CN114362268A - Wind-solar-charge two-stage prediction-based optimized scheduling method for comprehensive energy power supply system - Google Patents

Wind-solar-charge two-stage prediction-based optimized scheduling method for comprehensive energy power supply system Download PDF

Info

Publication number
CN114362268A
CN114362268A CN202210032942.2A CN202210032942A CN114362268A CN 114362268 A CN114362268 A CN 114362268A CN 202210032942 A CN202210032942 A CN 202210032942A CN 114362268 A CN114362268 A CN 114362268A
Authority
CN
China
Prior art keywords
wind
load
power
solar
prediction
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202210032942.2A
Other languages
Chinese (zh)
Inventor
黄晶晶
郑礼靖
张爱民
宋子健
徐畅甲
柴晓莉
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xian Jiaotong University
Original Assignee
Xian Jiaotong University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Xian Jiaotong University filed Critical Xian Jiaotong University
Priority to CN202210032942.2A priority Critical patent/CN114362268A/en
Publication of CN114362268A publication Critical patent/CN114362268A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E10/00—Energy generation through renewable energy sources
    • Y02E10/50—Photovoltaic [PV] energy
    • Y02E10/56—Power conversion systems, e.g. maximum power point trackers
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E10/00—Energy generation through renewable energy sources
    • Y02E10/70—Wind energy
    • Y02E10/76—Power conversion electric or electronic aspects
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
    • Y02E60/30—Hydrogen technology
    • Y02E60/36—Hydrogen production from non-carbon containing sources, e.g. by water electrolysis
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E70/00—Other energy conversion or management systems reducing GHG emissions
    • Y02E70/30—Systems combining energy storage with energy generation of non-fossil origin

Landscapes

  • Supply And Distribution Of Alternating Current (AREA)
  • Charge And Discharge Circuits For Batteries Or The Like (AREA)

Abstract

The invention relates to the field of optimized scheduling of an integrated energy power supply system, in particular to an optimized scheduling method of an integrated energy power supply system based on wind-solar-load two-stage prediction, which comprises the following steps: s1, obtaining schedulable capacity allocation of the power grid and the hydrogen energy storage system on the same day according to the wind power output prediction interval and the power load prediction interval on the previous day; according to the super-shortObtaining the wind-solar output accurate value P at any time of the day by the period prediction model* w&lAnd the accurate value P of the electricity load* load(ii) a And S2, combining the results obtained in the step S1 to switch the power supply mode of the comprehensive energy system. The method considers the randomness and uncertainty of the wind power and photovoltaic power generation processes, realizes accurate prediction of wind power, photovoltaic power output and power load, is convenient for arranging a power grid dispatching plan in advance, configuring fuel cells and electrolyzing water to produce hydrogen power, reduces the risk value and cost of random dispatching of a power supply system, and realizes optimized dispatching of a comprehensive energy power supply system.

Description

Wind-solar-charge two-stage prediction-based optimized scheduling method for comprehensive energy power supply system
Technical Field
The invention relates to the field of optimization scheduling of an integrated energy power supply system, in particular to an optimization scheduling method of an integrated energy power supply system based on wind-solar-load two-stage prediction.
Background
Along with the increase of global carbon dioxide emission, the greenhouse effect is increasingly intensified, global climate changes, extreme meteorological disasters frequently occur, and serious casualties and economic losses are caused to the world. Therefore, the application of green energy in the industrial field is promoted, the application proportion of renewable energy sources such as solar energy, wind energy, hydrogen energy and the like in an industrial park is improved, the electrification of industrial energy equipment is promoted, and the method becomes one of the main targets of the current social development.
In an industrial power supply system, renewable energy power generation modes such as photovoltaic and wind power are more and more widely applied by virtue of the advantages of green and low carbon. By 12 months in 2020, the integrated installation of grid-connected wind power and photovoltaic in China respectively reaches 2.81 hundred million kW and 2.53 hundred million kW, and the year-on-year growth is 33.1% and 23.9%. In 2020, the national wind power and photovoltaic cumulative power generation amount is 7270 hundred million kilowatt hours, the year-on-year increase is 15.1 percent, the total power generation amount is 9.5 percent, and the replacement effect of green electric energy of new energy is continuously enhanced. However, due to uncertainty of photovoltaic and wind power generation quantities, redundant electric energy cannot be timely consumed by a demand side, and accordingly the phenomenon of wind and light abandoning is serious, and great resource waste is caused.
Therefore, in order to solve the problem that the light is abandoned by abandoned wind of photovoltaic and wind power, the energy utilization rate is improved, and the electric energy generated by the photovoltaic and the wind power can be stored by introducing a hydrogen energy storage system. The hydrogen energy storage system can utilize surplus electric energy generated by the new energy to produce hydrogen, and the hydrogen is stored or used by downstream industries; when the load of the power system is in a peak, the stored hydrogen energy can be used for supplying power to the power utilization end by using the fuel cell, so that uninterrupted power supply is realized. Compared with other energy storage modes, the hydrogen energy storage has the advantages of high efficiency, environmental protection, no pollution, long-term storage and high energy density, and can further reduce the carbon emission in the energy storage process.
In a wind-solar-charged comprehensive energy system, due to uncertainty of wind-solar output and power load, system capacity configuration is over-ideal, and the utilization rate of equipment is low. There are studies showing that: the uncertainty of photovoltaic output and load is described by adopting interval prediction, the comprehensive energy system is subjected to day-ahead optimized scheduling based on an interval linear programming method, and the influence of photovoltaic and load fluctuation on energy scheduling in a day is not considered due to the fact that the prediction period is long and the prediction interval is not accurate enough.
And a wind and light output value can be accurately predicted by establishing a wind and light typical scene set-based two-level collaborative optimization configuration model of the comprehensive energy system, but the power load is not predicted, and the effect of load demand response on new energy consumption is ignored.
In conclusion, the existing optimization scheduling method of the comprehensive energy system has the problems of low wind-solar output prediction precision and weak wind-solar and load cooperative scheduling capability. The high wind and solar load prediction precision is beneficial to reasonably configuring the capacity of the energy storage equipment and reducing the resource waste in the power grid scheduling process; the wind, light and load multi-energy collaborative scheduling can enhance the flexibility of energy scheduling in time, fully exert the consumption capacity of the flexible load on new energy, and improve the energy utilization rate. Therefore, the research on the wind-solar-load high-precision prediction and the multi-energy collaborative scheduling strategy has great significance for the efficient operation of the comprehensive energy power supply system.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a wind-solar-charge two-stage prediction-based optimal scheduling method for a comprehensive energy power supply system.
The invention is realized by the following technical scheme:
a comprehensive energy power supply system optimal scheduling method based on wind-solar-charge two-stage prediction comprises the following steps:
s1, acquiring the current-day power grid configuration and the schedulable capacity of the hydrogen energy storage system according to the wind power output prediction interval and the power load prediction interval of the previous day; obtaining the wind-solar output accurate value P at any time of the day according to the ultra-short-term prediction model* w&lAnd the accurate value P of the electricity load* load；
S2, the power grid configuration of the day and the predicted wind-solar output accurate value P of the hydrogen energy storage system within the schedulable capacity and the interval preset time are carried out* w&lAnd the accurate value P of the electricity load* loadAnd combining to switch the power supply mode of the comprehensive energy system.
Preferably, in S1, the wind power output prediction section and the electric load prediction section on the previous day are both obtained by a section prediction model established by a particle swarm algorithm and a limit learning machine.
Preferably, the step of obtaining the wind power output prediction interval and the electric load prediction interval of the previous day is as follows: firstly, extreme learning and interval prediction model establishment are utilized, then the interval prediction model is trained, then iterative optimization is carried out on the interval prediction model by utilizing particle swarm optimization, finally, wind-light output data and power load data of the previous day are brought into the optimized interval prediction model, and a wind-light output prediction interval and a power load prediction interval of the previous day are obtained, wherein the wind-light output prediction interval is marked as Pw&l∈[Pw&l_min,Pw&l_max]，Pw&l_minFor minimum wind-solar output prediction, Pw&l_maxThe power load prediction interval is marked as P for the maximum wind and light output prediction valueload∈[Pload_min,Pload_max]，Pload_minFor minimum electrical load prediction, Pload_maxThe predicted value is the maximum power load.
Preferably, in S1, the power grid configuration of the day includes a power grid scheduling plan, a fuel cell configuration, and an electrolyzed water hydrogen production power.
Preferably, the power grid configuration rules are as follows:
Figure BDA0003467141820000031
wherein, the input power of the hydrogen energy storage system is the power P for producing hydrogen by electrolyzing waterH2(ii) a The output power of the hydrogen energy storage system, namely the discharge power of the fuel cell is PbatteryThe upper limit of the discharge power of the fuel pool is Pbattery_max(ii) a Power grid output power Pgrid。
Preferably, the wind-solar power precise value P* w&lAnd the accurate value P of the electricity load* loadThe acquisition steps are as follows: firstly, establishing an ultra-short-term prediction model, and then training the ultra-short-term prediction model by using wind-solar output data and electric load data of the previous day to obtain an optimized ultra-short-term prediction model; finally, wind and light output data and power load data at the previous moment are brought into the optimized ultra-short-term prediction model to obtain the ultra-short-term wind and light output accurate value P at the current moment* w&lAnd the accurate value P of the electricity load* load。
Preferably, the prediction interval time of the ultra-short term prediction model is 15 min.
Preferably, in S2, the ultra-short term prediction model is obtained by a quadratic programming method using a least squares support vector machine.
Preferably, the power supply mode of the integrated energy system comprises a wind-light spontaneous self-use mode, a wind-light residual power hydrogen production mode, a wind-light hydrogen-off-grid mode and a wind-light hydrogen-on-grid mode.
Preferably, the switching rule of the power supply mode of the integrated energy system is as follows:
when P is present* w&l＝P* loadWhen the wind-solar self-generation energy system is used, the comprehensive energy system enters a wind-solar self-generation self-use mode;
when P is present* load<P* w&l≤P* load+PH2The comprehensive energy system enters a wind-solar residual electricity hydrogen production mode;
when P is present* w&l<P* load≤P* w&l+PbatteryThe comprehensive energy system enters a wind-solar-hydrogen-off-grid mode;
when P is present* w&l+Pbattery<P* load≤P* w&l+Pbattery+PgridAnd the comprehensive energy system enters a wind-solar-hydrogen-grid-connected mode.
Compared with the prior art, the invention has the following beneficial effects:
the comprehensive energy power supply system optimal scheduling method based on wind-solar-charge two-stage prediction considers randomness and uncertainty of wind power and photovoltaic power generation processes, realizes accurate prediction of wind-solar power output and power load, is convenient for arranging a power grid scheduling plan in advance, configuring fuel cells and electrolyzed water hydrogen production power, reduces the risk value and cost of random scheduling of a power supply system, and realizes optimal scheduling of the comprehensive energy power supply system according to the principle of preferentially using green energy and improving the energy utilization rate.
At the power generation end, the power load is supplied by green energy such as wind and light, the dependence on the power supply of a large power grid is reduced, and the carbon emission is reduced.
At the power consumption end, the consumption capacity of the flexible load to new energy power generation is enhanced by reasonably arranging the power consumption load in each time period, and the energy utilization rate is improved.
At the energy storage end, when the load is in a low valley, the hydrogen energy storage system electrolyzes water to produce hydrogen by utilizing the surplus electric energy of wind power and photovoltaic power; when the load is in a peak, the stored hydrogen can supply power to the power load through the hydrogen fuel cell, so that the purposes of peak clipping, valley filling and uninterrupted power supply of a power supply system are achieved.
Furthermore, an interval prediction model is established by utilizing a particle swarm algorithm and a limit learning machine (PSO-ELM), the range of wind power output and power load output every hour in the next day is predicted, and a basis is provided for making an energy scheduling plan in advance.
Furthermore, an Extreme Learning Machine (ELM) is a novel single hidden layer forward network with excellent performance, the output weight of the network can be analyzed through one-step calculation, the generalization capability and the learning speed of the network are greatly improved, the ELM has strong nonlinear fitting capability, and the calculated amount and the search space are greatly reduced.
Furthermore, according to the prediction interval of wind, light and power output and the power load, a power grid dispatching plan is arranged in advance, and the fuel cell and the water electrolysis hydrogen production power are configured, so that sufficient power supply is ensured, and the amount of abandoned wind and light is reduced.
Furthermore, an ultra-short term prediction model is established by using a least square support vector machine (LS-SVM), and accurate values of wind-solar output and power load are predicted every 15min, so that accurate judgment conditions are provided for executing power supply mode switching.
Furthermore, a quadratic programming method is adopted by a least square support vector machine (LS-SVM) to solve the problem of function estimation, the quadratic programming method is directly converted into the problem of solving a linear equation set, the complexity of calculation can be simplified, the precision and the convergence speed of operation are improved, and the method is suitable for the ultra-short-term prediction scene of data.
Drawings
FIG. 1 is a schematic flow chart of an optimal scheduling method of a comprehensive energy power supply system based on wind-solar-load two-stage prediction according to the invention;
FIG. 2 is a schematic structural diagram of an integrated energy power supply system based on wind-solar-charge two-stage prediction.
Detailed Description
The present invention will now be described in further detail with reference to specific examples, which are intended to be illustrative, but not limiting, of the invention.
The invention discloses a wind-solar-load two-stage prediction-based optimal scheduling method for a comprehensive energy power supply system, which comprises the following steps of referring to FIG. 1:
s1, acquiring the power grid configuration and the schedulable capacity of the hydrogen energy storage system on the current day according to the wind power output prediction interval and the power load prediction interval on the previous day; obtaining the wind-solar output accurate value P at any time of the day according to the ultra-short-term prediction model* w&lAnd the accurate value P of the electricity load* loadThe prediction interval time of the ultra-short term prediction model is 15 min.
The method for acquiring the wind power output prediction interval and the electric load prediction interval in the previous day comprises the following steps:
firstly, establishing an interval prediction model by using an extreme learning machine, wherein the expression of the interval prediction model is as follows:
Figure BDA0003467141820000051
in the formula: i is the number of the hidden layer, K is the total number of the hidden layer, j is the number of input and output data, and N is the number of training samples; pjPredicting a model target output value for the interval; x is the number ofjNormalizing the input data for the interval prediction model; beta is aiIs the output weight connecting the hidden layer and the output layer; gi(ωi·xj+bi) As an activation function of the ith hidden layer node, ωiIs the weight connecting the ith hidden layer node and the input node; biIs the bias of the ith hidden layer of the network.
And then training the interval prediction model, wherein the steps are as follows:
according to the definition of the prediction interval, the target value P to be predictedjAt the nominal confidence level, the prediction target value falls within the constructed prediction interval with a probability of approaching the nominal confidence level, i.e.:
Figure BDA0003467141820000061
in the formula: pjPredicting a model target output value for the interval; x is the number ofjNormalizing the input data for the interval prediction model;
Figure BDA0003467141820000062
are respectively PjThe upper and lower bounds of the interval are predicted at a nominal confidence level.
The evaluation indexes of the prediction interval mainly comprise: prediction Interval Coverage (PICP), prediction interval normalized average bandwidth (PINAW), accumulated bandwidth deviation (AWD), Prediction Interval Coverage Error (PICE), and expressions of various indexes are as follows:
Figure BDA0003467141820000063
in the formula: α is the significance level; j is the input and output data number; n is the number of training samples; b isjB, if the prediction target value is contained in the upper and lower limits of the interval prediction for the Boolean quantity j1, otherwise Bj is 0; r is a predicted target value change range and is used for carrying out normalization processing on the average bandwidth; r is used for carrying out normalization processing on the accumulated bandwidth deviation;
Figure BDA0003467141820000064
are respectively PjPredicting upper and lower boundaries of the interval at a nominal confidence level;
Figure BDA0003467141820000071
accumulating bandwidth deviations for the prediction intervals;
Figure BDA0003467141820000072
for predicting the bandwidth deviation degree of the jth iteration of the interval, the expression is as follows:
Figure BDA0003467141820000073
for the wind power generation and power load prediction interval, the accuracy of the prediction interval is higher when the coverage error (PICE) of the prediction interval, the average bandwidth (PINAW) of the prediction interval and the accumulated bandwidth deviation (AWD) are satisfied. Therefore, the comprehensive optimization objective function F based on PICE, PINAW and AWD is defined as follows:
Figure BDA0003467141820000074
in the formula: gamma rayj、
Figure BDA0003467141820000075
λjAnd respectively adjusting the weight coefficients aiming at the predicted target coverage rate deviation, the predicted interval average bandwidth and the accumulated bandwidth deviation to control the influence ratio of different criteria on the optimization effect. | is an absolute value for PICE, PINAW, AWD.
For determining each hidden layer beta of ELM interval prediction modeli、ωi、biThe approximation degree of the prediction interval and the actual interval is evaluated by utilizing the comprehensive optimization objective function F, and the model is trained by combining the wind power output and the historical data of the power load before the day.
Iterative optimization is carried out on the interval prediction model by utilizing a Particle Swarm Optimization (PSO) algorithm to obtain beta of each hidden layeri、ωi、biFurther perfecting the interval prediction model.
And finally, substituting the wind-light output data and the electrical load data of the previous day into the optimized interval prediction model to obtain the wind-light output prediction interval and the electrical load prediction interval of the previous day, wherein the wind-light output prediction interval is marked as Pw&l∈[Pw&l_min,Pw&l_max]，Pw&l_minFor minimum wind-solar output prediction, Pw&l_maxThe power load prediction interval is marked as P for the maximum wind and light output prediction valueload∈[Pload_min,Pload_max]，Pload_minFor minimum electrical load prediction, Pload_maxThe predicted value is the maximum power load.
The current day of power grid configuration comprises a power grid dispatching plan, fuel cell configuration and water electrolysis hydrogen production power. In order to guarantee uninterrupted power supply and reduce wind and light abandonment, the configuration format of the schedulable capacity of the power grid and the hydrogen energy storage is as follows:
Figure BDA0003467141820000081
wherein the hydrogen energy storage systemInput power, namely power P for producing hydrogen by electrolyzing waterH2(ii) a The output power of the hydrogen energy storage system, namely the discharge power of the fuel cell is PbatteryThe upper limit of the discharge power of the fuel pool is Pbattery_max(ii) a Power grid output power Pgrid。
Wind-solar power precise value P* w&lAnd the accurate value P of the electricity load* loadThe acquisition steps are as follows: firstly, establishing an ultra-short term prediction model, wherein the expression is as follows:
Figure BDA0003467141820000082
in the formula: k is an input data number; m is the number of training samples; p*(x) Target output values for the ultra-short term prediction model; x is the number ofkInput data at the kth sampling moment; a iskLagrange multipliers for the kth sampling instant; k (x)kX) is a kernel function of a least squares support vector machine model,
Figure BDA0003467141820000083
e is the deviation.
Training the ultra-short-term prediction model by using the wind-solar output data and the electrical load data of the previous day to obtain an optimized ultra-short-term prediction model;
to determine a of an LS-SVM interval prediction modelkThe parameter values of delta and E need to train the model by using the wind-solar output and the daily historical data of the power load, the average error is used as an evaluation standard to judge the approach degree of the predicted value and the actual value of the ultra-short-term prediction model, and a when the minimum average error is met is obtainedkδ, E. The average error e of the ultra-short term prediction model is calculated as follows:
Figure BDA0003467141820000084
in the formula: e is the average error of the ultra-short term prediction model; k is an output value number; m is the number of training samples;
Figure BDA0003467141820000091
is the actual output value; skIs the predicted output value.
Finally, wind and light output data and power load data at the previous moment are brought into the optimized ultra-short-term prediction model to obtain the ultra-short-term wind and light output accurate value P at the current moment* w&lAnd the accurate value P of the electricity load* load。
S2, the power grid configuration of the day and the predicted wind-solar output accurate value P of the hydrogen energy storage system within the schedulable capacity and the interval preset time are carried out* w&lAnd the accurate value P of the electricity load* loadAnd combining to switch the power supply mode of the comprehensive energy system.
The power supply modes of the comprehensive energy system comprise a wind-light spontaneous self-use mode, a wind-light residual power hydrogen production mode, a wind-light hydrogen-off-grid mode and a wind-light hydrogen-grid-connected mode, and the switching rules are as follows:
when P is present* w&l＝P* loadWhen the wind-solar self-generation energy system is used, the comprehensive energy system enters a wind-solar self-generation self-use mode; namely, the wind-solar output is just equal to the electric load, and the generated energy of wind power and photovoltaic is completely absorbed by the electric load.
When P is present* load<P* w&l≤P* load+PH2The comprehensive energy system enters a wind-solar residual electricity hydrogen production mode; namely, the wind-solar output is greater than the power load, the generated energy of wind power and photovoltaic is sufficient, and the residual electric energy is available after the load is supplied. Therefore, surplus electric energy can be utilized to electrolyze water to produce hydrogen, the electric energy is stored in a hydrogen form, and the utilization rate of energy is improved.
When P is present* w&l<P* load≤P* w&l+PbatteryThe comprehensive energy system enters a wind-solar-hydrogen-off-grid mode; at this time, the hydrogen fuel cell power supply system is turned on. When P is present* w&l<P* load≤P* w&l+PbatteryTemporary, i.e. wind-solar power and hydrogen fuel cellsThe sum of the output power is more than or equal to the electric load. At the moment, the wind power, the photovoltaic and the fuel cell work in an off-grid mode to provide required electric energy for the load together.
When P is present* w&l+Pbattery<P* load≤P* w&l+Pbattery+PgridThe integrated energy system enters a wind-solar-hydrogen-grid-connected mode, namely the sum of wind-solar output and the output power of the fuel cell is smaller than the load, and the integrated energy power supply system enters the wind-solar-hydrogen-grid-connected mode. At the moment, the wind power, the photovoltaic and the fuel cell work in a grid-connected mode, and the power grid is used for supplementing the required electric energy for the load.
Referring to fig. 2, the invention also discloses a comprehensive energy power supply system based on wind-solar-charge two-stage prediction, which comprises a wind power generation system, a photovoltaic power generation system, a grid-connected and off-grid control system, an electrolytic water hydrogen production system, a fuel cell power supply system, a load, a direct current load, an alternating current bus and a direct current bus. The output end of the wind power photovoltaic system and the output end of the grid-connected and off-grid control system are connected with an alternating current bus, the input end of the electrolyzed water hydrogen production system is connected with a direct current bus, the output end of the electrolyzed water hydrogen production system is connected with the input end of a fuel cell power supply system, and the output end of the fuel cell power supply system is connected with the direct current bus. The alternating current load is connected with the alternating current bus, the direct current load is connected with the direct current bus, and the alternating current bus is connected with the direct current bus through the AC/DC converter.
On the alternating current side, the wind power generation system comprises a wind power generation device and an AC/AC converter, wherein the output end of the wind power generation device is connected to the AC/AC converter, and is connected to one end of an alternating current bus after voltage class conversion; the grid-connected control system comprises a grid-connected control switch and an alternating current power grid.
On the direct current side, the photovoltaic power generation system comprises a photovoltaic power generation device and a DC/DC converter, wherein the output end of the photovoltaic power generation device is connected to the DC/DC converter, and is connected to one end of a direct current bus after voltage grade conversion; the water electrolysis hydrogen production system comprises an electrolytic cell and a hydrogen storage tank, wherein a power input interface of the electrolytic cell is connected to one end of a direct current bus, a hydrogen outlet is connected to an inlet of the hydrogen storage tank, and an outlet of the hydrogen storage tank is connected to an input end of a fuel cell power supply system;
the fuel cell power supply system comprises a proton exchange membrane fuel cell power generation device and a DC/DC converter, wherein the input end of the fuel cell system is connected with the hydrogen storage tank, and the output end of the fuel cell system is connected to the DC/DC converter and is connected to one end of the direct current bus after voltage grade conversion.
At the load and bus end, an AC load is connected to one end of the AC bus, and a DC load is connected to one end of the DC bus. The alternating current bus and the direct current bus are connected through an AC/DC converter.
The wind power generation system drives the windmill blades to rotate by utilizing the wind power in the nature, and the rotating speed is increased by the speed increaser to enable the generator to generate alternating current; the photovoltaic power generation system directly converts light energy into electric energy through a photovoltaic effect, and the generated electric energy is direct current;
the water electrolysis hydrogen production system electrolyzes water to produce hydrogen by utilizing the surplus generated energy of photovoltaic and wind power and stores the generated hydrogen in a hydrogen storage tank; the fuel cell power supply system consumes fuel such as hydrogen, oxygen and the like to generate electric energy, and supplies power to an electric load through a DC/DC converter.
And the grid-connected control system controls the off-grid operation or the grid-connected operation of the comprehensive energy power supply system through a switch. When the new energy power generation can supply energy to the power load, the system is operated off-grid; and when the new energy power generation can not meet the demand of the power load, the system is in grid-connected operation.

Claims (9)

1. A comprehensive energy power supply system optimal scheduling method based on wind-solar-charge two-stage prediction is characterized by comprising the following steps:
s1, acquiring the current-day power grid configuration and the schedulable capacity of the hydrogen energy storage system according to the wind power output prediction interval and the power load prediction interval of the previous day; obtaining the wind-solar output accurate value P at any time of the day according to the ultra-short-term prediction model* w&lAnd the accurate value P of the electricity load* load；
S2, the power grid configuration of the day and the predicted wind-solar output accurate value P of the hydrogen energy storage system within the schedulable capacity and the interval preset time are carried out* w&lAnd the accurate value P of the electricity load* loadAnd combining to switch the power supply mode of the comprehensive energy system.
2. The method for optimizing and scheduling the comprehensive energy power supply system based on the wind-solar-electric two-stage prediction as claimed in claim 1, wherein in S1, the wind power output prediction interval and the electric load prediction interval of the previous day are both obtained through an interval prediction model established by a particle swarm algorithm and an extreme learning machine.
3. The wind-solar-electric two-stage prediction-based optimized scheduling method for the comprehensive energy power supply system according to claim 2, wherein the previous wind output prediction interval and power load prediction interval are obtained by the following steps: firstly, establishing an interval prediction model by using an extreme learning machine, then training the interval prediction model, then iteratively optimizing the interval prediction model by using a particle swarm algorithm, and finally substituting wind-solar output data and power load data of the previous day into the optimized interval prediction model to obtain a wind-solar output prediction interval and a power load prediction interval of the previous day, wherein the wind-solar output prediction interval is marked as Pw&l∈[Pw&l_min,Pw&l_max]，Pw&l_minFor minimum wind-solar output prediction, Pw&l_maxThe power load prediction interval is marked as P for the maximum wind and light output prediction valueload∈[Pload_min,Pload_max]，Pload_minFor minimum electrical load prediction, Pload_maxThe predicted value is the maximum power load.
4. The integrated energy supply system optimal scheduling method based on wind-solar-powered two-stage prediction as claimed in claim 3, wherein in S1, the power grid configuration of the day comprises a power grid scheduling plan, a fuel cell configuration and hydrogen production power by water electrolysis.
5. The wind-solar-electric two-stage prediction-based optimized scheduling method for the comprehensive energy power supply system according to claim 4, wherein the power grid configuration rule is as follows:
Figure FDA0003467141810000021
wherein, the input power of the hydrogen energy storage system is the power P for producing hydrogen by electrolyzing waterH2(ii) a The output power of the hydrogen energy storage system, namely the discharge power of the fuel cell is PbatteryThe upper limit of the discharge power of the fuel pool is Pbattery_max(ii) a Power grid output power Pgrid。
6. The method for optimizing and scheduling the comprehensive energy power supply system based on wind-solar-electric two-stage prediction as claimed in claim 1, wherein the accurate value P of wind-solar-electric output is* w&lAnd the accurate value P of the electricity load* loadThe acquisition steps are as follows: firstly, establishing an ultra-short-term prediction model, and then training the ultra-short-term prediction model by using wind-solar output data and electric load data of the previous day to obtain an optimized ultra-short-term prediction model; finally, wind and light output data and power load data at the previous moment are brought into the optimized ultra-short-term prediction model to obtain the ultra-short-term wind and light output accurate value P at the current moment* w&lAnd the accurate value P of the electricity load* load。
7. The method for optimizing and scheduling the comprehensive energy power supply system based on wind-solar-electric two-stage prediction as claimed in claim 6, wherein in S2, the ultra-short term prediction model is obtained by a quadratic programming method through a least square support vector machine.
8. The optimal scheduling method of the comprehensive energy power supply system based on wind-solar-charged two-stage prediction as claimed in claim 1, wherein the power supply modes of the comprehensive energy system comprise a wind-solar spontaneous self-use mode, a wind-solar residual electricity hydrogen production mode, a wind-solar hydrogen-off-grid mode and a wind-solar hydrogen-on-grid mode.
9. The wind-solar-charge two-stage prediction-based optimized scheduling method for the integrated energy power supply system according to claim 8, wherein the switching rule of the power supply mode of the integrated energy system is as follows:
when P is present* w&l＝P* loadWhen the wind-solar self-generation energy system is used, the comprehensive energy system enters a wind-solar self-generation self-use mode;
when P is present* load<P* w&l≤P* load+PH2The comprehensive energy system enters a wind-solar residual electricity hydrogen production mode;
when P is present* w&l<P* load≤P* w&l+PbatteryThe comprehensive energy system enters a wind-solar-hydrogen-off-grid mode;
when P is present* w&l+Pbattery<P* load≤P* w&l+Pbattery+PgridAnd the comprehensive energy system enters a wind-solar-hydrogen-grid-connected mode.
CN202210032942.2A 2022-01-12 2022-01-12 Wind-solar-charge two-stage prediction-based optimized scheduling method for comprehensive energy power supply system Pending CN114362268A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210032942.2A CN114362268A (en) 2022-01-12 2022-01-12 Wind-solar-charge two-stage prediction-based optimized scheduling method for comprehensive energy power supply system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210032942.2A CN114362268A (en) 2022-01-12 2022-01-12 Wind-solar-charge two-stage prediction-based optimized scheduling method for comprehensive energy power supply system

Publications (1)

Publication Number Publication Date
CN114362268A true CN114362268A (en) 2022-04-15

Family

ID=81109211

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210032942.2A Pending CN114362268A (en) 2022-01-12 2022-01-12 Wind-solar-charge two-stage prediction-based optimized scheduling method for comprehensive energy power supply system

Country Status (1)

Country Link
CN (1) CN114362268A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115238505A (en) * 2022-07-27 2022-10-25 国网山东省电力公司枣庄供电公司 Optimal configuration method and system for combined heat, power and hydrogen supply system based on wind and light station
CN116581742A (en) * 2023-05-16 2023-08-11 淮阴工学院 Chameleon algorithm-based flexible electric load scheduling system and chameleon algorithm-based flexible electric load scheduling method for regulating and controlling intelligent cloud platform
CN116683417A (en) * 2023-06-05 2023-09-01 国网浙江省电力有限公司杭州市钱塘区供电公司 Carbon flow optimization method and system for medium-low voltage flexible direct current power distribution network
CN117674375A (en) * 2023-11-15 2024-03-08 燕山大学 New energy multi-energy complementary hydrogen production system energy management method

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115238505A (en) * 2022-07-27 2022-10-25 国网山东省电力公司枣庄供电公司 Optimal configuration method and system for combined heat, power and hydrogen supply system based on wind and light station
CN115238505B (en) * 2022-07-27 2024-02-13 国网山东省电力公司枣庄供电公司 Thermoelectric hydrogen combined supply type system optimal configuration method and system based on wind-solar field station
CN116581742A (en) * 2023-05-16 2023-08-11 淮阴工学院 Chameleon algorithm-based flexible electric load scheduling system and chameleon algorithm-based flexible electric load scheduling method for regulating and controlling intelligent cloud platform
CN116581742B (en) * 2023-05-16 2024-02-23 淮阴工学院 Chameleon algorithm-based flexible electric load scheduling system and chameleon algorithm-based flexible electric load scheduling method for regulating and controlling intelligent cloud platform
CN116683417A (en) * 2023-06-05 2023-09-01 国网浙江省电力有限公司杭州市钱塘区供电公司 Carbon flow optimization method and system for medium-low voltage flexible direct current power distribution network
CN116683417B (en) * 2023-06-05 2024-01-30 国网浙江省电力有限公司杭州市钱塘区供电公司 Carbon flow optimization method and system for medium-low voltage flexible direct current power distribution network
CN117674375A (en) * 2023-11-15 2024-03-08 燕山大学 New energy multi-energy complementary hydrogen production system energy management method
CN117674375B (en) * 2023-11-15 2024-06-07 燕山大学 New energy multi-energy complementary hydrogen production system energy management method

Similar Documents

Publication Publication Date Title
CN110544935B (en) Electric-hydrogen multi-energy complementary direct-current micro-grid coordinated scheduling method
CN114362268A (en) Wind-solar-charge two-stage prediction-based optimized scheduling method for comprehensive energy power supply system
Zoulias et al. Techno-economic analysis of the integration of hydrogen energy technologies in renewable energy-based stand-alone power systems
Darras et al. Sizing of photovoltaic system coupled with hydrogen/oxygen storage based on the ORIENTE model
Darras et al. PV output power fluctuations smoothing: The MYRTE platform experience
CN113850474B (en) Thermoelectric hydrogen multi-energy flow comprehensive energy system and optimal scheduling method thereof
CN112086998B (en) Energy optimization scheduling method for direct-current micro-grid in island mode
CN112886645B (en) New energy power system operation simulation method based on ultrahigh hydrogen energy proportion
CN112803567B (en) Parameter optimization design method and system based on intelligent building optical storage power supply equipment
An et al. Coordinative optimization of hydro-photovoltaic-wind-battery complementary power stations
Huangfu et al. An optimal energy management strategy with subsection bi-objective optimization dynamic programming for photovoltaic/battery/hydrogen hybrid energy system
Belkhier et al. Novel design and adaptive coordinated energy management of hybrid fuel‐cells/tidal/wind/PV array energy systems with battery storage for microgrids
Wang et al. Research on capacity optimization and real-time control of island microgrid considering time-shifting load
Dong et al. Sizing of a stand-alone photovoltaic/wind energy system with hydrogen and battery storage based on improved ant colony algorithm
Ayat et al. Energy management based on a fuzzy controller of a photovoltaic/fuel cell/Li-ion battery/supercapacitor for unpredictable, fluctuating, high-dynamic three-phase AC load
CN116247745A (en) Hydrogen energy micro-grid rolling optimization scheduling method considering unplanned off-grid risk
CN116231694A (en) Comprehensive energy system with wind-solar complementary hydrogen production and cold-hot electric multifunctional complementation
Zidane et al. A new fuzzy logic solution for energy management of hybrid photovoltaic/battery/hydrogen system
Ran et al. Economic dispatch of off-grid photovoltaic generation system with hybrid energy storage
CN116231690A (en) Hydrogen production system and method by utilizing waste wind and waste light
Kord et al. An integrated hybrid power supply for off-grid applications fed by wind/photovoltaic/fuel cell energy systems
Zhang et al. Model Predictive Control Method of Multi-Energy Flow System Considering Wind Power Consumption
Rao et al. Implementation of minigrid with hybrid renewable energy sources for urban community buildings
Zhao et al. Energy Optimization Model of Multi Energy Interaction in Thermal Power Plants with Wind Power, Photovoltaic, Hydrogen Production and Hydrogen Fuel Cell System
CN116484553B (en) Multi-energy complementary-considered micro-grid optimization design method

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