and the gas load time sequence curve to obtain various types of loads, and the method specifically comprises the following steps:
load clustering is performed according to the following formula:
wherein F represents a clustering objective function, K represents the number of clusters, M represents the number of load time sequence curves, and hk,mRepresenting an identification variable, h when the load curve m belongs to the load class kk,mThe value is 1, h when the load curve m does not belong to the load class kk,mThe value is 0, alpha represents a weight coefficient,
representing the load characteristic distance between the load curve m and the load class k,
representing the load time-series curve distance, λ, between the load curve m and the load class k1Denotes a first coefficient, λ2Denotes a second coefficient, λ3Denotes the third coefficient, λ4Represents a fourth coefficient and satisfies λ1+λ2+λ3+λ4＝1；Kav,mDenotes the daily average load factor, K, of the load curve mav,kDenotes the daily average load factor, K, of the load class Kd,mDenotes the daily load factor, K, of the load curve md,kDenotes the daily load rate, ρ, of the load class kp,mRepresents the peak-time energy consumption rate, ρ, of the load curve mp,kRepresenting the peak-time energy consumption rate, ρ, of the load class kv,mThe energy consumption rate at valley time, ρ, expressed as the load curve mv,kThe valley time energy consumption rate of the load category k is represented; p ism,tRepresents the load value, P, of the load curve m to be clustered at time tk,tAnd representing the load value of the load category k to be clustered in the period t.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911035294.0A CN110796307B (en) | 2019-10-29 | 2019-10-29 | Distributed load prediction method and system for comprehensive energy system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911035294.0A CN110796307B (en) | 2019-10-29 | 2019-10-29 | Distributed load prediction method and system for comprehensive energy system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110796307A CN110796307A (en) | 2020-02-14 |
| CN110796307B true CN110796307B (en) | 2022-07-22 |
Family
ID=69441590
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911035294.0A Active CN110796307B (en) | 2019-10-29 | 2019-10-29 | Distributed load prediction method and system for comprehensive energy system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110796307B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111191856B (en) * | 2020-02-17 | 2023-12-08 | 天津大学 | Regional comprehensive energy system multi-energy load prediction method considering time sequence dynamic characteristics and coupling characteristics |
| CN111583059B (en) * | 2020-04-20 | 2024-01-23 | 上海电力大学 | Distributed energy station typical daily load acquisition method based on k-means clustering |
| CN111815026A (en) * | 2020-06-09 | 2020-10-23 | 国网山东省电力公司经济技术研究院 | Multi-energy system load prediction method based on feature clustering |
| CN111754037B (en) * | 2020-06-19 | 2023-01-20 | 国网河南省电力公司经济技术研究院 | Long-term load hybrid prediction method for regional terminal integrated energy supply system |
| CN112862179A (en) * | 2021-02-03 | 2021-05-28 | 国网山西省电力公司吕梁供电公司 | Energy consumption behavior prediction method and device and computer equipment |
| CN115081902B (en) * | 2022-06-30 | 2024-04-09 | 国网北京市电力公司 | Comprehensive planning method, device, equipment and medium based on source network load storage cooperation |
| CN117669839B (en) * | 2024-02-01 | 2024-04-30 | 山东赛马力发电设备有限公司 | Distributed load prediction method and system for comprehensive energy system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU2003220728A1 (en) * | 2002-07-22 | 2004-02-05 | Darby Matthew Ross | An intergrated energy utilization system |
| CN104820876A (en) * | 2015-05-21 | 2015-08-05 | 国家电网公司 | Short-term load forecasting method and system |
| CN106779129A (en) * | 2015-11-19 | 2017-05-31 | 华北电力大学（保定） | A kind of Short-Term Load Forecasting Method for considering meteorologic factor |
| CN107657266A (en) * | 2017-08-03 | 2018-02-02 | 华北电力大学（保定） | A kind of load curve clustering method based on improvement spectrum multiple manifold cluster |
| CN109670694A (en) * | 2018-12-10 | 2019-04-23 | 国网河南省电力公司经济技术研究院 | A kind of multipotency source supply system load forecasting method |
-
2019
- 2019-10-29 CN CN201911035294.0A patent/CN110796307B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU2003220728A1 (en) * | 2002-07-22 | 2004-02-05 | Darby Matthew Ross | An intergrated energy utilization system |
| CN104820876A (en) * | 2015-05-21 | 2015-08-05 | 国家电网公司 | Short-term load forecasting method and system |
| CN106779129A (en) * | 2015-11-19 | 2017-05-31 | 华北电力大学（保定） | A kind of Short-Term Load Forecasting Method for considering meteorologic factor |
| CN107657266A (en) * | 2017-08-03 | 2018-02-02 | 华北电力大学（保定） | A kind of load curve clustering method based on improvement spectrum multiple manifold cluster |
| CN109670694A (en) * | 2018-12-10 | 2019-04-23 | 国网河南省电力公司经济技术研究院 | A kind of multipotency source supply system load forecasting method |
Non-Patent Citations (1)
| Title |
|---|
| 基于径向基函数神经网络的综合能源系统;翟晶晶，吴晓蓓，王力立;《电力需求侧管理》;20190731;第21卷(第4期);23-27 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110796307A (en) | 2020-02-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110796307B (en) | Distributed load prediction method and system for comprehensive energy system | |
| CN113962364B (en) | Multi-factor power load prediction method based on deep learning | |
| Wang et al. | Research and application of a hybrid forecasting framework based on multi-objective optimization for electrical power system | |
| Hu et al. | A hybrid model based on CNN and Bi-LSTM for urban water demand prediction | |
| CN110909912B (en) | Park electric power system net load combination prediction method based on self-adaptive error feedback | |
| Shang et al. | Short-term load forecasting based on PSO-KFCM daily load curve clustering and CNN-LSTM model | |
| CN104598986B (en) | Methods of electric load forecasting based on big data | |
| CN111754037B (en) | Long-term load hybrid prediction method for regional terminal integrated energy supply system | |
| CN108376262B (en) | Analytical model construction method for typical characteristics of wind power output | |
| CN110380444B (en) | Capacity planning method for distributed wind power orderly access to power grid under multiple scenes based on variable structure Copula | |
| CN113642787A (en) | Regional power distribution network gridding load interval prediction method based on distribution and sale power competition situation | |
| CN111027775A (en) | Step hydropower station generating capacity prediction method based on long-term and short-term memory network | |
| CN104598985A (en) | Power load forecasting method | |
| CN112232575A (en) | Comprehensive energy system regulation and control method and device based on multivariate load prediction | |
| CN112365056A (en) | Electrical load joint prediction method and device, terminal and storage medium | |
| CN111476397B (en) | Bus load prediction method and system applied to medium and long term electric quantity safety check | |
| CN110991764B (en) | Day-ahead rolling optimization method for comprehensive energy system | |
| CN115689146A (en) | Multi-energy intelligent building load scheduling method and system | |
| CN111754029A (en) | Community load prediction system | |
| CN110135617A (en) | A kind of prediction technique of improvement short term under meteorological factor influence based on part throttle characteristics | |
| CN106934545B (en) | Power station group joint economic dispatching method and system | |
| CN115829418A (en) | Power consumer load characteristic portrait construction method and system suitable for load management | |
| CN115796915A (en) | Electricity price prediction method and system for electricity trading market | |
| Zhu et al. | Load Forecasting of the power system: an investigation based on the method of random forest regression | |
| CN114997470A (en) | Short-term power load prediction method based on LSTM neural network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |