CN-122000870-A - Output prediction method of wind-light power generation system, wind-light hydrogen production scheduling method and system

CN122000870ACN 122000870 ACN122000870 ACN 122000870ACN-122000870-A

Abstract

The invention discloses an output prediction method of a wind-light power generation system, a wind-light hydrogen production scheduling method and a wind-light hydrogen production scheduling system, and relates to the technical field of renewable energy source production control, comprising the steps of obtaining wind-light prediction output data, preprocessing the wind-light prediction output data, and interpolating and complementing missing part data; the method comprises the steps of obtaining wind and light prediction output data in 0 to a hours in the future, recording the wind and light prediction output data as first data, obtaining wind and light prediction output data in a period of 0 to b hours in the future, recording the wind and light prediction output data as second data, obtaining historical wind and light prediction output data, correcting the second data to generate third data, combining the first data and the third data into final prediction output data, outputting the final prediction output data in real time, carrying out real-time rolling correction on the final prediction output data, and carrying out real-time rolling correction on the final prediction output data in a period of preset time intervals. And by carrying out real-time feedback correction on wind-solar prediction output data, a high-precision basis for continuous updating is provided for a downstream power utilization system.

Inventors

GE QI
SUO LIANGCHEN
ZHANG FAN
DENG YING
LIU JIAJIN
XU DEJING
XU HUIQING
SONG YU

Assignees

中煤(深圳)研究院有限责任公司

Dates

Publication Date: 20260508
Application Date: 20260109

Claims (10)

1. The output prediction method of the wind-solar power generation system is characterized by comprising the following steps of: A1, obtaining wind-solar prediction output data, preprocessing the wind-solar prediction output data, and removing abnormal data; a2, interpolating and complementing the missing part in the wind-solar prediction output data; A3, acquiring wind and light prediction output data in 0 to a hours in the future, recording the wind and light prediction output data as first data, acquiring the wind and light prediction output data in a to b hours in the future, recording the wind and light prediction output data as second data, and correcting the second data by acquiring historical wind and light prediction output data to generate third data; and A4, carrying out real-time rolling correction on the final predicted force data, wherein the real-time rolling correction is carried out by repeatedly executing the step A3 by taking a preset time interval as a period.
2. The method for predicting the output of a wind-light power generation system according to claim 1, wherein the wind-light predicted output data is expected power obtained for wind power generation and photovoltaic power generation, and comprises a plurality of data points which respectively correspond to the expected power at the appointed moment; the interpolation complement is expressed as: ; Wherein, the Representing the point in time corresponding to the missing data point, ; Indicating the expected power at the specified time.
3. The method of claim 1, wherein the correcting process is expressed as: ; Wherein, the Representing the wind-solar prediction output data, namely the second data, corresponding to the designated time t and before correction; predicting output data, namely third data, for the corrected wind and solar energy; Obtaining a prediction error for the wind-solar prediction output data according to the history; Is a correction coefficient, and The value of (2) is 0.7 to 1.0.
4. A method of predicting output of a wind-solar power generation system according to claim 3, wherein the prediction error is calculated as: ; Wherein, the The wind-solar predicted output data representing the history at a specified time t, Actual wind-light output data of histories at a designated time t is represented; And Historical data at the same moment; And And predicting output data for the wind and light at the same appointed moment in the respective period for different periods.
5. The method according to claim 4, wherein the final predicted force data is a plurality of discrete data points fitted according to the preset time interval within 0-b hours, each data point representing the wind-solar predicted force data at a specified time, the length of the preset time interval being an integer and being a common factor of a and b, and b being an integer and being a factor of the period.
6. The wind-solar hydrogen production scheduling method, which is applied to the output prediction method of the wind-solar power generation system according to any one of claims 1 to 5, is characterized by comprising the following steps: B1, acquiring wind power predicted power and photoelectric predicted power within 24 hours according to the final predicted force data; B2, calculating a reference load of the hydrogen production of the electrolytic cell according to the wind power predicted power, calculating a wind power average predicted value according to the reference load, and calculating a daily average value of the hydrogen production of the electrolytic cell according to the wind power average predicted value; B3, dividing the photoelectric predicted power into three stages according to the daily average value, wherein the stage is marked as a platform stage when the photoelectric predicted power is higher than the daily average value, and the stages before and after the platform stage are respectively marked as a load lifting stage and a load dropping stage in the same day; Opening a first number of electrolytic cells in a normalized manner according to the wind power predicted power, opening a second number of electrolytic cells in the platform stage, gradually increasing the first number of electrolytic cells according to the increase of the photoelectric predicted power in the load lifting stage, and gradually closing the second number of electrolytic cells according to the decrease of the photoelectric predicted power in the load reducing stage; Converting the reference load into reference hydrogen production, converting the wind power predicted power into wind power hydrogen production, comparing the difference between the reference hydrogen production and the wind power hydrogen production, and obtaining the maximum value of the difference as a wind power hydrogen storage demand space; And B5, obtaining the effective capacity of the hydrogen storage tank, subtracting the maximum wind power hydrogen storage demand space to be used as a photoelectric hydrogen storage space, converting the photoelectric hydrogen storage space into a power value, and taking the part of the photoelectric predicted power exceeding the power value as the Internet surfing electric quantity.
7. The method for scheduling hydrogen production from wind and solar energy of claim 6, wherein the reference load The calculation of (2) is expressed as: n represents the number of specified moments in 24 hours, Representing the wind power predicted power corresponding to the ith appointed moment; The wind power average predicted value The calculation of (2) is expressed as: ; the daily average value The calculation of (2) is expressed as: ; Represents the energy consumption of the electrolytic cell for hydrogen production.
8. The wind-solar hydrogen production scheduling method according to claim 6, further comprising B6, performing power distribution for each of the electrolytic cells: B61, acquiring the first data, the inventory data of the hydrogen storage tank, the real-time power constraint information of the power grid and the operation condition of hydrogen utilization equipment, and constructing an optimization model; b62, collecting wind and light real-time output data, calculating the deviation between the wind and light predicted output data, and performing feedforward-feedback compensation on the total power data to obtain real-time total power; B63, distributing the real-time total power to each electrolytic cell according to the load-efficiency characteristic data of the electrolytic cells and with the lowest total hydrogen production and electricity consumption as a target, and generating an optimal power instruction for each electrolytic cell; and B64, repeating the steps B61 to B63 with the preset time interval as a period.
9. The wind-solar hydrogen production scheduling method according to claim 8, wherein in B63, the target is represented as: The constraint of the target comprises: ; wherein M represents the number of the electrolytic cells, and j represents the jth electrolytic cell; indicated as power allocated to the jth cell; Representing the real-time total power; And The lower limit and the upper limit of the operation power of the jth electrolytic tank are respectively set; indicating the power variation of the jth electrolytic tank; Indicating the maximum power change rate of the jth electrolytic cell; Load-efficiency characteristic data of the j-th electrolytic cell are shown.
10. A system, which applies the method for predicting the output of the wind-solar power generation system according to any one of claims 1 to 5, is characterized by comprising a meteorological data acquisition module, a prediction module, a power generation module, a prediction data processing module, a scheduling module and an electricity utilization module; The wind-solar prediction output data are generated by the prediction module based on the meteorological data and combined with historical meteorological data and historical power generation of the power generation module, the prediction data processing module is used for executing steps A1 to A4 to generate final prediction output data, the scheduling module generates a power scheduling plan for the power utilization module according to the final prediction output data, and the power utilization module executes power utilization according to the power scheduling plan.

Description

Output prediction method of wind-light power generation system, wind-light hydrogen production scheduling method and system Technical Field The invention relates to the technical field of renewable energy source production control, in particular to a method for predicting output of a wind-light power generation system, a method for scheduling wind-light hydrogen production and a system thereof. Background With the transition of the global energy structure to clean low carbon, renewable energy power generation technologies represented by wind energy and photovoltaic are rapidly developed. The green electricity is utilized to prepare green hydrogen through an electrolytic water process, and is one of key technical paths for realizing large-scale long-time storage and cross-region absorption of renewable energy, and the technology can convert electric energy into hydrogen energy, further serve as fuel or industrial raw materials, be applied to various fields of transportation, chemical industry, metallurgy and the like, and effectively promote deep decarburization. At present, the wind-light power generation hydrogen production system mainly comprises modes of wind-light part surfing, residual electricity hydrogen production, wind-light total off-grid independent hydrogen production and the like. Despite the broad prospect, the current wind-solar coupled hydrogen production technology still faces a series of technical bottlenecks to be solved in the aspect of large-scale and efficient application. First, wind and solar energy have natural volatility, intermittence and randomness, and the generated power is obviously affected by weather and weather. It is particularly pointed out that the output of photovoltaic power generation exhibits a strong regular spike in the day, starting from sunrise, the power rises, peaks in the middle of the day, and then drops rapidly to zero before sunset. The large drop of the daily power curve is overlapped with the wind power output which is relatively more likely to be continuous, so that the extremely complex and greatly-fluctuated hybrid energy input is formed. There is a fundamental conflict between this unstable power output and the continuous and smooth operation required for downstream chemical production (e.g., ammonia synthesis, methanol). Frequent power fluctuation, especially regular peak-to-valley variation from day to day, can force the hydrogen production electrolytic tank to be in an unstable working interval of repeated start-stop and frequent load lifting, can obviously reduce the electro-hydrogen conversion efficiency and the service life of equipment, and can bring serious safety challenges. At present, in order to ensure the stability of downstream production, a system is seriously dependent on high-precision prediction of wind-solar power generation power so as to make a production plan in advance. The current mainstream prediction method is highly dependent on historical weather and power data, and a prediction curve is generated through numerical weather prediction and a statistical model. However, the method has inherent limitations that on one hand, the historical data model is difficult to completely capture sudden changes of a weather system, so that a reference error exists in a prediction result, and on the other hand, the system generally lacks an efficient and automatic real-time monitoring and feedback correction mechanism, so that the prediction cannot be dynamically calibrated and corrected in time according to actual weather mutation (such as cloud cluster rapid movement) within an ultra-short time scale (such as 15 minutes to 1 hour in the future). In addition, the production scheduling strategies of most of the current systems are relatively extensive, and self-adaptive optimization cannot be performed on the dynamic characteristics of wind and light, particularly photovoltaic output, in different time scales in the day. For example, a differential hydrogen production operation plan cannot be dynamically formulated in coordination with wind power output according to a typical light Fu Gonglv curve of 'climbing in the morning-peak in the noon-declining in the afternoon'. This makes it difficult to match the supply of energy and the load of production, which often differ greatly from the actual production schedule formulated based on the predictions. Therefore, from the perspective of economy and energy efficiency of system operation, the problems directly result in two concurrent adverse phenomena, namely that on one hand, when wind and light output is abundant, the hydrogen production system cannot be completely consumed or cannot be immediately utilized downstream, so that 'waste wind and light' are abandoned, green electric power is wasted, and on the other hand, when wind and light output is insufficient or fluctuation is severe, stable and sufficient hydrogen cannot be provided for downstream, so that 'insufficient hydrogen supply' is ca