Method and system for predicting generating capacity of small hydropower station group

Abstract

The invention discloses a small hydropower station group generating capacity prediction method and system, and the method comprises the steps: analyzing historical rainfall data and historical runoff data, and constructing a rainfall runoff simulation initial model; analyzing the soil moisture characteristic curve to obtain a soil moisture infiltration rate influence factor; obtaining a soil moisture infiltration rate influence factor; obtaining a soil pore structure influence factor; correcting the rainfall runoff simulation initial model to obtain a rainfall runoff simulation prediction model; constructing a small hydropower station group output load prediction model through multiple regression analysis; and obtaining a generating capacity prediction result based on an output result of the small hydropower station group output load prediction model and the power dispatching operation data of the target small hydropower station group. According to the method for predicting the generating capacity of the small hydropower station group, the runoff convergence phenomenon generated under the rainfall condition is intelligently analyzed, a more accurate generating capacity prediction result can be obtained, and then accurate control over the small hydropower station group is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of power forecasting, and in particular to a method and system for forecasting power generation of a small hydropower group. Background Art

[0002] As a kind of renewable energy, small hydropower does not consume fossil fuels and does not produce greenhouse gas emissions during its power generation process, which helps reduce environmental pollution and global warming. Small hydropower group power generation can make fuller use of water resources and improve the utilization rate of renewable energy.

[0003] In order to improve the power supply conditions in remote mountainous areas or rural areas, small hydropower stations are often distributed in mountainous and hilly areas. The topography of these areas is complex and changeable, and it is easy to generate runoff during rainfall. In the prior art, the power generation prediction of small hydropower stations in mountainous and hilly areas only considers the impact of rainfall on the reservoir of the hydropower station, and ignores the runoff generated by the mountainous and hilly areas around the small hydropower station group, which leads to inaccurate prediction results of the power generation of small hydropower stations.

[0004] It can be seen that how to improve the accuracy of power generation prediction for small hydropower groups has become a technical problem that needs to be urgently solved by those skilled in the art. Summary of the invention

[0005] The present invention provides a method and system for predicting power generation of a small hydropower group. By intelligently analyzing the runoff inflow phenomenon generated under rainfall conditions, more accurate power generation prediction results can be obtained, thereby improving the precise control of the small hydropower group.

[0006] In order to solve the above technical problems, an embodiment of the present invention provides a method for predicting power generation of a small hydropower group, comprising: Based on the time series analysis method, the historical rainfall data and historical runoff data of the target area adjacent to the target small hydropower group are analyzed, and the initial rainfall-runoff simulation model is constructed according to the rainfall-runoff relationship fitting curve obtained by the analysis; Determine a soil moisture characteristic curve based on the soil geological data of the target area and the collected soil moisture sample data, and obtain a soil moisture infiltration rate influencing factor based on an analysis of the soil moisture characteristic curve; Obtaining the soil infiltration rate of the target area under different rainfall intensities, and obtaining the influencing factor of the soil moisture infiltration rate; Based on the time series analysis method, the pore change data of the target area under different rainfall intensities are analyzed to obtain the soil pore structure influencing factors; The initial rainfall-runoff simulation model is modified by the soil moisture infiltration rate influencing factor and the soil moisture infiltration rate influencing factor to obtain a rainfall-runoff simulation prediction model; based on the rainfall-runoff relationship fitting curve, the soil moisture characteristic curve and the soil pore structure influencing factor, a small hydropower group output load prediction model is constructed through multivariate regression analysis; The real-time data output by the rainfall runoff simulation prediction model is input into the small hydropower group output load prediction model, and the corresponding power generation prediction result is obtained based on the output result and the power dispatching operation data of the target small hydropower group.

[0007] As one preferred solution, the time series analysis method is used to analyze the historical rainfall data and historical runoff data of the target area adjacent to the target small hydropower group, including: Acquire historical rainfall data and corresponding original water level and flow change data of the target area, and obtain historical runoff data according to the relationship between the historical rainfall data and the original water level and flow change data; According to the surface coverage parameters of the target area, the underlying surface of the target area is subjected to rasterization processing, and the watershed confluence time is obtained from the rasterization processing result; Using a sliding time window to perform time series segmentation on the historical rainfall data and the historical runoff data, obtaining corresponding correlation values ​​according to the segmentation results, and determining the correlation values ​​that meet the conditions as rainfall-runoff conversion parameters; Based on the radial basis kernel function, a corresponding rule between the historical rainfall data and the historical runoff data is constructed, and a quadratic programming analysis is performed on the corresponding rule according to the rainfall-runoff conversion parameter to obtain a rainfall-runoff relationship curve.

[0008] As one of the preferred solutions, the soil moisture characteristic curve is determined according to the soil geological data of the target area and the collected soil moisture sample data, and the soil moisture infiltration rate influencing factor is obtained based on the analysis of the soil moisture characteristic curve, including: Correcting the soil moisture data collected by the soil moisture sensor to obtain soil moisture sample data; Based on the digital elevation model technology, the soil geological data of the target area is processed to obtain a soil moisture change curve, wherein the change area of ​​the soil moisture change curve is obtained by partitioning the target area by a rasterization method; Using a long short-term memory network to perform time series fitting analysis on the soil moisture change curve to obtain soil moisture characteristics, wherein the types of the soil moisture characteristics at least include hydraulic conductivity, capillary potential and water retention curve; The soil moisture variation curve is analyzed according to the soil moisture characteristics to obtain a soil moisture characteristic curve, and a soil moisture infiltration rate influencing factor is obtained from the coupling relationship among the hydraulic conductivity, capillary potential and water retention curve.

[0009] As one preferred solution, the soil moisture data collected by the soil moisture sensor is corrected to obtain soil moisture sample data, including: The soil moisture data is corrected according to the temperature data synchronously collected by the soil moisture sensor. If the corrected soil moisture content is within the saturated moisture content range and the rate of change between adjacent moments is less than a preset threshold, the corresponding data is determined as the soil moisture sample data.

[0010] As one of the preferred solutions, the step of obtaining the soil infiltration rate of the target area under different rainfall intensities and obtaining the soil moisture infiltration rate influencing factor includes: Under different rainfall intensities, the soil infiltration rate collected by the double-ring infiltrator was obtained, and the soil bulk density was measured using a ring knife; The soil porosity of the target area is measured based on a capillary method, and a soil structure stability index is calculated according to the soil bulk density value and the soil porosity, wherein the soil structure stability index reflects the degree of soil structure damage; Based on the soil structure stability index and the soil infiltration rate, an infiltration rate mapping relationship is established through a Gaussian kernel function; The infiltration rate mapping relationship and the rainfall intensity are analyzed to obtain a soil moisture infiltration rate influencing factor.

[0011] As one of the preferred solutions, the time series analysis method is used to analyze the pore change data of the target area under different rainfall intensities to obtain the soil pore structure influencing factors, including: Under different rainfall intensities, the surface soil feature points of the target area are obtained from the soil surface morphology collector, and the soil pore dynamic change data are obtained according to the matching of the surface soil feature points; and the soil moisture content data of the target area is collected from the soil moisture sensor; The soil pore dynamic change data and the soil water content data are time-series fitted using a long short-term memory network to obtain the soil pore structure influencing factor of the target area.

[0012] As one preferred solution, before inputting the real-time data output by the rainfall runoff simulation prediction model into the small hydropower group output load prediction model, the method further includes: Acquire real-time rainfall data of the target area monitored in real time; Preprocessing the real-time rainfall data, wherein the preprocessing at least includes mean filling the first abnormal missing data and / or deleting the second abnormal missing data; The pre-processed real-time rainfall data is input into the rainfall runoff simulation prediction model to obtain the real-time data output by the rainfall runoff simulation prediction model.

[0013] As one of the preferred solutions, the power dispatching operation data at least includes reservoir load data, grid load side demand data and hydropower generation cluster operation data; The corresponding power generation prediction result is obtained based on the output result and the power dispatching operation data of the target small hydropower group, including: Constructing the operation rules of the target small hydropower group according to the grid load side demand data and the hydropower generation cluster operation data; The power generation data to be corrected obtained from the output result and the reservoir load data is corrected based on the operation rule to obtain a corresponding power generation prediction result.

[0014] As one preferred solution, after obtaining the corresponding power generation prediction result, the method further includes: The generated reservoir control instructions matching the power generation prediction results are sent to the corresponding small hydropower group control terminal, wherein the reservoir control instructions at least include gate linkage control instructions, flood control instructions, drought control instructions and reservoir water level regulation instructions.

[0015] Another embodiment of the present invention provides a small hydropower group power generation prediction system, comprising: The initial model module is used to analyze the historical rainfall data and historical runoff data of the target area adjacent to the target small hydropower group based on the time series analysis method, and to construct an initial rainfall-runoff simulation model based on the rainfall-runoff relationship fitting curve obtained by the analysis; An infiltration rate influencing factor module is used to determine a soil moisture characteristic curve according to the soil geological data of the target area and the collected soil moisture sample data, and obtain a soil moisture infiltration rate influencing factor based on an analysis of the soil moisture characteristic curve; The infiltration rate influencing factor module is used to obtain the soil infiltration rate of the target area under different rainfall intensities and obtain the soil moisture infiltration rate influencing factor; The pore structure influencing factor module is used to analyze the pore change data of the target area under different rainfall intensities based on the time series analysis method to obtain the soil pore structure influencing factor; A load prediction model module is used to modify the initial rainfall runoff simulation model with the soil moisture infiltration rate influencing factor and the soil moisture infiltration rate influencing factor to obtain a rainfall runoff simulation prediction model; based on the rainfall runoff relationship fitting curve, the soil moisture characteristic curve and the soil pore structure influencing factor, a small hydropower group output load prediction model is constructed through multivariate regression analysis; The prediction module is used to input the real-time data output by the rainfall runoff simulation prediction model into the small hydropower group output load prediction model, and obtain the corresponding power generation prediction result based on the output result and the power dispatching operation data of the target small hydropower group.

[0016] Compared with the prior art, the embodiments of the present invention have the following advantages: (1) First, an initial rainfall-runoff simulation model is constructed based on historical data. Then, the factors affecting the infiltration rate and the infiltration rate under rainfall conditions are analyzed respectively, and the constructed initial model is corrected to obtain a more accurate rainfall-runoff simulation prediction model, thereby providing an accurate model support for the subsequent power generation calculation; (2) Through intelligent analysis of the runoff inflow phenomenon generated under rainfall conditions, more accurate power generation prediction results can be obtained. On the one hand, it helps the power grid dispatching center to better arrange power production and transmission, and reduce power grid fluctuations and load imbalances caused by inaccurate predictions. On the other hand, accurate prediction of the power generation capacity of small hydropower groups can help hydropower stations formulate more reasonable power generation plans, avoid waste of water resources and idle power generation equipment, and provide accurate data support for subsequent power generation control and carbon emission management, thus promoting the intelligent control process of small hydropower groups. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] FIG1 is a schematic flow chart of a method for predicting power generation of a small hydropower group in one embodiment of the present invention; FIG2 is a schematic diagram of the structure of a small hydropower group power generation prediction system in one embodiment of the present invention; Reference numerals: Among them, 11, initial model module; 12, infiltration rate influencing factor module; 13, infiltration rate influencing factor module; 14, pore structure influencing factor module; 15, load prediction model module; 16, prediction module. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] In the description of this application, the terms "first", "second", "third", etc. are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first", "second", "third", etc. may explicitly or implicitly include one or more of the feature. In the description of this application, unless otherwise specified, "plurality" means two or more.

[0020] In the description of the present application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be a connection between the two elements. The terms "vertical", "horizontal", "left", "right", "upper", "lower" and similar expressions used herein are only for illustrative purposes, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0021] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in the present invention have the same meaning as those commonly understood by those skilled in the art. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood by specific circumstances.

[0022] An embodiment of the present invention provides a method for predicting power generation of a small hydropower group. Specifically, please refer to FIG. 1 , which is a flow chart of a method for predicting power generation of a small hydropower group in one embodiment of the present invention, which includes steps S1 to S6: S1. Based on the time series analysis method, the historical rainfall data and historical runoff data of the target area adjacent to the target small hydropower group are analyzed, and the initial rainfall-runoff simulation model is constructed according to the rainfall-runoff relationship fitting curve obtained by the analysis; S2. Determine a soil moisture characteristic curve according to the soil geological data of the target area and the collected soil moisture sample data, and obtain a soil moisture infiltration rate influencing factor based on an analysis of the soil moisture characteristic curve; S3, obtaining the soil infiltration rate of the target area under different rainfall intensities, and obtaining the influencing factor of the soil moisture infiltration rate; S4. Based on the time series analysis method, the pore change data of the target area under different rainfall intensities are analyzed to obtain the soil pore structure influencing factor; S5. The initial rainfall-runoff simulation model is modified by the soil moisture infiltration rate influencing factor and the soil moisture infiltration rate influencing factor to obtain a rainfall-runoff simulation prediction model; based on the rainfall-runoff relationship fitting curve, the soil moisture characteristic curve and the soil pore structure influencing factor, a small hydropower group output load prediction model is constructed through multivariate regression analysis; S6. Input the real-time data output by the rainfall runoff simulation prediction model into the small hydropower group output load prediction model, and obtain the corresponding power generation prediction result based on the output result and the power dispatching operation data of the target small hydropower group.

[0023] It should be noted in advance that the small hydropower group in the embodiment of the present invention relies on the water storage tank of the reservoir during the power generation process, which is no different from the prior art. However, the present invention comprehensively considers the influence of runoff inflow under rainfall conditions. Therefore, the present invention also conducts geological analysis on the mountainous and hilly areas around the small hydropower group. Preferably, the relevant geological data introduced in the embodiment of this article is obtained by pre-set sensors, such as soil moisture sensors, etc., and their distribution positions and quantities are not specifically limited in the embodiment of the present invention. In addition, the embodiment of the present invention uses a rainfall monitoring device to obtain original rainfall data, and obtains original water level data through a water level monitoring device. In addition, the original rainfall and water level data can also be collected according to a fixed period through the automatic monitoring devices of the rain gauges and hydrological stations around the target area, and the collected original monitoring data are verified for rainfall data range and rainfall continuity. The water level data after inspection is converted into runoff data through the water level flow relationship curve, and the real-time rainfall and runoff data are calculated from the monitoring data that has passed the quality inspection. It will not be explained later.

[0024] In addition, in the above step S1, the embodiment of this invention first constructs an initial model for rainfall runoff simulation. This initial model is constructed based on historical data, so it is difficult to accurately reflect the actual runoff simulation prediction situation. In view of this, the embodiment of the present invention modifies the above initial model from two dimensions, namely, the soil moisture infiltration rate influencing factor and the soil moisture infiltration rate influencing factor. Soil moisture infiltration rate and soil moisture infiltration rate are two important parameters that describe the soil moisture penetration characteristics, which are closely related to the movement process of soil moisture. Soil moisture infiltration rate, also known as infiltration rate, refers to the intensity of water penetration in the soil, usually expressed in several millimeters per hour (mm / h). It reflects the amount of water that penetrates into the soil through the infiltration surface per unit area per unit time, focusing on the penetration capacity in the soil. Soil moisture infiltration rate, also known as soil infiltration rate or soil infiltration rate, refers to the amount of infiltration water per unit area of ​​soil on the surface per unit time. It is also expressed in several millimeters per hour (mm / h), but it focuses more on describing the absorption capacity of the soil surface for water, which is described in detail below.

[0025] For the above step S1, the above content explains how to obtain rainfall data. In order to obtain historical runoff data, historical runoff data can be obtained based on the relationship between historical rainfall data and original water level flow change data. A rainfall runoff correlation element acquisition device is used to measure vegetation coverage, soil permeability, and surface slope for the surface of the study area. The collected surface parameters are divided into vegetation coverage classification data, soil type classification data, and slope classification data according to the rasterization classification standard. Each type of underlying surface is rasterized according to the classification data and the infiltration rate and water storage capacity are calculated. The watershed confluence time is obtained from the calculation results of each grid unit. Preferably, a sliding time window based on 24 hours is used to segment the rainfall and runoff data sequence into time series at the annual scale, monthly scale, daily scale, and hourly scale. A long short-term memory network is used to calculate the correlation value of the segmented rainfall runoff data. If the correlation value exceeds 0.8, it is determined that there is a rainfall runoff correspondence in the period, and rainfall runoff conversion parameters of different time scales are obtained from the corresponding relationship data. The correspondence rules between rainfall and runoff values ​​(that is, historical rainfall data and historical runoff data) are constructed through the radial basis kernel function. The numerical correspondence rules are solved by quadratic programming according to the rainfall-runoff conversion parameters at the annual, monthly, daily and hourly scales. The rainfall-runoff numerical conversion rules are obtained from the solution results, and the rainfall-runoff relationship curves at different time scales are drawn according to the conversion rules.

[0026] In the above embodiments, the rainfall and water level data collection will be affected by environmental factors and produce abnormal values, such as negative or excessive rainfall, sudden changes or cliff-like drops in water level, and abnormal data will be eliminated through data quality inspection. The threshold interval for rainfall data range verification is set to 0-250 mm / hour, and data outside this range is marked as abnormal. The rainfall continuity test compares the rainfall change rate at adjacent moments, and data points with a change rate of more than 80% are marked as abnormal. The water level-flow relationship curve is established based on the cross-sectional measurement data, and the corresponding flow value is recorded for every 0.1 meter increase in water level to form a water level-flow correspondence table. Surface parameter collection targets three aspects: vegetation coverage, soil permeability, and surface slope. Vegetation coverage is obtained by remote sensing image interpretation, soil permeability is measured in situ by a permeameter, and surface slope is calculated using digital elevation data. Vegetation coverage is divided into five levels: 0-20% is bare land, 20-40% is sparse vegetation, 40-60% is medium vegetation, 60-80% is dense vegetation, and 80-100% is dense vegetation. Soil types are divided into three levels according to particle size: sand, loam, and clay. Slopes are divided into four levels: 0-5 degrees for flat land, 5-15 degrees for gentle slopes, 15-25 degrees for steep slopes, and more than 25 degrees for steep slopes. The rainfall runoff data series is divided according to the time scale. The annual scale is based on the hydrological year, the monthly scale is based on the natural month, the daily scale is based on 24 hours, and the hourly scale is based on 1 hour. The long short-term memory network receives rainfall data through the input layer, the forgetting gate controls the proportion of historical information retention, the input gate determines the proportion of new information reception, and the output gate generates the current moment runoff prediction value. The network parameters are optimized through back propagation. The radial basis kernel function maps rainfall and runoff to a high-dimensional feature space and constructs a linear classification hyperplane in the feature space. The kernel function parameters were determined by cross-validation, and the kernel width of the Gaussian kernel function was set to 0.8. In the process of solving the quadratic programming, the slack variable tolerance was set to 0.01, and the maximum number of iterations was 1000. The rainfall-runoff relationship curve shows different characteristics at different time scales. The annual scale curve reflects the overall runoff characteristics of the basin, the monthly scale curve reflects the seasonal variation law, and the daily and hourly scale curves depict the runoff characteristics of a single rainfall process. The infiltration rate is calculated based on the vegetation coverage and soil type. The infiltration rate increases by 15% for every 20% increase in vegetation coverage. The infiltration rate of sandy soil is 15 mm / hour, 8 mm / hour for loam, and 3 mm / hour for clay. The water storage capacity is related to the vegetation coverage and slope. The water storage capacity increases by 12% for every 20% increase in vegetation coverage, and the water storage capacity decreases by 8% for every 10-degree increase in slope. The confluence time is affected by the slope and the roughness of the underlying surface. The confluence time is shortened by 10% for every 5-degree increase in slope, and the confluence time is extended by 15% for every 20% increase in vegetation coverage.

[0027] For the above step S2, this embodiment is to obtain the soil moisture data collected by the soil moisture sensor. Preferably, the soil moisture data needs to be corrected. The specific method is: the soil moisture is corrected according to the temperature data synchronously collected by the soil moisture sensor. If the corrected soil moisture is within the saturated moisture content range and the change rate at adjacent moments is less than the preset threshold, the measured soil moisture (i.e., soil moisture sample data) is obtained. Then, based on the digital elevation model technology, the slope gradient value and the slope aspect azimuth are calculated according to the terrain digital elevation data, and the slope gradient and the slope aspect azimuth are partitioned by the rasterization method, and the soil moisture change curve is obtained for each slope and aspect combination. Then, the long short-term memory network is used to perform time series fitting on the soil moisture change curve. If the soil moisture content is less than the field water holding capacity, it is judged that the soil moisture is in the infiltration and replenishment process, and the soil moisture characteristic value is calculated by the soil temperature and porosity. Finally, a Gaussian radial basis function is established to analyze the soil moisture change curve according to the soil moisture characteristics to obtain the soil moisture characteristic curve, and the influencing factor of soil moisture infiltration rate is obtained from the coupling relationship among hydraulic conductivity, capillary potential and water retention curve.

[0028] Specifically, in the above embodiment, real-time soil moisture data is collected by soil moisture sensors arranged every 50 meters. The soil moisture sensors simultaneously collect soil temperature data and surface temperature and humidity data. The soil moisture content is corrected according to the temperature coefficient. The corrected soil moisture content is subjected to a value range test and a continuity test. The value range is between 0 and the saturated moisture content. The moisture content change rate at adjacent moments is less than 20%. The measured soil moisture content is obtained from the qualified test data. The slope gradient value and slope aspect azimuth of the study area were calculated based on the terrain digital elevation data. The rasterization method was used to divide the slope gradient into a gentle slope area soil moisture dataset according to 0-15 degrees, a medium slope area soil moisture dataset according to 15-25 degrees, and a steep slope area soil moisture dataset according to 25 degrees or more. The slope aspect azimuth was divided into east slope, south slope, west slope, and north slope soil moisture datasets according to 0-90 degrees, 90-180 degrees, 180-270 degrees, and 270-360 degrees. The soil moisture change curve was obtained for each slope aspect combination based on soil texture particle size classification. The number of hidden layer units of the long short-term memory network is 128, and the threshold of the forgetting gate is 0.6. The soil moisture change curve is time-series fitted, and the dynamic change law of different slopes and slope directions is calculated by the soil moisture change rate. If the soil moisture content is less than the field water holding capacity of 0.3, it is judged that the soil moisture is in the process of infiltration and recharge. The soil moisture characteristic values ​​composed of saturated moisture content, residual moisture content, and capillary rise height are calculated according to soil temperature and porosity. The Gaussian radial basis function is used to establish the mapping relationship between soil moisture content and infiltration rate. The kernel function parameters are determined by cross-validation. The soil moisture migration process is described according to the soil moisture characteristic value. The maximum soil moisture migration rate is characterized by saturated hydraulic conductivity, the capillary potential is characterized by the driving force of soil moisture migration, and the water retention curve is characterized by the soil water storage performance. The soil moisture migration law under different slope and aspect conditions is obtained from the coupling relationship between the three. The layout of soil moisture sensors takes into account the terrain characteristics and monitoring accuracy requirements. A monitoring point is set up every 50 meters in the study area. The altitude difference of the monitoring point is within 300 meters. The soil temperature and moisture content show a negative correlation. For every 1 degree Celsius increase in temperature, the soil moisture content decreases by 0.2%. When the soil temperature increases from 15 degrees Celsius to 25 degrees Celsius, the moisture content decreases from 28% to 26%. The slope gradient division is based on the critical slope of soil erosion. The soil erosion intensity in the gentle slope area of ​​0-15 degrees is small, and the fluctuation range of soil moisture content is less than 5%. The soil erosion intensity in the medium slope area of ​​15-25 degrees is moderate, and the fluctuation range of soil moisture content is between 5%-10%. The soil erosion intensity in the steep slope area above 25 degrees is large, and the fluctuation range of soil moisture content exceeds 10%. The influence of slope aspect angle on soil moisture is reflected in the amount of solar radiation. The east slope has sufficient sunshine in the morning, the south slope has long sunshine throughout the day, the west slope has strong sunshine in the afternoon, and the north slope has short sunshine.Soil texture is divided into sand, loam and clay according to particle size. Sand particle size is greater than 0.05 mm, porosity is between 35%-45%, water holding capacity is poor but permeability is good. Loam particle size is between 0.002-0.05 mm, porosity is between 45%-55%, water holding capacity and permeability are medium. Clay particle size is less than 0.002 mm, porosity is between 55%-65%, water holding capacity is good but permeability is poor. Soil water migration is affected by multiple factors. Saturated hydraulic conductivity reflects soil water conductivity. The saturated hydraulic conductivity of sand is 25 mm / hour, loam is 12 mm / hour, and clay is 4 mm / hour. Capillary potential represents the driving force of soil water migration, which decreases with increasing water content. When the water content increases from 10% to 30%, the capillary potential decreases from 800 cm of water column to 200 cm of water column. The water retention curve describes the characteristics of soil moisture. During the moisture absorption process, the water content increases with the increase of matrix potential, and during the dehumidification process, the water content decreases with the decrease of matrix potential, forming a hysteresis loop. The long short-term memory network controls the flow of information through a gating mechanism. The forgetting gate threshold of 0.6 means that 60% of the historical information is retained. The input gate and output gate are adaptively adjusted according to the current state. The kernel width of the Gaussian radial basis function is determined to be 0.8 through cross-validation. The kernel function maps the input space to the feature space and constructs a nonlinear classification boundary in the feature space. The soil moisture characteristics include three indicators: saturated moisture content, residual moisture content, and capillary rise height. The three together reflect the factors affecting the soil moisture infiltration rate.

[0029] For the above step S3, the embodiment of the present invention is to obtain the soil infiltration rate collected by the double-ring infiltration instrument under different rainfall intensities, and use a ring knife to measure the soil bulk density value. The soil porosity of the target area is measured based on the capillary method, and the soil structure stability index is calculated according to the soil bulk density value and the soil porosity. The soil structure stability index reflects the degree of soil structure damage. Exemplarily, the soil infiltration rate is monitored in real time by a double-ring infiltration instrument arranged on the surface, and the soil bulk density data is collected every 1 hour by a ring knife. The soil porosity is measured every 1 hour by the capillary method. The soil texture is divided into sandy soil with a sand content greater than 50%, clay soil with a clay content greater than 30%, and loam soil between the two according to the particle size distribution data measured by the laser particle size analyzer. The soil structure stability index is obtained from the classification data and aggregate analysis. Then, based on the soil structure stability index and the soil infiltration rate, the infiltration rate mapping relationship is established through the Gaussian kernel function, and the infiltration rate mapping relationship and the rainfall intensity are analyzed to obtain the soil water infiltration rate influencing factor.

[0030] Preferably, a long short-term memory network is used to perform time series fitting on the soil infiltration rate, wherein the number of hidden layer units is 128, the forgetting gate threshold is 0.6, and the scouring erosion intensity is calculated by the soil bulk density change rate. If the scouring depth exceeds 5 mm or the aggregate destruction rate exceeds 30%, it is determined to be soil structure damage. According to the degree of damage, the change level of soil infiltration capacity is divided into mild damage, moderate damage, and severe damage. A Gaussian kernel function is used to establish a mapping relationship between soil structure and infiltration rate, and the kernel parameter value is 0.8. According to the degree of soil structure damage, the infiltration rate is divided into fast infiltration greater than 50 mm / hour, medium infiltration 20-50 mm / hour, and slow infiltration less than 20 mm / hour. The soil infiltration law is obtained from the infiltration rate change curve under the combined conditions of rainfall intensity, runoff, and soil texture. The double-ring infiltration instrument consists of two concentric cylinders inside and outside, with an inner ring diameter of 20 cm and an outer ring diameter of 40 cm. The two rings are simultaneously filled with water to maintain a constant head, and the infiltration rate of the inner ring is measured. The outer ring is used to control lateral seepage. Soil bulk density reflects the compactness of the soil. When the bulk density of sand exceeds 1.6 g / cm3, the bulk density of loam exceeds 1.4 g / cm3, and the bulk density of clay exceeds 1.2 g / cm3, it indicates that the soil has been compacted. Soil porosity is measured by the capillary method. The soil sample is placed in a capillary and the height of water rise is measured. The porosity of sand is between 35%-45%, loam is between 45%-55%, and clay is between 55%-65%. Soil texture classification is based on particle size composition. The soil particle size distribution is measured by a laser particle size analyzer. The diameter of sand particles is greater than 0.05 mm, the diameter of clay particles is less than 0.002 mm, and the diameter of powder particles is between the two. Aggregate stability is measured by wet sieving. The soil sample is placed in water and shaken to measure the changes in the content of aggregates of different particle sizes. Aggregate destruction rate exceeding 30% indicates that the soil structure is unstable. The soil infiltration rate is stable under light rain intensity, and the infiltration rate of sand is maintained at 80 mm / h, loam is maintained at 40 mm / h, and clay is maintained at 15 mm / h. Under heavy rain conditions, raindrops splash and destroy soil aggregates, forming a surface sealing layer, and the infiltration rate drops significantly, dropping to 30 mm / h for sand, 15 mm / h for loam, and 5 mm / h for clay. The scouring effect of surface runoff on soil increases with the increase of flow velocity. When the flow velocity reaches 0.5 m / s, rill erosion begins to appear on the surface of the soil. As the scouring continues, the rills gradually deepen and widen. Soil runoff mainly occurs in soil pores, with a low flow rate and less damage to soil structure. The soil water stability index reflects the ability of soil to resist hydraulic damage. The index value is between 0 and 1. The larger the value, the more stable the structure. The long short-term memory network controls the flow of information through a gating mechanism. The forgetting gate threshold of 0.6 means that 60% of historical information is retained, which is suitable for processes such as soil infiltration that have long-term memory characteristics. The kernel parameter of the Gaussian kernel function determines the degree of nonlinearity of the feature space mapping. The parameter value of 0.8 avoids overfitting while maintaining the nonlinear mapping capability.The destruction of soil structure leads to a decrease in infiltration capacity. The more severe the damage, the more significant the decrease. The infiltration rate of severely damaged soil drops to less than 30% of the original value.

[0031] For the above step S4, this embodiment obtains the surface soil feature points of the target area from the soil surface morphology collector under different rainfall intensities, and obtains the soil pore dynamic change data according to the matching of the surface soil feature points; and collects the soil moisture content data of the target area from the soil moisture sensor; and uses a long short-term memory network to perform time series fitting on the soil pore dynamic change data and the soil moisture content data to obtain the soil pore structure influencing factor of the target area.

[0032] In addition, the pore distribution of soil samples can also be measured every hour by a soil porosity scanner to obtain soil pore dynamic change data, which is not limited here. In addition, in the embodiment of the present invention, different intensities of rainfall are set as light rain according to hourly rainfall less than 5 mm, moderate rain according to 5-10 mm, heavy rain according to 10-20 mm, and heavy rain according to 20 mm or more. The moisture sensor monitors the change of soil moisture content at depths of 5 cm, 15 cm, and 30 cm. The rainfall and runoff data are collected from the rain gauge and runoff trough. If the soil porosity decreases by more than 20% or the surface closure exceeds 50%, it is determined that significant structural damage has occurred. The influence mechanism of soil structure on runoff formation is established from the soil porosity change data and the infiltration rate data. The soil porosity scan obtains the internal structure image of the soil through the principle of X-ray tomography, scans once every 1 hour, and records the change of pore distribution. In the initial state, the soil pores are distributed in a reticular manner. As the rainfall continues, the pores are gradually destroyed and filled. Soil surface image acquisition records the micromorphological characteristics of the surface, and extracts feature points through edge detection. Feature points include pore boundaries, aggregate contours, etc., and use these feature points to achieve image registration. The long short-term memory network controls the flow of information through a gating mechanism and is suitable for processing data with time series characteristics such as soil porosity. The forgetting gate threshold of 0.6 means retaining 60% of historical information, allowing the model to remember long-term trends and respond to short-term fluctuations in a timely manner. The Gaussian kernel function captures the complex relationship between soil structure and runoff through nonlinear mapping. The kernel parameter of 0.8 avoids overfitting while maintaining nonlinear characteristics.

[0033] Furthermore, in the above embodiment, the method also includes preprocessing the real-time data to improve the accuracy. Before the real-time data output by the rainfall runoff simulation prediction model is input into the small hydropower group output load prediction model, it is first necessary to obtain the real-time rainfall data of the target area monitored in real time, and preprocess the real-time rainfall data. The preprocessing at least includes mean filling of the first abnormal missing data and / or deletion of the second abnormal missing data. Of course, the first abnormal missing data and the second abnormal missing data only represent the type of abnormal missing data. In addition, the preprocessing may also include data cleaning, data normalization, data integration and other links.

[0034] In an embodiment of the present invention, in order to obtain the final accurate power generation prediction result, it is necessary to comprehensively consider the influence of factors in multiple dimensions. Specifically, the power dispatching operation data at least includes reservoir load data, grid load side demand data and hydropower cluster operation data. In the process of obtaining the corresponding power generation prediction result, the operation rules of the target small hydropower group are constructed according to the grid load side demand data and the hydropower cluster operation data. This rule not only takes into account the power demand on the load side connected to the target small hydropower group, but also takes into account the operation of the hydropower units of the current small hydropower group, thereby improving the subsequent data accuracy. Based on the above-mentioned operation rules, the power generation data to be corrected obtained from the output result and the reservoir load data is corrected to obtain an accurate power generation prediction result.

[0035] In order to accurately control the subsequent power generation process and the management of carbon emissions, the embodiment of the present invention will also reasonably control reservoirs, etc. Specifically, it will generate reservoir control instructions that match the power generation forecast results and send them to the corresponding small hydropower group control terminal. The instructions include but are not limited to gate linkage control instructions, flood control instructions, drought control instructions and reservoir water level adjustment instructions. In a hydropower station, the gate is an important device for controlling water flow. Through linkage control instructions, the coordinated operation of multiple gates can be achieved to meet the changes in the operation of the hydropower station and the power grid demand. During the flood season, according to the dispatching command and supervision of the flood control command organization, the reservoir must obey the flood control instructions to ensure that the reservoir water level is within a safe range and prevent the occurrence of flood disasters. In the dry season, according to the drought relief plan and the emergency water dispatching implementation plan, the reservoir needs to store or release water according to the drought relief control instructions to meet the drought relief needs.

[0036] Another embodiment of the present invention provides a small hydropower group power generation prediction system. For details, please refer to FIG. 2 . FIG. 2 is a schematic diagram of the structure of the small hydropower group power generation prediction system in one embodiment of the present invention, which includes: The initial model module 11 is used to analyze the historical rainfall data and historical runoff data of the target area adjacent to the target small hydropower group based on the time series analysis method, and to construct an initial rainfall-runoff simulation model according to the rainfall-runoff relationship fitting curve obtained by the analysis; The infiltration rate influencing factor module 12 is used to determine the soil moisture characteristic curve according to the soil geological data of the target area and the collected soil moisture sample data, and obtain the soil moisture infiltration rate influencing factor based on the analysis of the soil moisture characteristic curve; The infiltration rate influencing factor module 13 is used to obtain the soil infiltration rate of the target area under different rainfall intensities and obtain the soil moisture infiltration rate influencing factor; The pore structure influencing factor module 14 is used to analyze the pore change data of the target area under different rainfall intensities based on a time series analysis method to obtain the soil pore structure influencing factor; The load prediction model module 15 is used to modify the initial rainfall runoff simulation model with the soil moisture infiltration rate influencing factor and the soil moisture infiltration rate influencing factor to obtain a rainfall runoff simulation prediction model; based on the rainfall runoff relationship fitting curve, the soil moisture characteristic curve and the soil pore structure influencing factor, a small hydropower group output load prediction model is constructed through multivariate regression analysis; The prediction module 16 is used to input the real-time data output by the rainfall runoff simulation prediction model into the small hydropower group output load prediction model, and obtain the corresponding power generation prediction result based on the output result and the power dispatching operation data of the target small hydropower group.

[0037] The method and system for predicting power generation of a small hydropower group provided by the embodiment of the present invention have the beneficial effects of at least one of the following: (1) First, an initial rainfall-runoff simulation model is constructed based on historical data. Then, the factors affecting the infiltration rate and the infiltration rate under rainfall conditions are analyzed respectively, and the constructed initial model is corrected to obtain a more accurate rainfall-runoff simulation prediction model, thereby providing an accurate model support for the subsequent power generation calculation; (2) Through intelligent analysis of the runoff inflow phenomenon generated under rainfall conditions, more accurate power generation prediction results can be obtained. On the one hand, it helps the power grid dispatching center to better arrange power production and transmission, and reduce power grid fluctuations and load imbalances caused by inaccurate predictions. On the other hand, accurate prediction of the power generation capacity of small hydropower groups can help hydropower stations formulate more reasonable power generation plans, avoid waste of water resources and idle power generation equipment, and provide accurate data support for subsequent power generation control and carbon emission management, thus promoting the intelligent control process of small hydropower groups.

[0038] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for predicting power generation of a small hydropower group, characterized in that: include: Based on the time series analysis method, the historical rainfall data and historical runoff data of the target area adjacent to the target small hydropower group are analyzed, and the initial rainfall-runoff simulation model is constructed according to the rainfall-runoff relationship fitting curve obtained by the analysis; Determine a soil moisture characteristic curve based on the soil geological data of the target area and the collected soil moisture sample data, and obtain a soil moisture infiltration rate influencing factor based on an analysis of the soil moisture characteristic curve; Obtaining the soil infiltration rate of the target area under different rainfall intensities, and obtaining the influencing factor of the soil moisture infiltration rate; Based on the time series analysis method, the pore change data of the target area under different rainfall intensities are analyzed to obtain the soil pore structure influencing factors; The initial rainfall-runoff simulation model is modified by the soil moisture infiltration rate influencing factor and the soil moisture infiltration rate influencing factor to obtain a rainfall-runoff simulation prediction model; based on the rainfall-runoff relationship fitting curve, the soil moisture characteristic curve and the soil pore structure influencing factor, a small hydropower group output load prediction model is constructed through multivariate regression analysis; The real-time data output by the rainfall runoff simulation prediction model is input into the small hydropower group output load prediction model, and the corresponding power generation prediction result is obtained based on the output result and the power dispatching operation data of the target small hydropower group.

2. The method for predicting power generation of a small hydropower group according to claim 1, characterized in that: The time series analysis method is used to analyze the historical rainfall data and historical runoff data of the target area adjacent to the target small hydropower group, including: Acquire historical rainfall data and corresponding original water level and flow change data of the target area, and obtain historical runoff data according to the relationship between the historical rainfall data and the original water level and flow change data; According to the surface coverage parameters of the target area, the underlying surface of the target area is subjected to rasterization processing, and the watershed confluence time is obtained from the rasterization processing result; Using a sliding time window to perform time series segmentation on the historical rainfall data and the historical runoff data, obtaining corresponding correlation values ​​according to the segmentation results, and determining the correlation values ​​that meet the conditions as rainfall-runoff conversion parameters; Based on the radial basis kernel function, a corresponding rule between the historical rainfall data and the historical runoff data is constructed, and a quadratic programming analysis is performed on the corresponding rule according to the rainfall-runoff conversion parameter to obtain a rainfall-runoff relationship curve.

3. The method for predicting power generation of a small hydropower group according to claim 1, characterized in that: Determining a soil moisture characteristic curve based on the soil geological data of the target area and the collected soil moisture sample data, and obtaining a soil moisture infiltration rate influencing factor based on an analysis of the soil moisture characteristic curve, includes: Correcting the soil moisture data collected by the soil moisture sensor to obtain soil moisture sample data; Based on the digital elevation model technology, the soil geological data of the target area is processed to obtain a soil moisture change curve, wherein the change area of ​​the soil moisture change curve is obtained by partitioning the target area by a rasterization method; Using a long short-term memory network to perform time series fitting analysis on the soil moisture change curve to obtain soil moisture characteristics, wherein the types of the soil moisture characteristics at least include hydraulic conductivity, capillary potential and water retention curve; The soil moisture variation curve is analyzed according to the soil moisture characteristics to obtain a soil moisture characteristic curve, and a soil moisture infiltration rate influencing factor is obtained from the coupling relationship among the hydraulic conductivity, capillary potential and water retention curve.

4. The method for predicting power generation of a small hydropower group according to claim 3, characterized in that: The soil moisture data collected by the soil moisture sensor is corrected to obtain soil moisture sample data, including: The soil moisture data is corrected according to the temperature data synchronously collected by the soil moisture sensor. If the corrected soil moisture content is within the saturated moisture content range and the rate of change between adjacent moments is less than a preset threshold, the corresponding data is determined as the soil moisture sample data.

5. The method for predicting power generation of a small hydropower group according to claim 1, characterized in that: The step of obtaining the soil infiltration rate of the target area under different rainfall intensities and obtaining the soil moisture infiltration rate influencing factor includes: Under different rainfall intensities, the soil infiltration rate collected by the double-ring infiltrator was obtained, and the soil bulk density was measured using a ring knife; The soil porosity of the target area is measured based on a capillary method, and a soil structure stability index is calculated according to the soil bulk density value and the soil porosity, wherein the soil structure stability index reflects the degree of soil structure damage; Based on the soil structure stability index and the soil infiltration rate, an infiltration rate mapping relationship is established through a Gaussian kernel function; The infiltration rate mapping relationship and the rainfall intensity are analyzed to obtain a soil moisture infiltration rate influencing factor.

6. The method for predicting power generation of a small hydropower group according to claim 1, characterized in that: The time series analysis method is used to analyze the pore change data of the target area under different rainfall intensities to obtain soil pore structure influencing factors, including: Under different rainfall intensities, the surface soil feature points of the target area are obtained from the soil surface morphology collector, and the soil pore dynamic change data are obtained according to the matching of the surface soil feature points; and the soil moisture content data of the target area is collected from the soil moisture sensor; The soil pore dynamic change data and the soil water content data are time-series fitted using a long short-term memory network to obtain the soil pore structure influencing factor of the target area.

7. The method for predicting power generation of a small hydropower group according to claim 1, characterized in that: Before inputting the real-time data output by the rainfall runoff simulation prediction model into the small hydropower group output load prediction model, the method further includes: Acquire real-time rainfall data of the target area monitored in real time; Preprocessing the real-time rainfall data, wherein the preprocessing at least includes mean filling the first abnormal missing data and / or deleting the second abnormal missing data; The pre-processed real-time rainfall data is input into the rainfall runoff simulation prediction model to obtain the real-time data output by the rainfall runoff simulation prediction model.

8. The method for predicting power generation of a small hydropower group according to claim 1, characterized in that: The power dispatching operation data at least includes reservoir load data, grid load side demand data and hydropower generation cluster operation data; The corresponding power generation prediction result is obtained based on the output result and the power dispatching operation data of the target small hydropower group, including: Constructing the operation rules of the target small hydropower group according to the grid load side demand data and the hydropower generation cluster operation data; The power generation data to be corrected obtained from the output result and the reservoir load data is corrected based on the operation rule to obtain a corresponding power generation prediction result.

9. The method for predicting power generation of a small hydropower group according to claim 1, characterized in that: After obtaining the corresponding power generation prediction result, the method further includes: The generated reservoir control instructions matching the power generation prediction results are sent to the corresponding small hydropower group control terminal, wherein the reservoir control instructions at least include gate linkage control instructions, flood control instructions, drought control instructions and reservoir water level regulation instructions.

10. A small hydropower group power generation prediction system, characterized in that: include: The initial model module is used to analyze the historical rainfall data and historical runoff data of the target area adjacent to the target small hydropower group based on the time series analysis method, and to construct an initial rainfall-runoff simulation model based on the rainfall-runoff relationship fitting curve obtained by the analysis; An infiltration rate influencing factor module is used to determine a soil moisture characteristic curve according to the soil geological data of the target area and the collected soil moisture sample data, and obtain a soil moisture infiltration rate influencing factor based on an analysis of the soil moisture characteristic curve; The infiltration rate influencing factor module is used to obtain the soil infiltration rate of the target area under different rainfall intensities and obtain the soil moisture infiltration rate influencing factor; The pore structure influencing factor module is used to analyze the pore change data of the target area under different rainfall intensities based on the time series analysis method to obtain the soil pore structure influencing factor; A load prediction model module is used to modify the initial rainfall runoff simulation model with the soil moisture infiltration rate influencing factor and the soil moisture infiltration rate influencing factor to obtain a rainfall runoff simulation prediction model; based on the rainfall runoff relationship fitting curve, the soil moisture characteristic curve and the soil pore structure influencing factor, a small hydropower group output load prediction model is constructed through multivariate regression analysis; The prediction module is used to input the real-time data output by the rainfall runoff simulation prediction model into the small hydropower group output load prediction model, and obtain the corresponding power generation prediction result based on the output result and the power dispatching operation data of the target small hydropower group.

Images