CN117477609A - Energy management system and method of hydrogen electric coupling system for park - Google Patents
Energy management system and method of hydrogen electric coupling system for park Download PDFInfo
- Publication number
- CN117477609A CN117477609A CN202311421071.4A CN202311421071A CN117477609A CN 117477609 A CN117477609 A CN 117477609A CN 202311421071 A CN202311421071 A CN 202311421071A CN 117477609 A CN117477609 A CN 117477609A
- Authority
- CN
- China
- Prior art keywords
- energy management
- power generation
- prediction
- hydrogen
- energy
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- UFHFLCQGNIYNRP-UHFFFAOYSA-N Hydrogen Chemical compound [H][H] UFHFLCQGNIYNRP-UHFFFAOYSA-N 0.000 title claims abstract description 68
- 229910052739 hydrogen Inorganic materials 0.000 title claims abstract description 68
- 239000001257 hydrogen Substances 0.000 title claims abstract description 68
- 238000000034 method Methods 0.000 title claims abstract description 49
- 230000008878 coupling Effects 0.000 title claims abstract description 26
- 238000010168 coupling process Methods 0.000 title claims abstract description 26
- 238000005859 coupling reaction Methods 0.000 title claims abstract description 26
- 238000010248 power generation Methods 0.000 claims abstract description 80
- 238000004146 energy storage Methods 0.000 claims abstract description 60
- 238000005457 optimization Methods 0.000 claims abstract description 57
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 claims abstract description 40
- 229910052744 lithium Inorganic materials 0.000 claims abstract description 40
- 238000012544 monitoring process Methods 0.000 claims abstract description 16
- 238000005096 rolling process Methods 0.000 claims abstract description 15
- 238000004519 manufacturing process Methods 0.000 claims abstract description 11
- 238000009499 grossing Methods 0.000 claims abstract description 10
- 238000003860 storage Methods 0.000 claims abstract description 10
- 238000006243 chemical reaction Methods 0.000 claims abstract description 7
- 238000007726 management method Methods 0.000 claims description 151
- 238000004422 calculation algorithm Methods 0.000 claims description 31
- 230000006870 function Effects 0.000 claims description 31
- 241000282461 Canis lupus Species 0.000 claims description 24
- HEZMWWAKWCSUCB-PHDIDXHHSA-N (3R,4R)-3,4-dihydroxycyclohexa-1,5-diene-1-carboxylic acid Chemical compound O[C@@H]1C=CC(C(O)=O)=C[C@H]1O HEZMWWAKWCSUCB-PHDIDXHHSA-N 0.000 claims description 16
- 230000002441 reversible effect Effects 0.000 claims description 16
- 238000012545 processing Methods 0.000 claims description 14
- 238000013507 mapping Methods 0.000 claims description 13
- 230000003993 interaction Effects 0.000 claims description 10
- 238000000354 decomposition reaction Methods 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 8
- 230000002159 abnormal effect Effects 0.000 claims description 7
- 230000000739 chaotic effect Effects 0.000 claims description 7
- 238000012937 correction Methods 0.000 claims description 7
- 238000009826 distribution Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000011049 filling Methods 0.000 claims description 6
- 230000000694 effects Effects 0.000 claims description 5
- 230000005611 electricity Effects 0.000 claims description 5
- 238000012417 linear regression Methods 0.000 claims description 5
- 238000013527 convolutional neural network Methods 0.000 claims description 4
- 230000007246 mechanism Effects 0.000 claims description 4
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 150000001875 compounds Chemical class 0.000 claims description 3
- 238000004134 energy conservation Methods 0.000 claims description 3
- 230000002195 synergetic effect Effects 0.000 claims description 3
- 241000282421 Canidae Species 0.000 claims description 2
- 238000004891 communication Methods 0.000 description 19
- 238000010586 diagram Methods 0.000 description 10
- 101001095088 Homo sapiens Melanoma antigen preferentially expressed in tumors Proteins 0.000 description 5
- 102100037020 Melanoma antigen preferentially expressed in tumors Human genes 0.000 description 5
- 238000004590 computer program Methods 0.000 description 5
- 239000000446 fuel Substances 0.000 description 5
- 230000002457 bidirectional effect Effects 0.000 description 4
- 238000003745 diagnosis Methods 0.000 description 4
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 4
- 230000001276 controlling effect Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 241001672018 Cercomela melanura Species 0.000 description 2
- 101000595182 Homo sapiens Podocan Proteins 0.000 description 2
- 101000860430 Homo sapiens Versican core protein Proteins 0.000 description 2
- 102100036036 Podocan Human genes 0.000 description 2
- 102100028437 Versican core protein Human genes 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- 238000007792 addition Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001035 drying Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000003912 environmental pollution Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000004806 packaging method and process Methods 0.000 description 1
- 230000002688 persistence Effects 0.000 description 1
- 238000002203 pretreatment Methods 0.000 description 1
- 238000000746 purification Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/28—Arrangements for balancing of the load in a network by storage of energy
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J13/00—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network
- H02J13/00006—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network characterised by information or instructions transport means between the monitoring, controlling or managing units and monitored, controlled or operated power network element or electrical equipment
- H02J13/00016—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network characterised by information or instructions transport means between the monitoring, controlling or managing units and monitored, controlled or operated power network element or electrical equipment using a wired telecommunication network or a data transmission bus
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/003—Load forecast, e.g. methods or systems for forecasting future load demand
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/004—Generation forecast, e.g. methods or systems for forecasting future energy generation
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/007—Arrangements for selectively connecting the load or loads to one or several among a plurality of power lines or power sources
- H02J3/0075—Arrangements for selectively connecting the load or loads to one or several among a plurality of power lines or power sources for providing alternative feeding paths between load and source according to economic or energy efficiency considerations, e.g. economic dispatch
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/24—Arrangements for preventing or reducing oscillations of power in networks
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/28—Arrangements for balancing of the load in a network by storage of energy
- H02J3/32—Arrangements for balancing of the load in a network by storage of energy using batteries with converting means
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
- H02J3/46—Controlling of the sharing of output between the generators, converters, or transformers
- H02J3/466—Scheduling the operation of the generators, e.g. connecting or disconnecting generators to meet a given demand
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/10—Power transmission or distribution systems management focussing at grid-level, e.g. load flow analysis, node profile computation, meshed network optimisation, active network management or spinning reserve management
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2300/00—Systems for supplying or distributing electric power characterised by decentralized, dispersed, or local generation
- H02J2300/20—The dispersed energy generation being of renewable origin
- H02J2300/22—The renewable source being solar energy
- H02J2300/24—The renewable source being solar energy of photovoltaic origin
- H02J2300/26—The renewable source being solar energy of photovoltaic origin involving maximum power point tracking control for photovoltaic sources
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Computer Networks & Wireless Communication (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention provides an energy management system and method of a hydrogen electric coupling system for a park, wherein the system architecture comprises the following components: a photovoltaic power generation system for providing clean power; the lithium battery energy storage system is used for smoothing power fluctuation of the photovoltaic power generation system and the load end; a hydrogen storage system for hydrogen-to-electricity conversion and smoothing power fluctuations; the energy management system EMS comprises an SCADA monitoring module and an energy management module, and can realize energy optimization scheduling of multiple time scales through global optimization before the day and rolling optimization in the day; a central control system. When the photovoltaic power generation exceeds the load end demand, the lithium battery is charged, hydrogen production and energy storage are carried out, surplus power is fully charged with the lithium battery and the hydrogen production and energy storage, and when the photovoltaic power generation is lower than the load end demand, the lithium battery energy storage system and the hydrogen energy storage system are subjected to discharge control through the central control system, so that power smoothing is realized. The invention can effectively improve the running reliability of the system, effectively reduce the dependence of users on the power grid and improve the utilization rate of renewable energy sources.
Description
Technical Field
The invention belongs to the technical field of new energy, and particularly relates to an energy management system and method of a hydrogen electric coupling system for a park.
Background
Due to the increasing environmental pollution, photovoltaic power generation and energy storage systems for parks are increasingly favored by the market. The photovoltaic and energy storage system can reduce dependence on the power grid to a certain extent for park users, and improve economic and environmental benefits of the users. However, photovoltaic power generation has instability and randomness due to weather factors such as solar irradiance, temperature, and the like. In addition, the introduction of different devices such as a lithium battery energy storage system and a hydrogen energy storage system also brings the problems of too many data nodes, redundant communication protocol types and the like. Therefore, reasonable energy management system architecture and methods are key to enabling stable and efficient operation of the hydrogen electrical coupling system.
Disclosure of Invention
Aiming at the actual demand of the prior art, the invention provides an energy management system and method of a hydrogen electric coupling system for a park. The method can realize the functions of real-time data acquisition, data transmission, data storage, operation monitoring, remote control, fault warning, output plan scheduling, statistical analysis, data encryption, operation log recording, user management and the like.
The data uploaded by the data acquisition terminal is subjected to research, judgment, processing and control instruction output by the energy management system, so that the safe and stable operation of the hydrogen electric coupling system is ensured.
The system architecture comprises: a photovoltaic power generation system for providing clean power; the lithium battery energy storage system is used for smoothing power fluctuation of the photovoltaic power generation system and the load end; a hydrogen storage system for hydrogen-to-electricity conversion and smoothing power fluctuations; the energy management system EMS comprises an SCADA monitoring module and an energy management module, and can realize energy optimization scheduling of multiple time scales through global optimization before the day and rolling optimization in the day; the central control system is used for controlling and monitoring the photovoltaic power generation system, the lithium battery energy storage system, the hydrogen energy storage system, the energy management system and other electrical equipment. When the photovoltaic power generation exceeds the load end demand, the lithium battery is charged, hydrogen production and energy storage are carried out, surplus power is fully charged with the lithium battery and the hydrogen production and energy storage, and when the photovoltaic power generation is lower than the load end demand, the lithium battery energy storage system and the hydrogen energy storage system are subjected to discharge control through the central control system, so that power smoothing is realized. The invention can effectively improve the running reliability of the system, effectively reduce the dependence of users on the power grid and improve the utilization rate of renewable energy sources.
The technical scheme adopted for solving the technical problems is as follows:
an energy management system for a hydrogen electrical coupling system for a campus, characterized by: comprising the following steps: a photovoltaic power generation system for providing clean power; the lithium battery energy storage system is used for smoothing power fluctuation of the photovoltaic power generation system and the load end; a hydrogen storage system for hydrogen-to-electricity conversion and smoothing power fluctuations; the energy management system EMS comprises an SCADA monitoring module and an energy management module, and realizes energy optimization scheduling of multiple time scales through global optimization before the day and rolling optimization in the day; the central control system is used for controlling and monitoring the photovoltaic power generation system, the lithium battery energy storage system, the hydrogen energy storage system, the energy management system and other electrical equipment; when the photovoltaic power generation exceeds the load end demand, the lithium battery is charged, hydrogen production and energy storage are carried out, surplus power is fully charged, the network is connected, and when the photovoltaic power generation is lower than the load end demand, the discharge control is carried out on the lithium battery energy storage system and the hydrogen energy storage system through the central control system, so that the power smoothness is realized.
Further, the energy management system participates in the information interaction of the economic operation scheduling of the power grid and the scheduling layer of the power distribution network, realizes the high-efficiency and low-cost operation of the whole system through a user-defined management strategy and an optimization algorithm, and responds to the user demands; according to the load and the power generation condition, the power generation of the system is realized to automatically track the power consumption requirement of a user; when the load changes randomly, the power system frequency is kept stable through PQ control or VF control, and renewable energy sources are used to the maximum extent.
Further, the energy management module receives a power grid dispatching instruction or provides data information for energy management and dispatching decision through photovoltaic power generation prediction and load prediction; two strategies of active management and passive management are provided:
the method comprises the steps of actively managing time resolution to be in a minute or hour level, providing data information for energy management and scheduling decision through photovoltaic power generation prediction and load prediction, establishing a model taking system economy and durability as optimization targets according to prediction data, time-of-use electricity price and equipment state information, and solving by using an improved gray wolf optimization algorithm IGWO under the condition that system electric energy requirements and all equipment constraint conditions are met so as to formulate a multi-time scale equipment output plan;
The passive management is in millisecond level or second level, and the working state of the equipment is adaptively adjusted according to the bus and equipment information through the synergistic effect of the energy management system and the central control system, so that the high-frequency on-off and mode switching of DCDC and DCAC are realized.
The energy management system has an energy management function and is divided into a day-ahead energy management optimization stage and a day-in energy management optimization stage; the energy management system reads the SQL Server database, predicts loads and photovoltaic power generation power before and during the day, and makes an energy management strategy and a rolling output plan according to the load prediction and photovoltaic power generation prediction results and a user-defined objective function, and refers to an equipment address table and issues a power request.
Based on the above system architecture design, it also provides:
an energy management module photovoltaic power generation prediction method comprises the following steps:
step S1: an initial meteorological and photovoltaic power generation sample data set with the time resolution of 5min is adopted and preprocessed;
step S2: the correlation degree of the meteorological parameters and the photovoltaic power generation power is quantized by adopting an MIC correlation coefficient method, and characteristic parameters are determined;
step S3: determining a proper cluster number through a fuzzy C-means algorithm;
step S4: and (5) establishing a CBA mixed depth model to predict the photovoltaic power generation power of one day in the future.
Further, in step S1, abnormal data is processed by adopting a k-nearest neighbor method correction and similar daily data filling method, and the data after data processing is normalized by a maximum and minimum normalization method;
in step S2, firstly, gridding processing is carried out and an information value IX is calculated; y, then for the information value I [ X ]; carrying out normalization treatment, and finally taking the normalized maximum mutual information value after grid removal as an MIC coefficient;
in step S3, initializing membership value to be a value between 0 and 1 after determining the objective function of FCM, calculating centroid of cluster, and calculating cluster Z i Centroid z of (2) i Updating the value of the membership function:
in step S4, the time sequence information is processed in the forward direction and the reverse direction, and the calculation result of the forward LSTM and the reverse output result of the reverse LSTM are overlapped according to a certain weight to obtain an output.
An energy management module load prediction method comprising the steps of:
step S1: an initial load sample data set with the time resolution of 1h is adopted and preprocessed;
step S2: decomposing the sample data into multi-frequency IMF components by an improved decomposition algorithm, namely a fully adaptive noise set empirical mode decomposition algorithm;
step S3: predicting a low-frequency component and a high-frequency component by adopting a Multiple Linear Regression (MLR) method and an improved time sequence convolutional neural network (TCN) method respectively;
Step S4: and superposing and restoring iteration prediction results of the MLR and the TCN to obtain the base load data.
Further, in step S1, abnormal data is processed by adopting a k-nearest neighbor method correction and similar daily data filling method, and the data after data processing is normalized by a maximum and minimum normalization method;
in the step S2, dividing the sequence signal into a high-frequency component and a low-frequency component according to the zero-crossing rate of the sequence signal, setting the zero-crossing rate to be 0.01 and classifying the zero-crossing rate to be the low-frequency component smaller than 0.01;
in the step S3, aiming at the low-frequency IMF with strong periodicity and weak fluctuation, the MLR is adopted for accurate prediction; the TCN is adopted to effectively train and predict high-frequency IMF components with poor stability and strong fluctuation, and IGWO (insulated gate bipolar transistor) for improving Tent chaos and elite reverse learning is adopted to optimize TCN hyper-parameters so as to improve prediction performance; in order to solve the problem that the numerical distribution generated by the Tent mapping lacks ergodic performance, the Tent mapping is improved:
wherein k represents the mapping times, Z k Representing the value to which the mapping corresponds.
After the numerical value of the chaotic map is generated, mapping and converting the chaotic variable and the variable participating in optimization:
x ik ＝l k +(u k -l k )×z k ,i∈[1,N],k∈[1,D]
wherein x is ik A value representing the kth dimension of the generated ith wolf, N representing the number of wolves, D representing the dimension of the problem, l k And u k Representing the upper and lower limits of the variable to be optimized.
Introducing an elite reverse learning mechanism on the basis of improving an initial sample generated by the Tent chaos;
in step S4, the iterative prediction results of the MLR and TCN are superimposed and output.
An energy management method based on load prediction and photovoltaic power generation prediction, comprising the following steps:
step S1: in the day-ahead energy management optimization stage, a power plan of a future 24h scheduling period is formulated by taking 1h as time resolution;
step S2: in an energy management optimizing stage in the day, rolling and optimizing a power plan of a future 3h scheduling period by taking 15min as time resolution;
step S3: respectively solving a day-ahead energy management model and a day-in energy management model by adopting an improved gray wolf optimization algorithm to obtain an optimal output plan;
step S4: and rolling and optimizing the daily output plan.
Further, in step S1, according to the load prediction and the photovoltaic power generation prediction results, the total running cost of the system is minimized, the self-balancing rate is maximized as an optimization target, and the energy conservation and the equipment output conditions are used as constraints to make a power output plan of a future 24h scheduling period;
in step S2, the difference value between the day-ahead output plan and the ultra-short-term prediction information is balanced by taking the minimum adjustment day-ahead plan as an optimization target and adjusting the charge and discharge of the lithium battery energy storage system, the charge and discharge of the hydrogen energy storage system and the power grid interaction power so as to reduce the influence of the prediction error on the optimization effect;
In step S3, hunting search strategy optimization based on dimension learning is introduced to improve a gray wolf optimization algorithm, and a day-ahead and day-in energy management model is solved;
R i (t)＝||X i (t)-X i_GWO (t+1)||
N i (t)＝{X j (t)|D i (X i (t),X j (t))≤R i (t),X j (t)∈pop}
X i_DLH,d (t+1)＝X i,d (t)+rand*(X n,d (t)-X r,d (t))
wherein X is i (t) represents a wolf group position; x is X i_GWO Representing the position of the t+1st iteration of the ith wolf obtained by updating the conventional GWO; r is R i (t) represents X i (t) and X i_GWO Euclidean distance between; n (N) i (t) represents that the compound satisfies R i (t) X in the radius range i A neighborhood of (t); d (D) i X represents i (t) and X j Euclidean distance of (t); x is X n,d (t) represents d-dimensional data of a wolf randomly selected from within the neighborhood, X r,d (t) represents random selection from within the populationD-dimensional data of a selected wolf, X i_DLH,d (t+1) represents the result of multi-neighborhood learning.
Compared with the prior art, the beneficial effects and main innovation points of the invention and the preferred scheme thereof are at least embodied in the following aspects:
1. the energy management system and the method for the hydrogen-electricity coupling system for the park can effectively improve the utilization rate of photovoltaic power generation, reduce the dependence degree of a user on a power grid, effectively provide self-balancing capability of the system and save cost.
2. The energy management model based on photovoltaic power generation and load prediction is provided, rolling correction is carried out on a 24-hour day-ahead plan according to a 3-hour day-ahead prediction result, and the influence caused by uncertainty of prediction information is reduced.
Drawings
The invention is described in further detail below with reference to the attached drawings and detailed description:
FIG. 1 is a schematic diagram of a hydrogen electric coupling system according to an embodiment of the present invention;
FIG. 2 is a functional system architecture diagram of an energy management system according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a load prediction flow of an energy management system according to an embodiment of the present invention;
FIG. 4 is a graph of load prediction results for an energy management system according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a photovoltaic power generation prediction flow of an energy management system according to an embodiment of the present invention;
FIG. 6 is a graph of the predicted outcome of photovoltaic power generation by the energy management system according to an embodiment of the present invention; wherein: (a) a sunny day prediction result; (b) overcast day prediction results; (c) rainy day prediction results.
Detailed Description
In order to make the features and advantages of the present patent more comprehensible, embodiments accompanied with figures are described in detail below:
it should be noted that the following detailed description is illustrative and is intended to provide further explanation of the present application. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
The following describes the embodiments of the present invention basically:
As shown in FIG. 1, the energy management system (1) is suitable for photovoltaic capacity +.6MWp, energy storage capacity: a small and medium scale mixed energy system of less than or equal to 5MWp in a park, industry, commerce and the like. The hydrogen electric coupling system comprises a photovoltaic power generation system (8), a lithium battery energy storage system, a hydrogen energy storage system, an energy management system (1), a central control system (2), DCDC (9-12), DCAC (14) and other various software and hardware necessary for completing the stable operation of the hydrogen electric coupling system.
And the photovoltaic power generation system (8) is used for converting solar energy into direct current and outputting the direct current. The photovoltaic module, the lightning protection junction box and the photovoltaic power generation system are made of materials necessary for operation, including but not limited to connectors, wire pipes PVC, cables and supports.
A lithium battery energy storage system. The system comprises a battery management system (3), a lithium battery pack (4) and software and hardware necessary for the operation of the lithium battery energy storage system. The lithium battery pack (4) is used for consuming surplus power of the photovoltaic power generation system and releasing power when the load exceeds the required time, and meanwhile, the voltage stability of the direct-current bus is maintained. The battery management system (3) is used for controlling and monitoring the lithium battery pack (4), and the embedded communication board card is used for packaging the acquired data flow and interacting data and control instructions with the energy management system (1).
The hydrogen energy storage system comprises a hydrogen production tank subsystem (5), a hydrogen storage tank (6), a fuel cell subsystem (7), a gas-liquid separator, a purifying device, a hydrogen channel valve, a flowmeter and the like, and software and hardware necessary for the operation of the hydrogen energy storage system. The electrolytic water hydrogen production tank subsystem (5) is used for absorbing surplus power after the lithium battery energy storage system stores energy, and the surplus power is used for the electrolytic water and is pressurized and stored in the hydrogen storage tank (6) after the processes of oxyhydrogen separation, drying, purification and the like. The hydrogen storage tank (6) supplies hydrogen to the fuel cell subsystem (7) when the load is exceeded, and the fuel cell subsystem (7) converts the hydrogen into electricity to meet the load demand.
As shown in fig. 2, the energy management system (1) realizes global optimization before the day and rolling optimization in the day according to the collected real-time data and state information, realizes the energy optimization scheduling function of multiple time scales, and simultaneously has the functions of real-time display of an HMI interface, fault alarm, data report, authority management and the like.
And the central control system (2) coordinates and controls the controllable distributed power supply and the controllable load according to a control strategy to realize the effective control of the source network charge storage.
The DCDC is mainly used for realizing the functions of energy flow and voltage rising/dropping, and can respectively act on a high-voltage side and a low-voltage side, and each side is provided with 5 working modes of constant current input/output, constant voltage output, MPPT input and a bus voltage-controlled current source. The DCDC (9) is a bidirectional power supply and is used for realizing bidirectional flow of energy and charge and discharge of the lithium battery pack (4).
The bidirectional grid-connected and off-grid inverter DCAC (14) is used for realizing energy bidirectional flow, AC-DC conversion and grid-connected/off-grid switching, and provides stable DC bus voltage during grid connection.
The battery management system BMS (3) adopts a master-slave distributed system architecture, comprises an upper-layer main control module BCU, a slave control module BMU and communication equipment RMU, and is communicated with each other through a CAN bus. The main control module BCU is connected with the energy management system (1) through the VCAN, is connected with the communication module RMU through the DCAN, so that interaction with the cloud server (18) is further realized, and is connected with the slave control module BMU through the PCAN. The battery management system (3) can realize functions of battery pack data acquisition, fault diagnosis, state estimation, active equalization and the like;
the energy management system (1) comprises a data acquisition and monitoring control system SCADA and an energy management module, and consists of equipment such as an industrial personal computer/server, a switch, a communication manager, a data acquisition terminal and the like. The field bus is used for communicating with each controllable device, and device state information is collected to determine an energy management strategy, so that functions of data acquisition, optimized operation, man-machine interaction, communication processing, power distribution, protection control and the like are realized. The host power supply is powered by an independent uninterruptible power supply UPS to realize a black start function.
The SCADA system comprises the functions of real-time data monitoring, fault alarm processing, equipment state monitoring, long-time data storage, user authority management, operation report generation, parameter setting and the like. All readable data of the bottom layer equipment are finally uploaded to the cloud server through various protocols, and part of the readable data are distributed to subcontrollers (such as a BCU of a battery management system (3), a FCU of a fuel cell and a DSP control board of DCDC) to carry out protocol analysis and edge calculation, so that the calculation pressure of scheduling layer equipment such as a central control system (2), an energy management system (1) and the like is relieved. And simultaneously, a north/south interface is provided for data collection, and a south interface provides SDK and data uploading protocol explanation, so that the system can be used for third-party gateway and equipment access.
The energy management system establishes communication with the equipment layer through the communication network shutdown by using communication protocols such as Modbus RTU, CAN, OPC, MQTT and the like, and is used for receiving the equipment layer uploading information and issuing a scheduling instruction of the energy management system. The energy management system uploads the device layer data to the local and cloud databases respectively through the Ethernet.
The energy management module is provided with two strategies, namely active management and passive management. The active management time resolution is in the order of minutes or hours, data information is provided for energy management and scheduling decision through photovoltaic power generation prediction and load prediction, a model taking system economy and durability as optimization targets is established according to the information such as prediction data, time-of-use electricity price, equipment state and the like, and under the condition that the system electric energy requirement and each equipment constraint condition are met, an improved gray wolf optimization algorithm IGWO is used for solving, so that a multi-time-scale equipment output plan is formulated. The passive management is in millisecond or second level, and the working state of the equipment is adaptively adjusted according to the bus and equipment information through the synergistic effect of the energy management system (1) and the central control system (2), so that the high-frequency on-off and mode switching of DCDC (9-12) and DCAC (14) can be realized.
The photovoltaic prediction adopts a CNN-BiLSTM-Attention (CBA) mixed depth model, an initial sample data set with 5min intervals is adopted and preprocessed, information is quantized through a MIC correlation coefficient method to select feature vectors, similar daily cluster analysis is carried out on samples aiming at three days of sunny days, cloudy days and rainy days, model training is carried out on the samples respectively, a prediction result is output, and a multi-time-scale equipment output plan is formulated.
The central control system (2) is connected with the energy management system through a TCP/IP protocol, is connected with the fuel cell subsystem (7) through a CAN protocol, is connected with the electrolytic water hydrogen production tank subsystem (5) through a Modbus TCP protocol, and adopts a Modbus RTU protocol for DCDC (9-12), DCAC (14) and intelligent ammeter (13/15), and other supporting auxiliary equipment adopts Modbus, OPC and other protocols.
The energy management system supports rich communication protocols and protocol conversion, and supports international standard communication protocols and power standards such as Modbus TCP, modbus RTU, MQTT, IEC60870-5-101, IEC60870-5-103, IEC60870-5-104, DL/T645-97, DL/T645-07 and the like. And the system supports cross-platform deployment and can stably run in a Windows, linux, UNIX, macOS system. The database adopts a local database and cloud database scheme, adopts a standard data interface, supports different data persistence schemes, can seamlessly migrate different mainstream relational databases (SQL Server, mySQL, oracle) and supports NoSQL databases (MongoDB, HBASE and the like).
The energy supply method comprises the steps that an energy management system EMS (1) is matched with a central control system (2) to monitor, display and analyze real-time data in the hydrogen electric coupling system, and the working modes of the lithium battery energy storage system and the hydrogen energy storage system are regulated to realize the real-time balance of energy in the system. When the photovoltaic power generation power is higher than the power load, the lithium battery energy storage system and the hydrogen energy storage system store energy sequentially, and the network is accessed if surplus power exists. When the photovoltaic power generation power is lower than the power consumption power, the central control system sends a power request to the lithium battery energy storage system and the hydrogen energy storage system, the lithium battery energy storage system and the hydrogen energy storage system are discharged successively, and the shortage power is supplemented by the power grid.
The system and method of embodiments of the present invention are further described below:
the system mainly comprises the following components: the system comprises an energy management system (1), a central control system (2), a photovoltaic power generation system (8), a lithium battery energy storage system, a hydrogen energy storage system, DCDC (9-12), DCAC (14) and other various software and hardware necessary for completing the stable operation of a hydrogen electric coupling system.
The energy management system (1) can provide effective support for the power grid, participate in the information interaction of the economic operation scheduling of the power grid and the scheduling layer of the power distribution network, realize the high-efficiency and low-cost operation of the whole system through a user-defined management strategy and an optimization algorithm, and can respond to the demands of users. According to the load and the power generation condition, the power generation system can automatically track the power consumption requirement of a user. When the load changes randomly, the power system frequency is kept stable through PQ control or VF control, and renewable energy sources are used to the maximum extent. The system comprises a SCADA and an energy management module.
Furthermore, the SCADA can be connected with various equipment in the hydrogen-electricity coupling systems such as DCDC (9-12), DCAC (14), intelligent ammeter (13/15), energy equipment, communication manager and the like, the telemetry and telemetry of measured data, equipment running state and the like can be sent to the communication manager, the communication manager uses a corresponding communication protocol to analyze, and all data are packed into a standard TCP/IP protocol according to a set sequence and sent to the SCADA. The system realizes the localized real-time monitoring, operation control and high-efficiency operation and maintenance of resources, monitors key parameters such as power flow, voltage, current, frequency and the like in real time, and realizes the safe and stable operation of the system through various protection means.
Further, the energy management module can receive a power grid dispatching instruction or provide data information for energy management and dispatching decision through photovoltaic power generation prediction and load prediction, an objective function considering optimization indexes such as economy, durability and the like of the system is constructed according to information such as load demand, power generation power, time-of-use electricity price and the like, the objective function is solved by using an improved gray wolf optimization algorithm under the condition that the system electric energy demand and each equipment constraint condition are met, and a multi-time-scale equipment output plan is formulated.
Preferably, the energy management system (1) can be selected from equipment such as an industrial PC or an industrial server, and the equipment comprises a movable front cover plate, a BIOS battery, a CFast-Card, a hard disk, an extended hard disk, power wiring, a plurality of network cards, a plurality of USB interfaces, a plurality of display interfaces and a plurality of serial interfaces.
The central control system (2) is a high-performance controller which is locally installed, is in communication connection with DCDC (9-12), an inverter (14), a lithium battery energy storage system, a hydrogen energy storage system, a power grid side ammeter (13), a load side ammeter (15), controllable loads and the like through a protocol conversion module, and is responsible for managing bottom equipment and access points, monitoring the running state of equipment, rapidly generating control instructions and issuing and executing. Through interaction cooperation with the energy management system (1), the functions of system power flow control, voltage frequency stable control, off-grid mode switching, power load cut-out, system black start and the like are realized, and safe, stable and efficient operation of the hydrogen-electricity coupling system is ensured.
Preferably, the central control system (2) is selected from an industrial PC or an industrial server, or is deployed in the same embedded controller with the energy management system (1), and comprises a plurality of serial interfaces, a display interface, a CFast-Card, a MicroSD Card, an RJ45 Ethernet Card (X000 and X001), a battery, a CPU diagnosis indicator, a USB interface, a power diagnosis indicator, a 24V controller power interface, a 24V bus power interface (same as 10), an I/O channel 24V power interface, an I/O channel 0V power interface, a disconnection guide rail connection brace and a PE interface
DCDC can set a variety of operating modes and can be switched between the various operating modes quickly and frequently. The working modes of the high-voltage and low-voltage power supply can be constant-current input, constant-current output, constant-voltage output, MPPT input and bus voltage-controlled current sources, and each working mode can respectively act on a high-voltage side or a low-voltage side, and 10 different combinations are provided.
The energy management system has load prediction and photovoltaic power generation prediction functions. The energy management system receives the 'three-remote' information uploaded by the equipment through TCP/IP and stores the information in the SQL server database. The energy management system configures an ODBC data source in the Windows operating system, so that predictive control software developed based on MATLAB can be connected with an SQL Server database to realize a predictive function, and corresponding equipment addresses and power request signals are determined through a Modbus protocol point table according to a user-defined algorithm and a target. The user can also access the database through the HMI interface and the workstation PC.
Further, the energy management module photovoltaic power generation prediction method comprises the following steps:
step S1: and (5) adopting an initial meteorological and photovoltaic power generation sample data set with the time resolution of 5min and preprocessing.
Step S2: and quantifying the association degree of the meteorological parameters and the photovoltaic power generation power by adopting an MIC correlation coefficient method and determining characteristic parameters.
Step S3: and determining the proper clustering number through a fuzzy C-means algorithm.
Step S4: and (5) establishing a CBA mixed depth model to predict the photovoltaic power generation power of one day in the future.
In step S1, abnormal data is processed by adopting a k-nearest neighbor method correction and similar daily data filling method, and the data after data processing is normalized by a maximum and minimum normalization method.
Further, in step S2, firstly, a gridding process is performed and an information value ix is calculated; y, then for the information value I [ X ]; and Y is normalized, and finally, the normalized maximum mutual information value after grid removal is used as an MIC coefficient.
Further, in step S3, after determining the objective function of the FCM, initializing the membership value to a value between 0 and 1, calculating the centroid of the cluster, and calculating the cluster Z i Centroid z of (2) i Updating the value of the membership function:
further, in step S4, the timing information is processed in both the forward direction and the reverse direction, and the calculation result of the forward LSTM and the reverse output result of the reverse LSTM are overlapped according to a certain weight to obtain an output.
h t,z ＝LSTM(h t-1 ,x t )
h i,f ＝LSTM(h i-1 ,x t )
y t ＝a t,b *h t,z +b t,b *h i,f +c t,b
Further, the energy management module load prediction method comprises the following steps:
step S1: an initial load sample dataset with a time resolution of 1h is used and pre-processed.
Step S2: the decomposition into multi-frequency IMF components is performed by a modified decomposition algorithm, a fully adaptive noise set empirical mode decomposition (CEEMDAN) algorithm.
Step S3: the low frequency and high frequency components are predicted using a multiple linear regression method (multiple linearregression, MLR) and a modified time series convolutional neural network (Temporal Convolutional Network, TCN), respectively.
Step S4: and superposing and restoring iteration prediction results of the MLR and the TCN to obtain the base load data.
Further, in step S1, the same pretreatment method as that of photovoltaic power generation prediction is adopted.
Further, in step S2, the sequence signal is divided into a high frequency component and a low frequency component according to the size of the zero crossing rate, the zero crossing rate is set to be 0.01, and the zero crossing rate is classified as the low frequency component which is smaller than 0.01. Wherein the calculation formula of the zero-crossing rate is as follows:
further, in step S3, for the low-frequency IMF with strong periodicity and weak fluctuation, the MLR is used for accurate prediction. The TCN is adopted to effectively train and predict high-frequency IMF components with poor stability and strong fluctuation, and the IGWO for improving Tent chaos and elite reverse learning is adopted to optimize TCN hyper-parameters so as to improve the prediction performance. In order to solve the problem that the numerical distribution generated by the Tent mapping lacks ergodic performance, the Tent mapping is improved:
After the numerical value of the chaotic map is generated, mapping and converting the chaotic variable and the variable participating in optimization:
x ik ＝l k +(u k -l k )×z k ,i∈[1,N],k∈[1,D]
an elite reverse learning mechanism is introduced on the basis of improving an initial sample of Tent chaotic generation.
Further, in step S4, the iterative prediction results of the MLR and TCN are superimposed and output.
The energy management system has an energy management function and can be divided into a day-ahead energy management optimization stage and a day-in energy management optimization stage. The energy management system can read the SQL Server database, forecast load and photovoltaic power generation power before and during the day, and formulate an energy management strategy and a rolling output plan according to the load forecast and photovoltaic power generation forecast result and a user-defined objective function, and consult an equipment address table and issue a power request.
Further, the energy management method based on load prediction and photovoltaic power generation prediction comprises the following steps:
step S1: in the day-ahead energy management optimization stage, a power plan of a future 24h scheduling period is formulated with 1h as time resolution.
Step S2: in the day-ahead energy management optimization stage, the power output plan of a future 3h scheduling period is optimized in a rolling mode with 15min as time resolution.
Step S3: and respectively solving the day-ahead and day-in energy management models by adopting a further improved gray wolf optimization algorithm, and obtaining an optimal output plan.
Step S4: and rolling and optimizing the daily output plan.
Further, in step S1, according to the load prediction and photovoltaic power generation prediction results, a power output plan of a future 24h scheduling period is formulated with minimum total running cost of the system, maximum self-balancing rate and the like as optimization targets and constraints of energy conservation, equipment output conditions and the like
Further, in step S2, the difference between the day-ahead output plan and the ultra-short-term prediction information is balanced by adjusting the charge and discharge of the lithium battery energy storage system, the charge and discharge of the hydrogen energy storage system and the power grid interaction power by taking the minimum adjustment day-ahead plan as an optimization target, so as to reduce the influence of the prediction error on the optimization effect.
Further, in step S3, a hunting search strategy based on dimension learning is introduced to further optimize the gray wolf optimization algorithm, and the day-ahead and day-in energy management models are solved.
R i (t)＝||X i (t)-X i_GWO (t+1)||
N i (t)＝{X j (t)|D i (X i (t),X j (t))≤R i (t),X j (t)∈pop}
X i_DLH,d (t+1)＝X i,d (t)+rand*(X n,d (t)-X r,d (t))
Wherein X is i (t) represents a wolf group position; x is X i_GWO Representing the position of the t+1st iteration of the ith wolf obtained by updating the conventional GWO; r is R i (t) represents X i (t) and X i_GWO Euclidean distance between; n (N) i (t) represents that the compound satisfies R i (t) X in the radius range i A neighborhood of (t); d (D) i X represents i (t) and X j Euclidean distance of (t); x is X n,d (t) represents d-dimensional data of a wolf randomly selected from within the neighborhood, X r,d (t) represents d-dimensional data of a wolf randomly selected from the population, X i_DLH,d (t+1) represents the result of multi-neighborhood learning.
The battery management system has the functions of energy management, fault diagnosis, active equalization and the like, adopts a master-slave distributed system architecture, and consists of an upper-layer main control module BCU, a slave control module BMU, a communication module RMU and necessary software and hardware.
Further, the battery management system main control module BCU is connected with the energy management system (1) through the VCAN, is connected with the communication module RMU through the DCAN, so that interaction of the cloud server (18) is realized, and is connected with the slave control module BMU through the PCAN.
Further, the battery management system main control module BCU is used for collecting system voltage and current, estimating battery SOX, actively balancing management, diagnosing faults in real time and performing protection treatment, performing thermal management on a battery pack, communicating with the energy management system (1) and other systems through a CAN bus, and the like.
In a preferred embodiment of the present embodiment:
as shown in fig. 1, the hydrogen electric coupling system adopts an alternating current-direct current hybrid scheme, and a direct current bus is connected with a lithium battery energy storage system, a hydrogen energy storage system and a photovoltaic power generation system through each DCDC. The alternating current bus is connected with the direct current bus through the DCAC, and intelligent electric meters with communication interfaces are arranged at grid connection points and loads.
When the power of the photovoltaic power generation system (8) is higher than the load demand, surplus power enters the lithium battery energy storage system through the direct current bus, and at the moment, the DCDC (9) at the side of the lithium battery energy storage system automatically recognizes the direct current bus voltage and adjusts the working mode, so that the lithium battery energy storage system is prevented from being frequently overcharged to prolong the service life of the battery. If surplus power still exists, the surplus power enters a hydrogen energy storage system, and the electrolytic water hydrogen production tank subsystem (5) works to generate hydrogen and stores the hydrogen in the hydrogen storage tank (6). If the lithium battery and the hydrogen storage tank can not be continuously stored, surplus power is converted into single-phase 220V alternating current through DCAC and connected with the grid.
As shown in fig. 2, the energy management system has functions of data monitoring, energy management, optimal scheduling, background management, safety protection, and the like. The data monitoring sources various sensors and panel cards in the equipment, monitors the electric power and meteorological data in the hydrogen electric coupling system in real time, including parameters such as voltage, current, power, frequency, power factors, electric energy quality and the like, transmits the data to a cloud or local database for processing and analysis, and displays the data on an HMI interface in real time. On the basis of meeting the passive energy management principle of the equipment layer, the energy management introduces an active energy management method based on the prediction data and makes a scheduling plan, and details are not repeated here. Background management is used for equipment management, personnel management, authority management, secondary application of data (data report) and the like, and can help a user to perform addition, deletion and verification of equipment and personnel lists, and support data report export of different time spans of year/month/day.
As shown in fig. 3, a load data set with time resolution of 1h is taken as an initial sample, abnormal data is processed by a k-nearest neighbor method correction and similar daily data filling method, and the data is normalized; decomposing the normalized data into a high-frequency component and a low-frequency component through an improved completely self-adaptive noise set empirical mode decomposition algorithm; setting the zero-crossing rate of the sequence signal to be 0.01, and classifying the sequence signal with the zero-crossing rate smaller than 0.01 as a low-frequency component; then respectively adopting a multiple linear regression method and an improved time sequence convolution neural network to predict a low-frequency component and a high-frequency component; the IGWO algorithm for improving Tent chaos and elite reverse learning by fusion is adopted to optimize the super parameters of the time sequence convolutional neural network so as to improve the prediction performance; and finally, superposing and restoring the iteration prediction result to obtain the base load prediction data and participate in energy scheduling.
As shown in fig. 4, comparing the CMTG method with the single LSTM, TCN and BPNN results, it can be intuitively seen that CMTG is closer to the true value than other algorithms, and has a higher degree of reduction in the 0-12 period, MAPE reaches 2.601%, and other periods may deviate, but the overall error is lower. Other algorithms have different degrees of divergence in both the early and late prediction phases. Overall, the MAPE of CMTG was 2.990% and the RMSE was 600.58kW, all improved to some extent over the other single algorithm.
As shown in fig. 5, taking a meteorological and photovoltaic power generation data set with a time resolution of 1h as an initial sample, and performing abnormal data processing and normalization by adopting the same method as the preprocessing of the photovoltaic initial sample; the correlation degree of the meteorological parameters and the photovoltaic power generation power is quantized by adopting an MIC correlation coefficient method; selecting key parameters with high association degree in meteorological parameters, calculating 5 statistical indexes, namely an average value, a standard deviation, a variation coefficient, kurtosis and skewness, by taking a day as a unit and combining the key parameters as characteristic vectors for clustering, clustering similar day samples based on FCM, and determining proper clustering sample numbers according to SC and DBI indexes; training a CBA model by using a data set of a similar day sample, and determining a model structure and parameters by adopting a trial-and-error improvement method; and selecting a prediction model trained by the same type of sample according to the weather data of the prediction day, and inputting the characteristics of the weather data of the prediction day and the like to obtain the photovoltaic power generation power.
As shown in fig. 6, the photovoltaic power generation power in three weather conditions of sunny days, cloudy days and rainy days is predicted by different algorithms. It can be seen that the prediction effect of different optimization algorithms in a sunny state is easy to extract curve features with small smooth and steady fluctuation, and the high-precision prediction can be realized by different algorithms. In a cloudy day state, the CBA and CB models show good time sequence feature extraction effects, can effectively track places with large fluctuation and inflection points, and are lower in MAPE and RMSE, wherein the CBA introduced with the Attention mechanism can be predicted more accurately. Under the rainy day state, data fluctuation is big and the stationarity is poor, LSTM and BiLSTM partial period can't effectively predict, and MAPE all exceeds 10%, and CBA still can press close to photovoltaic power generation, and MAPE is 7.102%, and RMSE is 0.1129kW.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The present patent is not limited to the above-mentioned preferred embodiments, and any person who can obtain other energy management systems and methods of various types of hydrogen-electricity coupling systems for parks under the teaching of the present patent shall not be limited to the above-mentioned preferred embodiments.
Claims (10)
1. An energy management system for a hydrogen electrical coupling system for a campus, characterized by: comprising the following steps: a photovoltaic power generation system for providing clean power; the lithium battery energy storage system is used for smoothing power fluctuation of the photovoltaic power generation system and the load end; a hydrogen storage system for hydrogen-to-electricity conversion and smoothing power fluctuations; the energy management system EMS comprises an SCADA monitoring module and an energy management module, and realizes energy optimization scheduling of multiple time scales through global optimization before the day and rolling optimization in the day; the central control system is used for controlling and monitoring the photovoltaic power generation system, the lithium battery energy storage system, the hydrogen energy storage system, the energy management system and other electrical equipment; when the photovoltaic power generation exceeds the load end demand, the lithium battery is charged, hydrogen production and energy storage are carried out, surplus power is fully charged, the network is connected, and when the photovoltaic power generation is lower than the load end demand, the discharge control is carried out on the lithium battery energy storage system and the hydrogen energy storage system through the central control system, so that the power smoothness is realized.
2. The energy management system for a campus hydrogen electrical coupling system of claim 1 wherein: the energy management system participates in the information interaction of the economic operation scheduling of the power grid and the scheduling layer of the power distribution network, realizes the high-efficiency and low-cost operation of the whole system through a user-defined management strategy and an optimization algorithm, and responds to the demands of users; according to the load and the power generation condition, the power generation of the system is realized to automatically track the power consumption requirement of a user; when the load changes randomly, the voltage and the frequency of the power system are kept stable through PQ control or VF control, and renewable energy sources are used to the maximum extent.
3. The energy management system for a campus hydrogen electrical coupling system of claim 1 wherein:
the energy management module receives a power grid dispatching instruction or provides data information for energy management and dispatching decision through photovoltaic power generation prediction and load prediction; two strategies of active management and passive management are provided:
the method comprises the steps of actively managing time resolution to be in a minute or hour level, providing data information for energy management and scheduling decision through photovoltaic power generation prediction and load prediction, establishing a model taking system economy and durability as optimization targets according to prediction data, time-of-use electricity price and equipment state information, and solving by using an improved gray wolf optimization algorithm IGWO under the condition that system electric energy requirements and all equipment constraint conditions are met so as to formulate a multi-time scale equipment output plan;
The passive management is in millisecond level or second level, the working state of the equipment is adaptively adjusted according to the bus and equipment information through the synergistic effect of the energy management system and the central control system, so that the high-frequency on-off and mode switching of DCDC and DCAC are realized, the power fluctuation is further smoothed, and the voltage frequency stability of the bus is maintained.
4. The energy management system for a campus hydrogen electrical coupling system of claim 1 wherein: the energy management system has an energy management function and is divided into a day-ahead energy management optimization stage and a day-in energy management optimization stage; the energy management system reads the SQL Server database, predicts loads and photovoltaic power generation power before and during the day, and makes an energy management strategy and a rolling output plan according to the load prediction and photovoltaic power generation prediction results and a user-defined objective function, and refers to an equipment address table and issues a power request.
5. An energy management module photovoltaic power generation prediction method, characterized in that the energy management system based on the hydrogen electric coupling system for parks as claimed in claim 1 comprises the following steps:
step S1: an initial meteorological and photovoltaic power generation sample data set with the time resolution of 5min is adopted and preprocessed;
Step S2: the correlation degree of the meteorological parameters and the photovoltaic power generation power is quantized by adopting an MIC correlation coefficient method, and characteristic parameters are determined;
step S3: determining a proper cluster number through a fuzzy C-means algorithm;
step S4: and (5) establishing a CBA mixed depth model to predict the photovoltaic power generation power of one day in the future.
6. The energy management module photovoltaic power generation prediction method according to claim 5, characterized in that:
in the step S1, abnormal data is processed by adopting a k-nearest neighbor method correction and similar daily data filling method, and the data processed by the data is normalized by a maximum and minimum normalization method;
in step S2, firstly, gridding processing is carried out and an information value IX is calculated; y, then for the information value I [ X ]; carrying out normalization treatment, and finally taking the normalized maximum mutual information value after grid removal as an MIC coefficient;
in step S3, initializing membership value to be a value between 0 and 1 after determining the objective function of FCM, calculating centroid of cluster, and calculating cluster Z i Centroid z of (2) i Updating the value of the membership function:
in step S4, the time sequence information is processed in the forward direction and the reverse direction, and the calculation result of the forward LSTM and the reverse output result of the reverse LSTM are overlapped according to a certain weight to obtain an output.
7. A method of energy management module load prediction, based on the campus hydrogen electrical coupling system of claim 1, comprising the steps of:
step S1: an initial load sample data set with the time resolution of 1h is adopted and preprocessed;
step S2: decomposing the sample data into multi-frequency IMF components by an improved decomposition algorithm, namely a fully adaptive noise set empirical mode decomposition algorithm;
step S3: predicting a low-frequency component and a high-frequency component by adopting a Multiple Linear Regression (MLR) method and an improved time sequence convolutional neural network (TCN) method respectively;
step S4: and superposing and restoring iteration prediction results of the MLR and the TCN to obtain the base load data.
8. The energy management module load prediction method of claim 7, wherein:
in the step S1, abnormal data is processed by adopting a k-nearest neighbor method correction and similar daily data filling method, and the data processed by the data is normalized by a maximum and minimum normalization method;
in the step S2, dividing the sequence signal into a high-frequency component and a low-frequency component according to the zero-crossing rate of the sequence signal, setting the zero-crossing rate to be 0.01 and classifying the zero-crossing rate to be the low-frequency component smaller than 0.01;
in the step S3, aiming at the low-frequency IMF with strong periodicity and weak fluctuation, the MLR is adopted for accurate prediction; the TCN is adopted to effectively train and predict high-frequency IMF components with poor stability and strong fluctuation, and IGWO (insulated gate bipolar transistor) for improving Tent chaos and elite reverse learning is adopted to optimize TCN hyper-parameters so as to improve prediction performance; in order to solve the problem that the numerical distribution generated by the Tent mapping lacks ergodic performance, the Tent mapping is improved:
Where k represents the number of mappings, z k Representing a value corresponding to the mapping;
after the numerical value of the chaotic map is generated, mapping and converting the chaotic variable and the variable participating in optimization:
x ik ＝l k +(u k -l k )×z k ,i∈[1,N],k∈[1,D]
wherein x is ik A value representing the kth dimension of the generated ith wolf, N representing the number of wolves, D representing the dimension of the problem, l k And u k Representing upper and lower limits of the variable to be optimized;
introducing an elite reverse learning mechanism on the basis of improving an initial sample generated by the Tent chaos;
in step S4, the iterative prediction results of the MLR and TCN are superimposed and output.
9. An energy management method based on load prediction and photovoltaic power generation prediction, characterized in that the energy management system based on the campus hydrogen electric coupling system according to claim 3 comprises the following steps:
step S1: in the day-ahead energy management optimization stage, a power plan of a future 24h scheduling period is formulated by taking 1h as time resolution;
step S2: in an energy management optimizing stage in the day, rolling and optimizing a power plan of a future 3h scheduling period by taking 15min as time resolution;
step S3: respectively solving a day-ahead energy management model and a day-in energy management model by adopting an improved gray wolf optimization algorithm to obtain an optimal output plan;
step S4: and rolling and optimizing the daily output plan.
10. The load prediction and photovoltaic power generation prediction based energy management method of claim 9, wherein:
in the step S1, according to the load prediction and photovoltaic power generation prediction results, the total running cost of the system is minimized, the self-balancing rate is maximized as an optimization target, and the energy conservation and equipment output conditions are used as constraints to make a power output plan of a future 24h scheduling period;
in step S2, the difference value between the day-ahead output plan and the ultra-short-term prediction information is balanced by taking the minimum adjustment day-ahead plan as an optimization target and adjusting the charge and discharge of the lithium battery energy storage system, the charge and discharge of the hydrogen energy storage system and the power grid interaction power so as to reduce the influence of the prediction error on the optimization effect;
in step S3, hunting search strategy optimization based on dimension learning is introduced to improve a gray wolf optimization algorithm, and a day-ahead and day-in energy management model is solved;
R i (t)＝||X i (t)-X i_GWO (t+1)||
N i (t)＝{X j (t)|D i (X i (t),X j (t))≤R i (t),X j (t)∈pop}
X i_DLH,d (t+1)＝X i,d (t)+rand*(X n,d (t)-X r,d (t))
wherein X is i (t) represents a wolf group position; x is X i_GWO Representing the position of the t+1st iteration of the ith wolf obtained by updating the conventional GWO; r is R i (t) represents X i (t) and X i_GWO Euclidean distance between; n (N) i (t) represents that the compound satisfies R i (t) X in the radius range i A neighborhood of (t); d (D) i X represents i (t) and X j Euclidean distance of (t); x is X n,d (t) represents d-dimensional data of a wolf randomly selected from within the neighborhood, X r,d (t) represents d-dimensional data of a wolf randomly selected from the population, X i_DLH,d (t+1) represents the result of multi-neighborhood learning.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311421071.4A CN117477609A (en) | 2023-10-30 | 2023-10-30 | Energy management system and method of hydrogen electric coupling system for park |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311421071.4A CN117477609A (en) | 2023-10-30 | 2023-10-30 | Energy management system and method of hydrogen electric coupling system for park |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117477609A true CN117477609A (en) | 2024-01-30 |
Family
ID=89637245
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311421071.4A Pending CN117477609A (en) | 2023-10-30 | 2023-10-30 | Energy management system and method of hydrogen electric coupling system for park |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117477609A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118247085A (en) * | 2024-05-21 | 2024-06-25 | 江苏国氢氢能源科技有限公司 | Operation supervision control system and method for electric-hydrogen energy supply system |
| CN118333342A (en) * | 2024-06-12 | 2024-07-12 | 华能江苏综合能源服务有限公司 | Power generation amount settlement management method and system for photovoltaic power station |
-
2023
- 2023-10-30 CN CN202311421071.4A patent/CN117477609A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118247085A (en) * | 2024-05-21 | 2024-06-25 | 江苏国氢氢能源科技有限公司 | Operation supervision control system and method for electric-hydrogen energy supply system |
| CN118247085B (en) * | 2024-05-21 | 2024-08-27 | 江苏国氢氢能源科技有限公司 | Operation supervision control system and method for electric-hydrogen energy supply system |
| CN118333342A (en) * | 2024-06-12 | 2024-07-12 | 华能江苏综合能源服务有限公司 | Power generation amount settlement management method and system for photovoltaic power station |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109713712B (en) | Optical storage intelligent management integrated machine and optical storage intelligent management method | |
| CN106877338B (en) | The alternating current-direct current micro-capacitance sensor uncertain optimization operation method of the intermittent energy source containing high density | |
| Li et al. | Optimal planning of energy storage system in active distribution system based on fuzzy multi-objective bi-level optimization | |
| CN107910863B (en) | Power distribution network scheduling method comprehensively considering photovoltaic output and load demand prediction interval | |
| CN107798441B (en) | Little grid system of intelligence based on distributed generation | |
| CN102591278B (en) | Microgrid control system having power generation and load forecast functions and microgrid control method | |
| CN104319768B (en) | A kind of micro-capacitance sensor is powered and method for supervising | |
| CN103904695B (en) | A kind ofly close on island many microgrids dynamic dispatching method based on MCS-PSO | |
| Zhang et al. | Research on bi-level optimized operation strategy of microgrid cluster based on IABC algorithm | |
| CN112865075B (en) | AC/DC hybrid micro-grid optimization method | |
| CN104268806A (en) | Micro grid power monitoring system | |
| CN113285479A (en) | Ultra-large-capacity regional micro-grid system and operation method | |
| CN105305488B (en) | A kind of evaluation method for considering new-energy grid-connected and power transmission network utilization rate being influenced | |
| CN117477609A (en) | Energy management system and method of hydrogen electric coupling system for park | |
| CN117411011A (en) | Flexible power load regulating system of multiple power generation systems | |
| CN108075471B (en) | Multi-objective constraint optimization power grid scheduling strategy based on stochastic power output prediction | |
| Lei et al. | Design of energy storage control strategy to improve the PV system power quality | |
| Huangfu et al. | An optimal energy management strategy with subsection bi-objective optimization dynamic programming for photovoltaic/battery/hydrogen hybrid energy system | |
| CN115483701A (en) | Regional autonomous hierarchical regulation and control method for active power distribution network based on power distribution automation master station system | |
| CN105262143A (en) | Photovoltaic source network control method based on optimal regional economy and system | |
| CN116885840A (en) | Distributed new energy online monitoring method and system based on real-time data | |
| Luo et al. | Research on optimal configuration of distributed photovoltaic with typical day scene generation technology | |
| CN114626604B (en) | Distributed photovoltaic observation method and system based on reference station perception | |
| Wu et al. | Research on optimal storage capacity of DC micro‐grid system in PV station | |
| Kusuma et al. | Leveraging PSO algorithms to achieve optimal stand-alone microgrid performance with a focus on battery lifetime |
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 |