CN117439051A - Photovoltaic second-level power prediction method, system and medium based on irradiation monitoring - Google Patents
Photovoltaic second-level power prediction method, system and medium based on irradiation monitoring Download PDFInfo
- Publication number
- CN117439051A CN117439051A CN202311310684.0A CN202311310684A CN117439051A CN 117439051 A CN117439051 A CN 117439051A CN 202311310684 A CN202311310684 A CN 202311310684A CN 117439051 A CN117439051 A CN 117439051A
- Authority
- CN
- China
- Prior art keywords
- photovoltaic
- node
- growth
- monitoring
- irradiance
- 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
- 238000012544 monitoring process Methods 0.000 title claims abstract description 232
- 238000000034 method Methods 0.000 title claims abstract description 92
- 238000004891 communication Methods 0.000 claims abstract description 75
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 47
- 230000008569 process Effects 0.000 claims abstract description 45
- 230000006855 networking Effects 0.000 claims abstract description 37
- 238000012549 training Methods 0.000 claims abstract description 32
- 238000013528 artificial neural network Methods 0.000 claims abstract description 25
- QVFWZNCVPCJQOP-UHFFFAOYSA-N chloralodol Chemical compound CC(O)(C)CC(C)OC(O)C(Cl)(Cl)Cl QVFWZNCVPCJQOP-UHFFFAOYSA-N 0.000 claims abstract description 11
- 239000011159 matrix material Substances 0.000 claims description 14
- 238000004590 computer program Methods 0.000 claims description 13
- 238000012545 processing Methods 0.000 claims description 10
- 238000003860 storage Methods 0.000 claims description 9
- 239000013598 vector Substances 0.000 claims description 7
- 238000004458 analytical method Methods 0.000 claims description 6
- 230000008859 change Effects 0.000 claims description 6
- 230000000694 effects Effects 0.000 claims description 6
- 238000009499 grossing Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 230000005855 radiation Effects 0.000 claims description 4
- 238000010586 diagram Methods 0.000 description 9
- 238000005286 illumination Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 4
- 238000004146 energy storage Methods 0.000 description 3
- 230000007774 longterm Effects 0.000 description 3
- 238000012876 topography Methods 0.000 description 3
- 238000012937 correction Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 238000011835 investigation Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y10/00—Economic sectors
- G16Y10/35—Utilities, e.g. electricity, gas or water
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y20/00—Information sensed or collected by the things
- G16Y20/10—Information sensed or collected by the things relating to the environment, e.g. temperature; relating to location
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y30/00—IoT infrastructure
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y40/00—IoT characterised by the purpose of the information processing
- G16Y40/10—Detection; Monitoring
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y40/00—IoT characterised by the purpose of the information processing
- G16Y40/20—Analytics; Diagnosis
-
- 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
- H02J3/38—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
- H02S—GENERATION OF ELECTRIC POWER BY CONVERSION OF INFRARED RADIATION, VISIBLE LIGHT OR ULTRAVIOLET LIGHT, e.g. USING PHOTOVOLTAIC [PV] MODULES
- H02S50/00—Monitoring or testing of PV systems, e.g. load balancing or fault identification
- H02S50/10—Testing of PV devices, e.g. of PV modules or single PV cells
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Biomedical Technology (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- General Business, Economics & Management (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Power Engineering (AREA)
- Molecular Biology (AREA)
- Water Supply & Treatment (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Human Resources & Organizations (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Environmental & Geological Engineering (AREA)
- Toxicology (AREA)
- Public Health (AREA)
- Remote Monitoring And Control Of Power-Distribution Networks (AREA)
Abstract
The invention discloses a photovoltaic second-level power prediction method, a system and a medium based on irradiation monitoring, which comprise the following steps: the photovoltaic monitoring base station acquires irradiance data around the photovoltaic power station through the photovoltaic monitoring terminals, wherein a plurality of photovoltaic monitoring terminals are arranged around the photovoltaic monitoring base station, and LoRa wireless communication is established between the photovoltaic monitoring base station and the photovoltaic monitoring terminals based on a networking algorithm of a maximum reliable path and an optimal path; the photovoltaic monitoring base station trains the LSTM long-short-period neural network according to irradiance data around the photovoltaic power station, corrects the power output data by the photovoltaic power station in the training process, iterates training until the network residual error enters a steady state or meets the precision requirement, and outputs irradiance and photovoltaic output predicted values of each photovoltaic monitoring terminal at a certain moment in the future by adopting the LSTM long-period neural network. The invention realizes the prediction of the second-level photovoltaic power, and has high prediction precision.
Description
Technical Field
The invention relates to the technical field of photovoltaic power generation, in particular to a photovoltaic second-level power prediction method, a system and a medium based on irradiation monitoring.
Background
Along with the construction of a novel power system, a large number of photovoltaic power stations are connected into a power grid, and randomness, fluctuation and intermittence of the photovoltaic power stations bring great challenges to power grid power and electric quantity balance and safe and stable operation. An important means for solving uncertainty of the photovoltaic power is to accurately predict and forecast the output of the photovoltaic power when the photovoltaic power is complementarily combined with a plurality of energy sources such as a conventional power source, an energy storage source and the like.
Currently, a great deal of research has been carried out for photovoltaic power prediction, mainly including two types of methods: direct prediction methods and indirect prediction methods. The direct prediction method is a prediction method based on data statistics, and is a method for predicting photovoltaic output conditions under different weather conditions in the future according to the rule of photovoltaic historical output data, and a large amount of historical data needs to be collected and counted. The indirect prediction method is a method for directly predicting according to weather forecast data without any photovoltaic historical data, and has the main advantages of reducing a large amount of historical data statistics, but having high dependence on the processing capacity of the log weather forecast and cloud pictures on the prediction accuracy.
Currently, photovoltaic pre-day short-term prediction and intra-day 4-hour ultra-short-term prediction are of more interest according to application scheduling plan requirements. However, when the photovoltaic, the hydropower, the pumped storage, the energy storage and the like form a multi-energy complementary system to jointly generate power, the photovoltaic power prediction needs to be improved to the second level when the photovoltaic and the hydropower, the pumped storage, the energy storage and the like are jointly involved in the stability control, and related researches are less. In addition, in southwest area, photovoltaic is distributed on river canyons and mountains, communication conditions are poor, the conventional wired networking mode is quite unrealistic for the situation of mountain areas, and in addition, the coverage of public network signals such as GPRS/4G and the like in the mountain areas is quite uneven, so that communication quality is difficult to guarantee.
Disclosure of Invention
The technical problems to be solved by the invention are that the existing photovoltaic power prediction method is low in prediction precision, cannot reach the second level, and the photovoltaic power prediction is affected by poor quality due to poor networking of photovoltaic distribution communication in mountain areas.
The invention aims to provide a photovoltaic second-level power prediction method, a system and a medium based on irradiation monitoring, aiming at the photovoltaic second-level power fluctuation prediction requirement, the photovoltaic fluctuation characteristic monitoring complete equipment is used for networking monitoring irradiance data around a photovoltaic power station, the irradiance data is combined with photovoltaic duration output data, and a long-period memory coding algorithm is used for carrying out second-level photovoltaic power prediction, so that the prediction accuracy is high.
The invention is realized by the following technical scheme:
in a first aspect, the present invention provides a photovoltaic second level power prediction method based on irradiation monitoring, the method comprising:
the photovoltaic monitoring base station acquires irradiance data around the photovoltaic power station through the photovoltaic monitoring terminals, wherein a plurality of photovoltaic monitoring terminals are arranged around the photovoltaic monitoring base station, and LoRa wireless communication is established between the photovoltaic monitoring base station and the photovoltaic monitoring terminals based on a networking algorithm of a maximum reliable path and an optimal path;
the photovoltaic monitoring base station trains the LSTM long-short-period neural network according to irradiance data around the photovoltaic power station, corrects the power output data by the photovoltaic power station in the training process, iterates training until the network residual error enters a steady state or meets the precision requirement, and outputs irradiance and photovoltaic output predicted values of each photovoltaic monitoring terminal at a certain moment in the future by adopting the LSTM long-period neural network.
Further, irradiance data around the photovoltaic power station comprises illumination intensity, temperature and the like, and the invention considers that the temperature is unchanged for the second-level time scale, and only the illumination intensity is considered.
A plurality of photovoltaic monitoring terminals are arranged around photovoltaic monitoring basic station, includes: a plurality of photovoltaic monitoring terminals are distributed according to a multi-tree network, a photovoltaic monitoring base station is used as a master node, the photovoltaic monitoring terminals are used as slave nodes, and the slave nodes are divided into multiple levels of slave nodes according to different network levels;
the master node carries a scheduling task of the downlink communication network, and an intermediate slave node in the multi-stage slave nodes realizes ordered relay.
Further, the networking algorithm based on the maximum reliable path and the optimal path comprises the following steps:
adopting a multi-tree network growth strategy to perform multi-tree network growth on the nodes to be assembled to obtain communication links of the nodes to be assembled; the multi-tree network growth comprises one-level node growth and multi-level node growth;
the node to be assembled refers to a photovoltaic monitoring terminal which is not connected to the network.
Further, the multi-tree network growth strategy comprises an optimal path growth strategy and a maximum reliable path growth strategy;
in the process of growing the optimal path growth strategy into the multi-way tree, the node to be networked continuously searches for a more optimal path through a preferred algorithm until the path traversal is completed; the method comprises the following steps: when snr_m+snr_s > =snr_m '+snr_s' and min (snr_s, snr_m) > =min (snr_m ', snr_s') is satisfied, the link of the node to be networked is replaced with a link (snr_s, snr_m), otherwise it remains unchanged; wherein, SLink ' (SNR_S ', SNR_M ') is the former link of the node S to be networked; SLink (snr_s, snr_m) is the current link;
In the process of the maximum reliable path growth strategy being the multi-way tree growth, the node to be networked continuously searches for a better path through a preferred algorithm, when the current communication link (SNR_S, SNR_M) reaches the standard that the communication quality is good or above, the node stops growing, otherwise, the node is stopped until the path traversal is completed; the method comprises the following steps: a threshold with excellent communication quality is defined as W, and a good threshold is defined as G. The communication quality is excellent when snr_s > W and snr_m > W, otherwise, the communication quality is good when snr_s > G and snr_m > G, otherwise, the communication quality is poor.
Further, the specific steps of the first-level node growth are as follows:
step 11, the master node initiates a networking request to the node Si to be networked, and starts overtime waiting for receiving a reply message; if the received reply message is overtime, the current slave node to be networked fails to be networked, the step 14 is entered, and otherwise, the step 12 is entered;
step 12, receiving a reply message within the timeout period, analyzing the reply message, if the analysis is unsuccessful, the current node to be networked fails to be networked, entering step 14, otherwise, entering step 13;
step 13, marking the node Si as a primary slave node and in a network access state, grading the communication quality according to the communication link of the node Si, and entering step 14;
Step 14, searching the next node to be networked in the slave node set { S }, if the slave node set { S } is traversed, ending the primary networking process, entering step 15, otherwise, entering step 11;
step 15, the node set with the nodes in the slave node set { S } being primary slave nodes and the communication quality being good or excellent is recorded as a first intermediate node set { L1}, and the non-networked slave node set is recorded as a second intermediate node set { U1}; if { U1} is empty, all slave nodes enter the network to finish the growth process, otherwise, enter the secondary node growth process.
Further, the specific steps of the multi-stage node growth are as follows:
step 21, the master node initiates a network growth request to a node L1i in the first intermediate node set { L1 };
step 22, the node L1i sends networking request information to the node U1i in the second intermediate node set { U1}, and starts to wait for receiving the reply message over time; if the received reply message is overtime, the current slave node to be networked fails to be networked, the step 24 is entered, and otherwise, the step 23 is entered;
step 23, receiving the reply message within the timeout period, and analyzing the reply message; if the analysis is unsuccessful, the current node to be networked fails to be networked, and the step 25 is entered, otherwise, the step 24 is entered;
Step 24, if the communication link (snr_s, snr_m) of the node U1i meets the growth policy, marking as a secondary slave node and in a network access state, otherwise, marking as a non-network access state, and entering step 25;
step 25, go on traversing the second intermediate node set { U1}, if the second intermediate node set { U1} is not traversed, search for the next node to be networked from the second intermediate node set { U1}, enter step 22, otherwise: if no network node is not accessed from the node set { S }, the secondary node growth is completed, all the nodes are completed, otherwise: continuing traversing the first intermediate node set { L1}, searching the next level node from the first intermediate node set { L1} if the traversing is not completed, and entering a step 21, otherwise: if the first intermediate node set { U1} has no non-network node, the second-level node growth is completed, all the nodes are completed, and if not, the third-level node growth process is entered.
Further, outputting irradiance and photovoltaic output predicted values of each photovoltaic monitoring terminal at a future moment, including:
carrying out interpolation smoothing processing on irradiance data of each photovoltaic monitoring terminal (namely photovoltaic monitoring points) in combination with coordinate information of the irradiance data, fitting the irradiance data and the photovoltaic output data to a two-dimensional curved surface, obtaining irradiance data and photovoltaic output data at any position on the two-dimensional curved surface, and realizing prediction of photovoltaic output; wherein the predicted value of the photovoltaic output The calculation formula of (2) is as follows:
wherein X represents irradiance column vectors of all photovoltaic monitoring terminals (namely photovoltaic monitoring points); a is that t The influence matrix at the time t is represented, and represents irradiance of each photovoltaic monitoring terminal at the time t in the past, and influence degree of irradiance change in the future is represented; a is that t-j The influence matrix at the moment t-j is represented, j is automatically optimized according to the algorithm effect, namely irradiance at the moment j in the past is representedAll the trends are affected, and data from the past 10 moments can be taken as input for training.
In a second aspect, the present invention further provides a photovoltaic second level power prediction method based on irradiation monitoring, the method comprising:
the photovoltaic monitoring terminal collects irradiance data around the photovoltaic power station and transmits the irradiance data around the photovoltaic power station to the photovoltaic monitoring base station; the photovoltaic monitoring terminals are arranged around the photovoltaic monitoring base station, and LoRa wireless communication is established between the photovoltaic monitoring terminals and the photovoltaic monitoring base station based on a networking algorithm of the maximum reliable path and the optimal path;
the photovoltaic monitoring terminal receives irradiance and a photovoltaic output predicted value at a certain moment in the future; the irradiance and photovoltaic output predicted value at a certain moment in the future is obtained by training the LSTM long-short-period neural network by the photovoltaic monitoring base station according to irradiance data around the photovoltaic power station, correcting the power output data by the photovoltaic power station in the training process, iterating the training until the network residual error enters a steady state or meets the precision requirement, and outputting by the LSTM long-short-period neural network.
Further, a plurality of photovoltaic monitoring terminals set up around photovoltaic monitoring basic station, include:
a plurality of photovoltaic monitoring terminals are distributed according to a multi-tree network, a photovoltaic monitoring base station is used as a master node, the photovoltaic monitoring terminals are used as slave nodes, and the slave nodes are divided into multiple levels of slave nodes according to different network levels;
the master node carries a scheduling task of the downlink communication network, and an intermediate slave node in the multi-stage slave nodes realizes ordered relay.
Further, the networking algorithm based on the maximum reliable path and the optimal path comprises the following steps: adopting a multi-tree network growth strategy to perform multi-tree network growth on the nodes to be assembled to obtain communication links of the nodes to be assembled; the multi-tree network growth comprises one-level node growth and multi-level node growth;
the node to be assembled refers to a photovoltaic monitoring terminal which is not connected with the network;
the multi-tree network growth strategy comprises an optimal path growth strategy and a maximum reliable path growth strategy;
in the process of growing the optimal path growth strategy into the multi-way tree, the node to be networked continuously searches for a more optimal path through a preferred algorithm until the path traversal is completed;
in the process of the maximum reliable path growth strategy being the multi-way tree growth, the node to be networked continuously searches for a better path through a preferred algorithm, when the current communication link reaches the standard with good communication quality or above, the node stops growing, otherwise, the node is stopped until the path traversal is completed. In a third aspect, the present invention further provides a photovoltaic second level power prediction system based on radiation monitoring, the system comprising:
The photovoltaic monitoring base station is used for acquiring irradiance data around the photovoltaic power station from the photovoltaic monitoring terminal, training the LSTM long-short-period neural network according to the irradiance data around the photovoltaic power station, correcting the power output data of the photovoltaic power station in the training process, performing iterative training until the network residual error enters a steady state or meets the precision requirement, outputting irradiance and photovoltaic output predicted values of each photovoltaic monitoring terminal at a certain moment in the future by adopting the LSTM long-period neural network, and transmitting the irradiance and the photovoltaic output predicted values to the corresponding photovoltaic monitoring terminal;
the photovoltaic monitoring terminal is used for collecting irradiance data around the photovoltaic power station and transmitting the irradiance data around the photovoltaic power station to the photovoltaic monitoring base station; receiving irradiance and photovoltaic output predicted values at a future moment transmitted by a photovoltaic monitoring base station;
and a plurality of photovoltaic monitoring terminals are arranged around the photovoltaic monitoring base station, and LoRa wireless communication is established between the photovoltaic monitoring base station and the photovoltaic monitoring terminals based on a networking algorithm of the maximum reliable path and the optimal path.
Further, arrange a plurality of photovoltaic monitoring terminals around the photovoltaic monitoring basic station, include:
a plurality of photovoltaic monitoring terminals are distributed according to a multi-tree network, a photovoltaic monitoring base station is used as a master node, the photovoltaic monitoring terminals are used as slave nodes, and the slave nodes are divided into multiple levels of slave nodes according to different network levels;
The master node carries a scheduling task of the downlink communication network, and an intermediate slave node in the multi-stage slave nodes realizes ordered relay.
Further, the networking algorithm based on the maximum reliable path and the optimal path comprises the following steps:
adopting a multi-tree network growth strategy to perform multi-tree network growth on the nodes to be assembled to obtain communication links of the nodes to be assembled; the multi-tree network growth comprises one-level node growth and multi-level node growth;
the node to be assembled refers to a photovoltaic monitoring terminal which is not connected with the network;
the multi-tree network growth strategy comprises an optimal path growth strategy and a maximum reliable path growth strategy;
in the process of growing the optimal path growth strategy into the multi-way tree, the node to be networked continuously searches for a more optimal path through a preferred algorithm until the path traversal is completed;
in the process of the maximum reliable path growth strategy being the multi-way tree growth, the node to be networked continuously searches for a better path through a preferred algorithm, when the current communication link reaches the standard with good communication quality or above, the node stops growing, otherwise, the node is stopped until the path traversal is completed.
Further, outputting irradiance and photovoltaic output predicted values of each photovoltaic monitoring terminal at a future moment, including:
Carrying out interpolation smoothing processing on irradiance data of each photovoltaic monitoring terminal in combination with coordinate information of the irradiance data, fitting the irradiance data into a two-dimensional curved surface, obtaining irradiance data and photovoltaic output data at any position on the two-dimensional curved surface, and realizing prediction of photovoltaic output; wherein the predicted value of the photovoltaic outputThe calculation formula of (2) is as follows:
wherein X represents the radiation of each photovoltaic monitoring terminalAn illuminance column vector; a is that t The influence matrix at the time t is represented, and represents irradiance of each photovoltaic monitoring terminal at the time t in the past, and influence degree of irradiance change in the future is represented; a is that t-j The influence matrix at the moment t-j is represented, j is automatically optimized according to the algorithm effect, namely irradiance at the moment j in the past has influence on future trend.
In a fourth aspect, the present invention further provides a computer readable storage medium, where a computer program is stored, where the computer program when executed by a processor implements a photovoltaic second level power prediction method based on radiation monitoring as described above.
Compared with the prior art, the invention has the following advantages and beneficial effects:
according to the photovoltaic second-level power prediction method, system and medium based on irradiation monitoring, photovoltaic fluctuation characteristic monitoring complete equipment is used for networking monitoring irradiance data around a photovoltaic power station according to photovoltaic second-level power fluctuation prediction requirements, the irradiance data is combined with photovoltaic duration output data, and a long-period memory coding algorithm is used for carrying out second-level photovoltaic power prediction, so that prediction accuracy is high.
Drawings
The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention. In the drawings:
FIG. 1 is a diagram of the overall system architecture of a photovoltaic monitoring terminal of the present invention;
FIG. 2 is a network diagram of a photovoltaic monitoring terminal arrangement of the present invention;
FIG. 3 is a schematic view of the installation position of a photovoltaic fluctuation monitoring terminal of a photovoltaic power station according to the present invention;
fig. 4 is a schematic diagram of actual measurement data of each illuminance monitoring terminal of a certain photovoltaic power station according to the present invention;
fig. 5 is a graph of photovoltaic second level predictions and their errors (predicted future 5 s) according to the present invention.
Detailed Description
For the purpose of making apparent the objects, technical solutions and advantages of the present invention, the present invention will be further described in detail with reference to the following examples and the accompanying drawings, wherein the exemplary embodiments of the present invention and the descriptions thereof are for illustrating the present invention only and are not to be construed as limiting the present invention.
The prediction accuracy is not high, the second level cannot be achieved, and the problems that the photovoltaic power prediction is affected due to poor quality and poor photovoltaic distribution communication networking in mountain areas are solved. The invention designs a photovoltaic second-level power prediction method, a photovoltaic second-level power prediction system and a photovoltaic second-level power prediction medium based on irradiation monitoring, aiming at photovoltaic second-level power fluctuation prediction requirements, networking monitoring irradiance data around a photovoltaic power station through a photovoltaic fluctuation characteristic monitoring complete device, and performing second-level photovoltaic power prediction by combining irradiance data with photovoltaic duration output data and adopting a long-period memory coding algorithm, wherein the prediction accuracy is high.
Specifically, the photovoltaic fluctuation characteristic monitoring complete device comprises a photovoltaic monitoring terminal and a photovoltaic monitoring base station, wherein the photovoltaic monitoring terminal monitors irradiance data (such as illumination intensity) around a photovoltaic power station in real time and transmits the irradiance data to the photovoltaic monitoring base station in a wireless mode, and the photovoltaic monitoring base station sends the data to a photovoltaic prediction system in a wired mode after receiving the data of the photovoltaic monitoring terminal. And the photovoltaic monitoring base station and the photovoltaic monitoring terminal are in wireless communication through LoRa. The system architecture is shown in fig. 1 below.
The photovoltaic monitoring terminals are distributed as shown in fig. 2. In order to realize reliable communication between the photovoltaic monitoring terminal and the photovoltaic monitoring base station, the invention establishes a LoRa communication private network between the photovoltaic monitoring terminal and the photovoltaic monitoring base station, and provides a networking algorithm based on a maximum reliable path and an optimal path.
Irradiance data of each photovoltaic monitoring terminal is obtained according to the networking algorithm, the duration output data of the photovoltaic power station is combined as a sample, and the LSTM long-term neural network algorithm is adopted for training, so that a photovoltaic power predicted value of 5-10 seconds in the future can be obtained.
Example 1
As shown in fig. 1, the photovoltaic second-level power prediction method based on irradiation monitoring of the present invention comprises:
The photovoltaic monitoring base station acquires irradiance data around the photovoltaic power station through the photovoltaic monitoring terminals, wherein a plurality of photovoltaic monitoring terminals are arranged around the photovoltaic monitoring base station, and LoRa wireless communication is established between the photovoltaic monitoring base station and the photovoltaic monitoring terminals based on a networking algorithm of a maximum reliable path and an optimal path;
the photovoltaic monitoring base station trains the LSTM long-short-period neural network according to irradiance data around the photovoltaic power station, corrects the power output data by the photovoltaic power station in the training process, iterates training until the network residual error enters a steady state or meets the precision requirement, and outputs irradiance and photovoltaic output predicted values of each photovoltaic monitoring terminal at a certain moment in the future by adopting the LSTM long-period neural network.
As shown in fig. 2, specifically, the photovoltaic monitoring terminals (the illumination monitoring sensor and the communication base station) are uniformly arranged around the photovoltaic power station in a square matrix form with the photovoltaic monitoring base station (i.e. the photovoltaic power station) as a center, and the photovoltaic monitoring data are collected to the communication base station installed on the main control roof of the photovoltaic power station.
According to the arrangement of the photovoltaic power stations, the interval between the adjacent photovoltaic monitoring terminals is generally about 1-2 km, the proper sparsity can be arranged in the area with the gentle topography, the number of distributed points can be increased in the area with the poorer topography compared with the concave-convex communication condition, and the specific interval is based on actual investigation.
Taking a photovoltaic power station in Sichuan Aba state as an example, through on-site actual investigation, the signal link and the topography condition are comprehensively considered, the on-site deployment condition of irradiance monitoring terminals is shown in figure 3, and the GPS coordinate information of each photovoltaic monitoring terminal is shown in table 1.
Table 1 the GPS coordinates of each photovoltaic monitoring terminal are as follows:
specifically, the networking algorithm based on the maximum reliable path and the optimal path is as follows:
firstly, distributing a plurality of photovoltaic monitoring terminals according to a multi-tree network, taking a photovoltaic monitoring base station as a master node and a photovoltaic monitoring terminal as a slave node, and dividing the slave node into multiple slave nodes according to different network levels; as shown in fig. 2, B, C, D, E is a primary slave node, F, G, H, I is a secondary slave node, and J, K is a tertiary slave node. The master node carries the scheduling task of the downlink communication network, and the intermediate slave nodes in the multi-stage slave nodes realize orderly relay, so that the multi-fork tree has a growth characteristic.
To describe the growth process of the "multi-tree", the following terms are defined:
1) From the node set { S }: the set of slave nodes is known in the network, such as primary slave node B, C, D, E, secondary slave node F, G, H, I, etc. as shown in fig. 2.
2) Node Si to be networked: nodes not networked in { S }.
3) Si communication link SLink: the master node can establish a communication link path with Si, and A-B-I-J is a communication link of the slave node J.
4) Communication quality of SLink (snr_s, snr_m): wherein, snr_s is the signal-to-noise ratio of Si correctly receiving the networking request of the parent node, and snr_m is the signal-to-noise ratio of Si correctly receiving the networking acknowledgement frame returned by Si. For (B, I), B is a parent node, denoted as 'M', I is a child node, denoted as 'S'.
(1) "Multi-tree" network growth strategy
The multi-tree network growth strategy comprises an optimal path growth strategy and a maximum reliable path growth strategy;
in the process of growing the optimal path growth strategy into the multi-way tree, the node to be networked continuously searches for a more optimal path through a preferred algorithm until the path traversal is completed; the method comprises the following steps: when snr_m+snr_s > =snr_m '+snr_s' and min (snr_s, snr_m) > =min (snr_m ', snr_s') is satisfied, the link of the node to be networked is replaced with a link (snr_s, snr_m), otherwise it remains unchanged; wherein, SLink ' (SNR_S ', SNR_M ') is the former link of the node S to be networked; SLink (snr_s, snr_m) is the current link;
in the process of the maximum reliable path growth strategy being the multi-way tree growth, the node to be networked continuously searches for a better path through a preferred algorithm, when the current communication link (SNR_S, SNR_M) reaches the standard that the communication quality is good or above, the node stops growing, otherwise, the node is stopped until the path traversal is completed; the method comprises the following steps: a threshold with excellent communication quality is defined as W, and a good threshold is defined as G. The communication quality is excellent when snr_s > W and snr_m > W, otherwise, the communication quality is good when snr_s > G and snr_m > G, otherwise, the communication quality is poor.
(2) "Multi-tree" network growth algorithm
According to the standard, the master node can predict the information of the slave node and carry the scheduling task of the growth of the multi-way tree. In addition, a "multi-tree" network growth strategy needs to be preset, which is defined as ST, and ST can be an optimal path growth strategy or a maximum reliable path growth strategy.
Adopting a multi-tree network growth strategy to perform multi-tree network growth on the nodes to be assembled to obtain communication links of the nodes to be assembled; the multi-tree network growth comprises one-level node growth and multi-level node growth; the node to be assembled refers to a photovoltaic monitoring terminal which is not connected to the network.
A) First-level node growth process
Specifically, the specific steps of the first-level node growth are as follows:
step 11, the master node initiates a networking request to the node Si to be networked, and starts overtime waiting for receiving a reply message; if the received reply message is overtime, the current slave node to be networked fails to be networked, the step 14 is entered, and otherwise, the step 12 is entered;
step 12, receiving a reply message within the timeout period, analyzing the reply message, if the analysis is unsuccessful, the current node to be networked fails to be networked, entering step 14, otherwise, entering step 13;
step 13, marking the node Si as a primary slave node and in a network access state, grading the communication quality according to a communication link Slink (SNR_S, SNR_M) of the node Si, and entering step 14;
Step 14, searching the next node to be networked in the slave node set { S }, if the slave node set { S } is traversed, ending the primary networking process, entering step 15, otherwise, entering step 11;
step 15, the node set with the nodes in the slave node set { S } being primary slave nodes and the communication quality being good or excellent is recorded as a first intermediate node set { L1}, and the non-networked slave node set is recorded as a second intermediate node set { U1}; if { U1} is empty, all slave nodes enter the network to finish the growth process, otherwise, enter the secondary node growth process.
B) Multistage node growth process
The multi-level node growth is based on the completion of the previous-level node growth. The secondary node growth process is described below as an example. The multi-stage node growth process is analogized.
The specific steps of the secondary node growth are as follows:
step 21, the master node initiates a network growth request to a node L1i in the first intermediate node set { L1 };
step 22, the node L1i sends networking request information to the node U1i in the second intermediate node set { U1}, and starts to wait for receiving the reply message over time; if the received reply message is overtime, the current slave node to be networked fails to be networked, the step 24 is entered, and otherwise, the step 23 is entered;
step 23, receiving the reply message within the timeout period, and analyzing the reply message; if the analysis is unsuccessful, the current node to be networked fails to be networked, and the step 25 is entered, otherwise, the step 24 is entered;
Step 24, if the communication link (snr_s, snr_m) of the node U1i meets the growth policy, marking as a secondary slave node and in a network access state, otherwise, marking as a non-network access state, and entering step 25;
step 25, go on traversing the second intermediate node set { U1}, if the second intermediate node set { U1} is not traversed, search for the next node to be networked from the second intermediate node set { U1}, enter step 22, otherwise: if no network node is not accessed from the node set { S }, the secondary node growth is completed, all the nodes are completed, otherwise: continuing traversing the first intermediate node set { L1}, searching the next level node from the first intermediate node set { L1} if the traversing is not completed, and entering a step 21, otherwise: if the first intermediate node set { U1} has no non-network node, the second-level node growth is completed, all the nodes are completed, and if not, the third-level node growth process is entered.
In particular, irradiance data around the photovoltaic power station comprises illumination intensity, temperature and the like, and the temperature is considered to be unchanged for the second-level time scale, and only the illumination intensity is considered.
Specifically, the photovoltaic monitoring base station trains the LSTM long-short-period neural network according to irradiance data around the photovoltaic power station, corrects the power output data by the photovoltaic power station in the training process, iterates training until the network residual error enters a steady state or meets the precision requirement, and outputs irradiance and photovoltaic output predicted values of each photovoltaic monitoring terminal at a certain moment in the future by adopting the LSTM long-period neural network.
According to irradiance data around the photovoltaic power station, the duration output data of the photovoltaic power station is combined as a sample, an LSTM long-term neural network is adopted for training, a photovoltaic power predicted value of 5-10 seconds in the future can be obtained, and single independent prediction can be carried out based on each photovoltaic monitoring terminal (namely a photovoltaic monitoring point) or overall prediction can be carried out.
Specifically, the single independent prediction is to predict irradiance at the next time point by taking a monitoring point of a photovoltaic monitoring terminal as an input; inputting the last ten sampling points into an STM long-term neural network, and outputting a predicted value; and according to the photovoltaic monitoring points of the square matrix, the difference value of the two-dimensional curved surface is obtained. Wherein the elapsed force data is used as a correction feedback to make the correction.
Specifically, the integral prediction is to input the duration output data and irradiance data of each monitoring point into the STM long-short-term neural network together, train a neural network model, output the predicted value of each photovoltaic monitoring point at one time, and then interpolate.
The specific algorithm flow is as follows:
let X denote irradiance row vector of each photovoltaic monitoring terminal, then there is predicted value of photovoltaic outputThe method comprises the following steps:
wherein X represents irradiance column vectors of all photovoltaic monitoring terminals (namely photovoltaic monitoring points); a is that t The influence matrix at the time t is represented, and represents irradiance of each photovoltaic monitoring terminal at the time t in the past, and influence degree of irradiance change in the future is represented; a is that t-j The influence matrix at the time t-j is represented, j is automatically optimized according to the algorithm effect, namely irradiance at the time j in the past has influence on future trend, and data at the time 10 in the past can be taken as input for training.
The method comprises the steps of taking the duration irradiance data of each photovoltaic monitoring terminal at a plurality of past points as the input of an LSTM long-short-term neural network, training, correcting by adopting actually measured historical output data in the training process, performing training iteration until a network residual enters a steady state or meets the precision requirement, and outputting irradiance and photovoltaic output predicted values of each photovoltaic monitoring terminal at a certain moment in the future.
And (3) carrying out interpolation smoothing processing on irradiance data of each photovoltaic monitoring terminal (namely photovoltaic monitoring points) in combination with coordinate information of the irradiance data, fitting the irradiance data and the photovoltaic output data to a two-dimensional curved surface, obtaining irradiance data and photovoltaic output data at any position on the two-dimensional curved surface, and realizing prediction of photovoltaic output.
In this embodiment, irradiance monitoring data obtained by 32 photovoltaic monitoring terminals on site of a photovoltaic power station in Sichuan of 2021 is shown in fig. 4. The photovoltaic output for the next 5 seconds was predicted as shown in fig. 5. The predicted error is mostly within +/-1 MW, the maximum error is not more than 1.5MW, and the photovoltaic rated capacity is 50MW, namely the error is not more than 3%.
Example 2
As shown in fig. 1, the difference between the present embodiment and embodiment 1 is that the present embodiment provides a photovoltaic second level power prediction system based on irradiation monitoring, which uses a photovoltaic second level power prediction method based on irradiation monitoring of embodiment 1; the system comprises:
the photovoltaic monitoring base station is used for acquiring irradiance data around the photovoltaic power station from the photovoltaic monitoring terminal, training the LSTM long-short-period neural network according to the irradiance data around the photovoltaic power station, correcting the power output data of the photovoltaic power station in the training process, performing iterative training until the network residual error enters a steady state or meets the precision requirement, outputting irradiance and photovoltaic output predicted values of each photovoltaic monitoring terminal at a certain moment in the future by adopting the LSTM long-period neural network, and transmitting the irradiance and the photovoltaic output predicted values to the corresponding photovoltaic monitoring terminal;
the photovoltaic monitoring terminal is used for collecting irradiance data around the photovoltaic power station and transmitting the irradiance data around the photovoltaic power station to the photovoltaic monitoring base station; receiving irradiance and photovoltaic output predicted values at a future moment transmitted by a photovoltaic monitoring base station;
and a plurality of photovoltaic monitoring terminals are arranged around the photovoltaic monitoring base station, and LoRa wireless communication is established between the photovoltaic monitoring base station and the photovoltaic monitoring terminals based on a networking algorithm of the maximum reliable path and the optimal path.
Further, arrange a plurality of photovoltaic monitoring terminals around the photovoltaic monitoring basic station, include:
a plurality of photovoltaic monitoring terminals are distributed according to a multi-tree network, a photovoltaic monitoring base station is used as a master node, the photovoltaic monitoring terminals are used as slave nodes, and the slave nodes are divided into multiple levels of slave nodes according to different network levels;
the master node carries a scheduling task of the downlink communication network, and an intermediate slave node in the multi-stage slave nodes realizes ordered relay.
Further, the networking algorithm based on the maximum reliable path and the optimal path comprises the following steps:
adopting a multi-tree network growth strategy to perform multi-tree network growth on the nodes to be assembled to obtain communication links of the nodes to be assembled; the multi-tree network growth comprises one-level node growth and multi-level node growth;
the node to be assembled refers to a photovoltaic monitoring terminal which is not connected with the network;
the multi-tree network growth strategy comprises an optimal path growth strategy and a maximum reliable path growth strategy;
in the process of growing the optimal path growth strategy into the multi-way tree, the node to be networked continuously searches for a more optimal path through a preferred algorithm until the path traversal is completed;
in the process of the maximum reliable path growth strategy being the multi-way tree growth, the node to be networked continuously searches for a better path through a preferred algorithm, when the current communication link reaches the standard with good communication quality or above, the node stops growing, otherwise, the node is stopped until the path traversal is completed.
Further, outputting irradiance and photovoltaic output predicted values of each photovoltaic monitoring terminal at a future moment, including:
carrying out interpolation smoothing processing on irradiance data of each photovoltaic monitoring terminal in combination with coordinate information of the irradiance data, fitting the irradiance data into a two-dimensional curved surface, obtaining irradiance data and photovoltaic output data at any position on the two-dimensional curved surface, and realizing prediction of photovoltaic output; wherein the predicted value of the photovoltaic outputThe calculation formula of (2) is as follows:
wherein X represents irradiance column vectors of all photovoltaic monitoring terminals; a is that t The influence matrix at the time t is represented, and represents irradiance of each photovoltaic monitoring terminal at the time t in the past, and influence degree of irradiance change in the future is represented; a is that t-j The influence matrix at the moment t-j is represented, j is automatically optimized according to the algorithm effect, namely irradiance at the moment j in the past has influence on future trend.
The specific implementation of the first-level node growth and the multi-level node growth is performed according to the steps of the photovoltaic second-level power prediction method based on irradiation monitoring in embodiment 1, and the detailed description is omitted in this embodiment.
Meanwhile, the invention also provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program realizes the photovoltaic second level power prediction method based on irradiation monitoring of the embodiment 1 when being executed by a processor.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The foregoing description of the embodiments has been provided for the purpose of illustrating the general principles of the invention, and is not meant to limit the scope of the invention, but to limit the invention to the particular embodiments, and any modifications, equivalents, improvements, etc. that fall within the spirit and principles of the invention are intended to be included within the scope of the invention.
Claims (15)
1. The photovoltaic second-level power prediction method based on irradiation monitoring is characterized by comprising the following steps of:
the photovoltaic monitoring base station acquires irradiance data around the photovoltaic power station through the photovoltaic monitoring terminals, wherein a plurality of photovoltaic monitoring terminals are arranged around the photovoltaic monitoring base station, and LoRa wireless communication is established between the photovoltaic monitoring base station and the photovoltaic monitoring terminals based on a networking algorithm of a maximum reliable path and an optimal path;
and the photovoltaic monitoring base station trains the LSTM long-short-period neural network according to irradiance data around the photovoltaic power station, corrects the continuous output data of the photovoltaic power station in the training process, iterates training until the network residual error enters a steady state or meets the precision requirement, and outputs irradiance and photovoltaic output predicted values of each photovoltaic monitoring terminal at a certain moment in the future by adopting the LSTM long-period neural network.
2. The photovoltaic second-level power prediction method based on irradiation monitoring according to claim 1, wherein a plurality of photovoltaic monitoring terminals are arranged around a photovoltaic monitoring base station, and the method comprises the steps of: distributing a plurality of photovoltaic monitoring terminals according to a multi-tree network, taking the photovoltaic monitoring base station as a master node and the photovoltaic monitoring terminal as a slave node, and dividing the slave node into multiple levels of slave nodes according to different network levels;
The master node carries a scheduling task of a downlink communication network, and an intermediate slave node in the multi-stage slave nodes realizes ordered relay.
3. The photovoltaic second level power prediction method based on irradiation monitoring according to claim 2, wherein the networking algorithm based on the maximum reliable path and the optimal path comprises:
adopting a multi-tree network growth strategy to perform multi-tree network growth on the nodes to be assembled to obtain communication links of the nodes to be assembled; the multi-tree network growth comprises one-level node growth and multi-level node growth;
the node to be assembled refers to a photovoltaic monitoring terminal which is not connected to the network.
4. A photovoltaic second level power prediction method based on irradiance monitoring according to claim 3, wherein the multi-tree network growth strategy comprises an optimal path growth strategy and a maximum reliable path growth strategy;
in the process of growing the optimal path growth strategy into the multi-way tree, the node to be networked continuously searches for a better path through a preferred algorithm until path traversal is completed;
in the process of the maximum reliable path growth strategy being the multi-way tree growth, the node to be networked continuously searches for a better path through a preferred algorithm, when the current communication link reaches the standard with good communication quality or above, the node stops growing, otherwise, the node is not grown until the path traversal is completed.
5. A photovoltaic second level power prediction method based on irradiation monitoring according to claim 3, wherein the specific steps of the first level node growth are as follows:
step 11, the master node initiates a networking request to the node Si to be networked, and starts overtime waiting for receiving a reply message; if the received reply message is overtime, the current slave node to be networked fails to be networked, the step 14 is entered, and otherwise, the step 12 is entered;
step 12, receiving a reply message within the timeout period, analyzing the reply message, if the analysis is unsuccessful, the current node to be networked fails to be networked, entering step 14, otherwise, entering step 13;
step 13, marking the node Si as a primary slave node and in a network access state, grading the communication quality according to the communication link of the node Si, and entering step 14;
step 14, searching the next node to be networked in the slave node set { S }, if the traversal of the slave node set { S } is completed, ending the primary networking process, entering step 15, otherwise, entering step 11;
step 15, the node set with the nodes in the slave node set { S } being primary slave nodes and the communication quality being good or excellent is recorded as a first intermediate node set { L1}, and the slave node set without networking is recorded as a second intermediate node set { U1}; if { U1} is empty, all slave nodes enter the network to finish the growth process, otherwise, enter the secondary node growth process.
6. The photovoltaic second level power prediction method based on irradiation monitoring according to claim 5, wherein the specific steps of the multi-level node growth are as follows:
step 21, the master node initiates a network growth request to a node L1i in the first intermediate node set { L1 };
step 22, the node L1i sends networking request information to the node U1i in the second intermediate node set { U1}, and starts to wait for receiving the reply message over time; if the received reply message is overtime, the current slave node to be networked fails to be networked, the step 24 is entered, and otherwise, the step 23 is entered;
step 23, receiving a reply message within a timeout period, and analyzing the reply message; if the analysis is unsuccessful, the current node to be networked fails to be networked, and the step 25 is entered, otherwise, the step 24 is entered;
step 24, if the communication link of the node U1i meets the growth policy, marking as a secondary slave node and in a network access state, otherwise, marking as a non-network access state, and entering step 25;
step 25, go on traversing the second intermediate node set { U1}, if the second intermediate node set { U1} is not traversed, search for the next node to be networked from the second intermediate node set { U1}, enter step 22, otherwise: if no network node is not accessed from the node set { S }, the secondary node growth is completed, all the nodes are completed, otherwise: continuing traversing the first intermediate node set { L1}, searching the next level node from the first intermediate node set { L1} if the traversing is not completed, and entering a step 21, otherwise: if the first intermediate node set { U1} has no non-network node, the second-level node growth is completed, all the nodes are completed, and if not, the third-level node growth process is entered.
7. The method for predicting photovoltaic second-level power based on irradiation monitoring according to claim 1, wherein outputting predicted values of irradiance and photovoltaic output at a future time of each photovoltaic monitoring terminal comprises:
carrying out interpolation smoothing processing on irradiance data of each photovoltaic monitoring terminal in combination with coordinate information of the irradiance data, fitting the irradiance data into a two-dimensional curved surface, obtaining irradiance data and photovoltaic output data at any position on the two-dimensional curved surface, and realizing prediction of photovoltaic output; wherein the predicted value of the photovoltaic outputThe calculation formula of (2) is as follows:
wherein X represents irradiance column vectors of all photovoltaic monitoring terminals; a is that t The influence matrix at the time t is represented, and represents irradiance of each photovoltaic monitoring terminal at the time t in the past, and influence degree of irradiance change in the future is represented; a is that t-j The influence matrix at the moment t-j is represented, j is automatically optimized according to the algorithm effect, namely irradiance at the moment j in the past has influence on future trend.
8. The photovoltaic second-level power prediction method based on irradiation monitoring is characterized by comprising the following steps of:
the photovoltaic monitoring terminal collects irradiance data around the photovoltaic power station and transmits the irradiance data to the photovoltaic monitoring base station; the photovoltaic monitoring terminals are arranged around the photovoltaic monitoring base station, and LoRa wireless communication is established between the photovoltaic monitoring terminals and the photovoltaic monitoring base station based on a networking algorithm of a maximum reliable path and an optimal path;
The photovoltaic monitoring terminal receives irradiance and a photovoltaic output predicted value at a certain moment in the future; the irradiance and photovoltaic output predicted value at a certain moment in the future is obtained by training an LSTM long-short-period neural network by a photovoltaic monitoring base station according to irradiance data around the photovoltaic power station, correcting by adopting the photovoltaic power station duration output data in the training process, iterating training until a network residual error enters a steady state or meets the precision requirement, and adopting the LSTM long-short-period neural network to output.
9. The method for predicting photovoltaic second level power based on irradiation monitoring of claim 8, wherein the plurality of photovoltaic monitoring terminals are disposed around the photovoltaic monitoring base station, comprising:
distributing a plurality of photovoltaic monitoring terminals according to a multi-tree network, taking the photovoltaic monitoring base station as a master node and the photovoltaic monitoring terminal as a slave node, and dividing the slave node into multiple levels of slave nodes according to different network levels;
the master node carries a scheduling task of a downlink communication network, and an intermediate slave node in the multi-stage slave nodes realizes ordered relay.
10. The photovoltaic second level power prediction method based on irradiation monitoring according to claim 8, wherein the networking algorithm based on the maximum reliable path and the optimal path comprises: adopting a multi-tree network growth strategy to perform multi-tree network growth on the nodes to be assembled to obtain communication links of the nodes to be assembled; the multi-tree network growth comprises one-level node growth and multi-level node growth;
The node to be assembled refers to a photovoltaic monitoring terminal which is not connected with the network;
the multi-tree network growth strategy comprises an optimal path growth strategy and a maximum reliable path growth strategy;
in the process of growing the optimal path growth strategy into the multi-way tree, the node to be networked continuously searches for a better path through a preferred algorithm until path traversal is completed;
in the process of the maximum reliable path growth strategy being the multi-way tree growth, the node to be networked continuously searches for a better path through a preferred algorithm, when the current communication link reaches the standard with good communication quality or above, the node stops growing, otherwise, the node is not grown until the path traversal is completed.
11. A photovoltaic second level power prediction system based on irradiance monitoring, the system comprising:
the photovoltaic monitoring base station is used for acquiring irradiance data around the photovoltaic power station from the photovoltaic monitoring terminal, training the LSTM long-short-period neural network according to the irradiance data around the photovoltaic power station, correcting the continuous output data of the photovoltaic power station in the training process, iterating the training until the network residual error enters a steady state or meets the precision requirement, outputting irradiance and photovoltaic output predicted values of each photovoltaic monitoring terminal at a certain moment in the future by adopting the LSTM long-period neural network, and transmitting the irradiance and the photovoltaic output predicted values to the corresponding photovoltaic monitoring terminal;
The photovoltaic monitoring terminal is used for collecting irradiance data around the photovoltaic power station and transmitting the irradiance data around the photovoltaic power station to the photovoltaic monitoring base station; receiving irradiance and photovoltaic output predicted values at a future moment transmitted by a photovoltaic monitoring base station;
the method comprises the steps that a plurality of photovoltaic monitoring terminals are arranged around a photovoltaic monitoring base station, and LoRa wireless communication is established between the photovoltaic monitoring base station and the photovoltaic monitoring terminals based on a networking algorithm of a maximum reliable path and an optimal path.
12. The photovoltaic second-level power prediction system based on irradiation monitoring according to claim 11, wherein a plurality of photovoltaic monitoring terminals are arranged around the photovoltaic monitoring base station, comprising:
distributing a plurality of photovoltaic monitoring terminals according to a multi-tree network, taking the photovoltaic monitoring base station as a master node and the photovoltaic monitoring terminal as a slave node, and dividing the slave node into multiple levels of slave nodes according to different network levels;
the master node carries a scheduling task of a downlink communication network, and an intermediate slave node in the multi-stage slave nodes realizes ordered relay.
13. The photovoltaic second level power prediction system based on irradiation monitoring according to claim 11, wherein the networking algorithm based on the maximum reliable path and the optimal path comprises:
Adopting a multi-tree network growth strategy to perform multi-tree network growth on the nodes to be assembled to obtain communication links of the nodes to be assembled; the multi-tree network growth comprises one-level node growth and multi-level node growth;
the node to be assembled refers to a photovoltaic monitoring terminal which is not connected with the network;
the multi-tree network growth strategy comprises an optimal path growth strategy and a maximum reliable path growth strategy;
in the process of growing the optimal path growth strategy into the multi-way tree, the node to be networked continuously searches for a better path through a preferred algorithm until path traversal is completed;
in the process of the maximum reliable path growth strategy being the multi-way tree growth, the node to be networked continuously searches for a better path through a preferred algorithm, when the current communication link reaches the standard with good communication quality or above, the node stops growing, otherwise, the node is not grown until the path traversal is completed.
14. A photovoltaic second level power prediction system based on irradiance monitoring of claim 11, wherein,
outputting irradiance and photovoltaic output predicted values of each photovoltaic monitoring terminal at a future moment, wherein the method comprises the following steps:
the irradiance data of each photovoltaic monitoring terminal is combined with the coordinate information to interpolate Smoothing, namely fitting the two-dimensional curved surface to obtain irradiance data and photovoltaic output data at any position on the two-dimensional curved surface, and predicting the photovoltaic output; wherein the predicted value of the photovoltaic outputThe calculation formula of (2) is as follows:
wherein X represents irradiance column vectors of all photovoltaic monitoring terminals; a is that t The influence matrix at the time t is represented, and represents irradiance of each photovoltaic monitoring terminal at the time t in the past, and influence degree of irradiance change in the future is represented; a is that t-j The influence matrix at the moment t-j is represented, j is automatically optimized according to the algorithm effect, namely irradiance at the moment j in the past has influence on future trend.
15. A computer readable storage medium storing a computer program, wherein the computer program when executed by a processor implements a photovoltaic second level power prediction method based on radiation monitoring as claimed in any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311310684.0A CN117439051A (en) | 2023-10-10 | 2023-10-10 | Photovoltaic second-level power prediction method, system and medium based on irradiation monitoring |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311310684.0A CN117439051A (en) | 2023-10-10 | 2023-10-10 | Photovoltaic second-level power prediction method, system and medium based on irradiation monitoring |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117439051A true CN117439051A (en) | 2024-01-23 |
Family
ID=89547227
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311310684.0A Pending CN117439051A (en) | 2023-10-10 | 2023-10-10 | Photovoltaic second-level power prediction method, system and medium based on irradiation monitoring |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117439051A (en) |
-
2023
- 2023-10-10 CN CN202311310684.0A patent/CN117439051A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2021238505A1 (en) | Federated learning-based regional photovoltaic power probability prediction method, and cooperative regulation and control system | |
| Rajanna et al. | Modeling of integrated renewable energy system for electrification of a remote area in India | |
| CN110068878B (en) | Optimal integrated forecasting method for intelligent temperature grid | |
| Saeed et al. | An optimal approach of wind power assessment using Chebyshev metric for determining the Weibull distribution parameters | |
| CN106532778B (en) | Method for calculating maximum access capacity of distributed photovoltaic grid connection | |
| Yang et al. | A novel ARX-based multi-scale spatio-temporal solar power forecast model | |
| US11264948B2 (en) | Techniques for forecasting solar power generation | |
| CN107437149B (en) | Method and system for determining output of photovoltaic power station | |
| Bosisio et al. | Improving DTR assessment by means of PCA applied to wind data | |
| CN108293193A (en) | A kind of efficient computational methods for completing drive test data | |
| CN106055816A (en) | Multi-factor photovoltaic equipment optimized laying method based on shadow radiation analysis | |
| KR20210066642A (en) | Consulting system through big data based prediction of solar power generation | |
| Shirbhate et al. | Time-series energy prediction using hidden Markov model for smart solar system | |
| CN105762843A (en) | Power output allocation method for non-storage type dispatchable photovoltaic power station | |
| CN117439051A (en) | Photovoltaic second-level power prediction method, system and medium based on irradiation monitoring | |
| US20230252384A1 (en) | Techniques for forecasting solar power generation | |
| Dong et al. | Very short-term photovoltaic power forecasting using uncertain basis function | |
| Silva et al. | Probabilistic models for one-day ahead solar irradiance forecasting in renewable energy applications | |
| CN103473476B (en) | The wind energy resources projectional technique of wind data is surveyed based on several anemometer towers in region | |
| CN104701841A (en) | Customer-side energy information sensing method of smart power grid | |
| CN107918920A (en) | The output correlation analysis method of more photovoltaic plants | |
| CN115239178A (en) | Rural low-voltage distribution type photovoltaic site selection and volume fixing method, system and medium | |
| Kou et al. | Sparse heteroscedastic Gaussian process for shortterm wind speed forecasting | |
| Ntlela et al. | Solar irradiation forecasting for the city of Durban using time series analysis | |
| CN112651540A (en) | Power distribution network planning project investment optimization 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 |