CN115712800A - Reservoir water level fluctuation processing method - Google Patents
Reservoir water level fluctuation processing method Download PDFInfo
- Publication number
- CN115712800A CN115712800A CN202211481284.1A CN202211481284A CN115712800A CN 115712800 A CN115712800 A CN 115712800A CN 202211481284 A CN202211481284 A CN 202211481284A CN 115712800 A CN115712800 A CN 115712800A
- Authority
- CN
- China
- Prior art keywords
- data
- reservoir
- water level
- flow
- dam
- 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.)
- Granted
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A10/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE at coastal zones; at river basins
- Y02A10/40—Controlling or monitoring, e.g. of flood or hurricane; Forecasting, e.g. risk assessment or mapping
Abstract
The invention relates to a reservoir water level fluctuation processing method, which comprises the following steps: s1, acquiring the dam front water level historical data of a target reservoir; s2, removing coarse data in the historical data; s3, filling missing data in the historical data after the operation of the S2; s4, filtering and smoothing the historical data after the operation of the S3 by adopting an SG filtering mode; s5, calculating a runoff inverse hanging rate P0 by using the original dam front water level historical data, calculating a runoff inverse hanging rate P1 by using the dam front water level data operated in the S4, judging whether the P1 is smaller than the P0, if so, outputting a value of the P1 and the dam front water level data processed by the S4, and if not, returning to execute the S4; the runoff inverse hanging rate indicates the frequency that the warehousing flow of the target reservoir is smaller than the ex-warehousing flow of the upstream adjacent reservoir. The method can improve the accuracy of the water level in front of the reservoir dam, thereby reducing the scheduling risk.
Description
Technical Field
The invention relates to the technical field of water resource management, in particular to a reservoir water level fluctuation processing method.
Background
With the development of social economy and technology, people develop and utilize water resources more and more fully and perfectly, and a hydropower station is an important mode. And the hydropower station has a very complicated monitoring system for safe and economic operation. However, in the construction of various index systems, the measurement of the water level data before the dam is an important component participating in the safe and economic operation of the reservoir of the hydropower station, because the reservoir capacity of the reservoir is difficult to directly measure in the daily work of reservoir scheduling and the like, the water level before the dam is required to be taken as indirect measurement data, and then the indirect measurement data is converted into the reservoir capacity through a water level reservoir capacity curve and is applied to the calculation of the water quantity of the reservoir. However, in the prior art, the dam front water level data is directly acquired and stored in a database by instantly measuring the water level, and is applied to various works such as daily scheduling, risk evaluation and the like. Because the water level before the dam has large fluctuation, the evaluation and the processing of the measurement characteristic of the water level before the dam are lacked in the prior art, and the calculation of the storage capacity by adopting the directly measured water level data before the dam has large errors, the scheduling risk is increased, and the feasibility of the reservoir scheduling instruction is reduced. Therefore, a method for processing the measured data of the dam front water level of the reservoir is urgently needed to process the dam front water level fluctuation so as to reduce the dam front water level error and reduce the scheduling risk.
Disclosure of Invention
The invention aims to solve the technical problem of providing a reservoir water level fluctuation processing method for processing the water level fluctuation before the dam so as to reduce the water level error before the dam and reduce the scheduling risk.
The technical scheme adopted by the invention for solving the technical problems is as follows: a method for processing fluctuation of reservoir water level comprises the following steps,
s1, acquiring the dam front water level historical data of a target reservoir;
s2, removing coarse data in the historical data;
s3, filling missing data in the historical data after the operation of the S2;
s4, filtering and smoothing the historical data after the operation of the S3 by adopting an SG filtering mode;
s5, calculating a runoff inverse hanging rate P0 by using the original dam front water level historical data, calculating a runoff inverse hanging rate P1 by using the dam front water level data operated in the S4, judging whether the P1 is smaller than the P0, if so, outputting a value of the P1 and the dam front water level data processed by the S4, and if not, returning to execute the S4; the runoff inverse hanging rate indicates the frequency that the warehousing flow of the target reservoir is smaller than the ex-warehousing flow of the upstream adjacent reservoir.
Further, in S2, the manner of removing the coarse data is as follows: and eliminating data exceeding an interval range (mu-3 sigma, mu +3 sigma) in the historical data, wherein mu represents the average value of the historical data, and sigma represents the standard deviation of the historical data.
Furthermore, the manner of filling the missing data in S3 is,
wherein l represents filled missing data and has a unit of m, l 1 、l 2 Representing dam front water level data adjacent to the missing data in the historical data sequence, wherein the unit is m, t 1 、t 2 And t represents the expected time of the missing data.
Further, the filtering and smoothing process performed by using the SG filtering method in S4 specifically includes,
and fitting the historical data subjected to the coarse data processing and the missing data processing by adopting the following fitting formula:
wherein x is i Display filterThe ith dam front water level data to be fitted in the wave window, y i Represents the ith fitted data, a 0 ，a 1 ，…，a k-1 K-1 is a variable to be solved, fitting by using a k-1 polynomial, wherein k is a natural number and k is more than or equal to 2;
wherein, a 0 ，a 1 ，…，a k-1 The solution of (c) is as follows:
if the length of the data sequence in the filtering window is N, N =2m +1, m is a natural number, and N is greater than or equal to k, N polynomials can be generated in the filtering window according to the fitting formula, where k variables to be solved, that is, k element linear equations, are determined in the N polynomials, and the least square formula is used:
can solve to obtain a 0 ，a 1 ，…，a k-1 。
Further, the calculation method of the filtering window data sequence length N is: dispersing the saint-venon equation set by adopting a Preissmann four-point weight implicit difference format, then simplifying the equation set into a linear equation set by neglecting second-order micro quantity, solving the linear equation set, and obtaining the flow reaching time used by the sudden change flow reaching a downstream target reservoir when the flow of an upstream adjacent reservoir is suddenly changed, wherein the flow reaching time is divided by the sampling frequency of the water level data in front of the dam, and the length N of the filtering window data sequence is obtained by taking the nearest odd number upwards.
Further, the calculation method of the radial flow hang-up rate in S5 is as follows:
wherein P represents the runoff inverse hanging rate, I t' Represents the flow rate, I 'of the upstream river channel of the target reservoir in the period t' t′ Representing the warehousing flow of the target reservoir reversely deducted from the storage quantity change and the ex-warehouse flow in the T 'time period, wherein T' represents the time period, and T represents the total calculation time period;
in the above formula I t' ＝O t'-j ，
Wherein O is t'-j The flow of the upstream river channel control node is represented, namely the flow of the upstream adjacent reservoir out of the reservoir, j represents the number of time periods required for the flow of the upstream river channel control node to reach the target reservoir, and l t'+1 Representing the dam front water level of the target reservoir in a T '+1 period, f is the mapping relation between the dam front water level of the target reservoir and the reservoir capacity and is used for converting the dam front water level into the reservoir capacity, and T is the calculated total period length of O' t′ The flow of the target reservoir is calculated in the way of
Wherein
The generated flow rate of the target reservoir in the period t' is shown,
the water level data can be obtained by inquiring an NHQ curve according to the load data and the dam front water level data,
representing the gate flow of the target reservoir in the period t',
the water level data before the dam and the gate opening can be inquired according to the water level data before the dam and the discharge curve can be calculated.
Further, in the P1 calculation process, the missing load data is filled up by using the following formula:
wherein w represents missing load data in MW, w 1 、w 2 Load data adjacent to the missing load data in MW, t ″) 1 、t″ 2 Correspondence of load data adjacent to each other before and after the load data indicating the missingT "represents the desired time scale for missing load data.
Further, in the P1 calculation process, the missing gate opening data is filled by using the following formula:
g＝g 1
wherein g represents the missing gate opening data and the unit is m, g 1 The gate opening data adjacent to the gate opening data indicating the missing gate opening data is expressed by m.
Compared with the prior art, the invention has the advantages that: the invention carries out missing data processing and coarse data processing on the dam front water level historical data, compares the runoff inverse hanging rate calculated by the original historical data with the runoff inverse hanging rate calculated by the dam front water level data after the series of operation processing after filtering and smoothing processing is carried out by adopting an SG filtering method, and repeats the SG filtering method operation when the calculated runoff inverse hanging rate does not meet the requirement until the calculated runoff inverse hanging rate meets the requirement, thereby improving the accuracy of the dam front water level data, reducing the dam front water level data error, enhancing the dam front water level data quality, further improving the accuracy of the reservoir warehousing flow and other data indirectly calculated by the dam front water level, and reducing the scheduling risk.
Drawings
Fig. 1 is a flowchart of a reservoir level fluctuation processing method according to the present application.
FIG. 2 is a comparison diagram of the water level before dam before and after processing.
FIG. 3 is a schematic diagram of reservoir warehousing flow comparison before and after dam front water level data processing.
Detailed Description
The invention is described in further detail below with reference to the embodiments of the drawing, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are illustrative and intended to be illustrative of the invention and are not to be construed as limiting the invention.
In the description of the present application, it should be noted that, for the terms of orientation, such as "central", "lateral", "longitudinal", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", etc., indicate orientations and positional relationships based on the orientations or positional relationships shown in the drawings, which are only for convenience of description and simplification of the description, but 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 should not be construed as limiting the specific scope of the present application. The terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or to implicitly indicate the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of the present invention, "a plurality" means two or more unless specifically defined otherwise.
As shown in fig. 1, which is a flow chart of the reservoir level fluctuation processing method of the present invention, as shown in the figure, the reservoir level fluctuation processing method includes the following steps,
s1, acquiring the dam front water level historical data of a target reservoir;
s2, removing coarse data in the historical data;
s3, filling missing data in the historical data after the operation of the S2;
s4, filtering and smoothing the historical data after the operation of the S3 by adopting an SG filtering mode;
s5, calculating a runoff inverse hanging rate P0 before processing by using the original dam water level historical data, calculating a runoff inverse hanging rate P1 after processing by using the dam water level data after S4 operation, judging whether P1 is smaller than P0, if so, outputting the value of P1 and the dam water level data after S4 operation, and if not, returning to execute S4; the runoff inverse hanging rate indicates the frequency that the warehousing flow of the target reservoir is smaller than the ex-warehousing flow of the upstream adjacent reservoir.
When the hydropower station has a long operation period and accumulates more data, the same type of group data only contains random errors, the random errors are calculated to obtain standard deviation, an interval is determined according to a certain probability, the errors exceeding the interval are considered not to belong to the random errors but to be coarse errors, and the data containing the coarse errors are removed. Specifically, the hydropower station reservoir dam front water level data is subjected to normal distribution test, the probability of numerical distribution in an interval (mu-sigma, mu + sigma) is 0.6826, the probability of numerical distribution in an interval (mu-2 sigma, mu +2 sigma) is 0.9545, and the probability of numerical distribution in an interval (mu-3 sigma, mu +3 sigma) is 0.9973, wherein sigma represents standard deviation and mu represents mean value. It can be seen that the dam front water level data is almost completely concentrated in the interval (μ -3 σ, μ +3 σ), so that the data outside the interval (μ -3 σ, μ +3 σ) in the historical data is defined as coarse data and eliminated, which is generally referred to as the three-sigma criterion.
In each data acquisition process of a hydropower station, communication or instrument and equipment faults may cause that measurement data does not form a discontinuous time sequence at a fixed time frequency, namely, missing data is formed, and meanwhile, the data can also form the missing data after coarse data processing, so that the application difficulty of the data sequence is increased, the data sequence is not beneficial to automatic continuous calculation or use, and the missing dam front water level data needs to be filled when the dam front water level is processed.
Most of the conditions of the dam front water level of the reservoir are used as the characteristic number of the reservoir capacity of the reservoir, and the difference between the reservoir capacity and the flow magnitude of the inlet and outlet in a short time is large, so that the reservoir capacity of the reservoir can be considered to be continuously changed in a short time, namely, the reservoir capacity is strongly correlated with dam front water level data at the adjacent sampling moments before and after, and missing data in dam front water level historical data can be filled by adopting the following formula:
wherein, l represents the filled missing data and the unit is m, l 1 、l 2 Representing a gap in the dam data sequenceThe unit of the adjacent water level data before and after the leakage data is m, t 1 、t 2 Indicates the time t corresponding to the adjacent water level data before and after the missing data 0 Indicating the desired time scale for missing data.
In this embodiment, the filter smoothing process in step S4 by using the SG filter method is performed,
and fitting the historical data subjected to the coarse data processing and the missing data processing by adopting the following fitting formula:
wherein x is i Representing the ith dam front water level data to be fitted in the filtering window, y i Representing the corresponding data after the ith fitting, namely dam front water level data a after SG filtering smoothing 0 ，a 1 ，…，a k-1 For the variable to be solved, k-1 represents fitting by using k-1 degree polynomial, k is a natural number and is more than or equal to 2, the fitting effect is better if the k value is larger, the nonlinearity degree is higher, but if the k value is too large, the distortion is easy to occur, and the value range of the k value is generally [3,10 ] according to experience]Natural numbers between;
a 0 ，a 1 ，…，a k-1 the solution of (c) is as follows:
if the sequence length of dam front water level historical data in a filtering window is N, N =2m +1, m is a natural number, and N is greater than or equal to k, N polynomials can be generated in the filtering window according to a fitting formula, k variables needing to be solved in the N polynomials, namely k element linear equations are obtained through a least square formula:
can solve to obtain a 0 ，a 1 ，…，a k-1 . Since the method of solving variables by using the least square method belongs to the prior art, the detailed process of solving is not described in detail here.
The calculation mode of the length N of the data sequence of the filtering window is as follows: dispersing the saint-venon equation set by adopting a Preissmann four-point weight implicit difference format, then simplifying the equation set into a linear equation set by neglecting second-order micro quantity, solving the linear equation set, obtaining the flow arrival time used by the sudden change flow to reach a downstream target reservoir when the flow of the upstream adjacent reservoir is suddenly changed, dividing the flow arrival time by the data sampling frequency, and taking the nearest odd number upwards to obtain the filtering window data sequence length N.
Specifically, the water level change before the dam is related to the inflow and outflow processes in the same time period, and the outflow is also a time sequence with high correlation with the inflow. Therefore, the related data analysis can be carried out on the inflow sequence, and the length N of the filtering window data sequence is correspondingly calculated by adopting the length of the autocorrelation data sequence. In the course of water flow channel evolution, the length of the autocorrelation data sequence can be equivalent to the runoff evolution time. Determining upstream and downstream boundary conditions under different conditions, and analyzing the relation between runoff evolution time and the boundary conditions through river one-dimensional hydrodynamics.
The river channel one-dimensional hydrodynamic model can finely describe the evolution process of water flow in a river channel of a reservoir area, so that the time-dependent change process of the water level and the flow of each section along the course is calculated. The basic control equation of the one-dimensional water flow motion of the riverway is an St.Vietnam equation set, and the equation set comprises a continuous equation and a momentum equation:
wherein, B is the width of the surface of the water passing section, and the unit is m; z is water level and the unit is m; t is time in units of s; q is the flow rate in m 3 S; x is the longitudinal channel distance along the main flow direction, and the unit is m; q is the side inflow in m 3 S; alpha is a momentum correction coefficient; a is the water passing area and the unit is m 2 (ii) a g is gravity acceleration; s. the f For the reduction of the friction resistance, it can be represented by the following formula:
in the formula: n is a radical of an alkyl radical c The Manning roughness coefficient of the water delivery channel; r is the hydraulic radius and is expressed in m.
The holy-south equation set belongs to a first-order quasi-linear hyperbolic partial differential equation set, an analytic solution of the holy-south equation set cannot be obtained at present, and an approximate solution of the holy-south equation set can only be obtained by a numerical discrete method. Finite Difference Method (FDM) is the most widely applied method in numerical simulation, which replaces continuous function values of the whole solution region with values on finite nodes. Finite difference methods can be classified into explicit formats and implicit formats according to whether the calculation process depends on unknown quantities of the time to be solved. In the embodiment, the saint-wien equation set is discretized by adopting a Preissmann four-point weighted implicit differential format with high convergence speed and good stability. The method comprises the following specific steps:
the saint-Vinan equation system is discretized by adopting a Preissmann four-point weighted implicit difference format, and the difference form of a dependent variable and a derivative function thereof in space and time is as follows:
wherein theta is a weighting coefficient, and theta is more than or equal to 0 and less than or equal to 1.0; f represents a function value; the subscripts i ', i' +1 denote the i 'th and i' +1 th sections, respectively; superscripts n and n +1 respectively represent the nth and n +1 th moments; Δ t is a time discrete step; Δ x is a spatially discrete step size.
In the Preissmann format, when 0.5. Ltoreq. Theta. Ltoreq.1.0, the format is unconditionally stable. In the present model, θ is taken to be 0.75. In order to control the phase error generated by numerical dispersion, the stability of the format can be further improved through the reasonable collocation of the time discrete step length delta t and the space discrete step length delta x. In the present model, Δ x is the rootAutomatically dividing according to the set delta t, and taking
To calculate the initial water depth of the downstream control section of the channel section. In the actual calculation process, when the set time step is adjusted, the space step calculated by the model is changed correspondingly.
After the Saint Vietnam equation set is dispersed, the second-order trace is ignored, the second-order trace is simplified into a linear equation set, the linear equation set can be directly solved, and the unknown quantity to be solved is the water level and the flow of each calculation section at each moment. The channel control equation between the i 'th and i' +1 th sections can be linearized into the following format:
wherein the content of the first and second substances,
is the flow and water level of the ith' section at the time of (n + 1),
the flow and water level of the (i' + 1) th section at the (n + 1) th moment; coefficient C i' ,D i' ,E i' ,G i' ,F i' ,Φ i' All can be obtained by calculating hydraulic parameters and hydraulic elements at the nth moment, and the formula is as follows:
wherein, the first and the second end of the pipe are connected with each other,
represents the average value of the surface widths of the ith 'section and the ith' +1 section at the nth moment, u is the average flow velocity of the sections,
the average value of the flow rate of the i '-th cross section and the i' + 1-th cross section is shown, and the other variables have the same meanings as above.
In the saint-Venen equation system solving process, an upstream selected flow boundary is the ex-warehouse flow of an upstream adjacent reservoir, a downstream selected water level boundary is the dam front water level of a downstream adjacent reservoir, the ex-warehouse flow of the upstream adjacent reservoir changes with the flow of the upstream adjacent reservoir, the time of the changed flow reaching the downstream adjacent reservoir is calculated, further, after the change of the flow of the upstream adjacent reservoir, the time when the flow change of a downstream target reservoir is 99% of the corresponding flow change of the upstream adjacent reservoir is defined as the flow reaching time, the flow reaching time is divided by the sampling frequency of the dam front water level data of the downstream target reservoir, and the nearest base number is taken upwards to obtain the filtering window sequence length N.
In this embodiment, the manner of calculating the reverse hanging rate of the radial flow in S5 is as follows:
Wherein P represents the runoff inverse hanging rate, I t' Representing the upstream river inflow of the target reservoir in the t' period, I t "represents the warehousing flow of the target reservoir which is reversely deduced from the storage quantity change and the ex-warehouse flow in the T 'time period, T' represents the time period, and T represents the total time period number;
in the above formula I t' ＝O t'-j ，
Wherein O is t'-j The flow of the upstream river channel control node of the target reservoir is represented, namely the flow of the upstream adjacent reservoir of the target reservoir, j represents the number of time periods required for the flow of the upstream river channel control node to flow to the target reservoir, and l t'+1 Representing the dam front water level of the target reservoir in T' +1 time period, wherein f is the mapping relation between the dam front water level of the target reservoir and the reservoir capacity, the dam front water level is converted into the reservoir capacity, the mapping relation can be obtained by inquiring the reservoir capacity curve of the target reservoir, T is the calculated total time period number, and O t "is the flow of the target reservoir, and the calculation mode is
Wherein
The generated flow rate of the target reservoir in the period t' is shown,
the water flow control device can be obtained by inquiring an NHQ curve according to load data and dam water level data and calculating, wherein the NHQ curve is also called a unit output-water head-flow relation curve,
representing the target hydropower station gate flow during the time period t',
the calculation can be carried out according to the gate opening and dam front water level data inquiry discharge curve.
Calculating to obtain a runoff inverse hanging rate P0 before treatment by using original dam water level historical data according to a runoff inverse hanging rate calculation formula, calculating to obtain a treated runoff inverse hanging rate P1 by using dam water level data subjected to water level fluctuation treatment, if P1 is smaller than P0, indicating that the quality of the dam water level data subjected to fluctuation treatment is improved, outputting P1 and the treated dam water level data, and if P1 is larger than or equal to the value of P0, adjusting the value of a filtering parameter k, returning to the step S4, and continuing to perform filtering smoothing until P1 is smaller than P0.
In this embodiment, in the P1 calculation process, the following formula is adopted to fill up missing load data:
wherein w represents missing load data in MW, w 1 、w 2 Representing the load data missing in the load data sequence and the adjacent load data in the unit of MW, t 1 ”、t' 2 'represents a time scale corresponding to adjacent and nearest water level data before and after the missing water level data, and t' represents a desired time scale of the missing water level data. The load data is filled, so that the calculated power generation flow of the target reservoir is more accurate.
In this embodiment, in the P1 calculation process, the missing gate opening data is filled by using the following formula:
g＝g 1
wherein g represents the missing gate opening data and the unit is m, g 1 And the unit of the gate opening data which is adjacent to the gate opening data before the missing gate opening data in the gate opening data sequence is m. And the gate opening data is filled, so that the calculated gate flow of the target reservoir is more accurate.
Taking two adjacent hydropower stations of a certain watershed cascade system as an example, the method is used for processing the water level data in front of the dam of the downstream target reservoir, wherein the upstream hydropower station comprises 4 183MW hydraulic turbines, 5 holes of 8 × 16m flood gates and four holes of sand washing bottom holes, the downstream adjacent hydropower station comprises 6 59.5MW hydraulic turbines and 5 holes of 13 × 16m flood gates, the length of a river channel between the two stations is about 30km, and no branch interference exists between the two stations. The implementation flow of the water level fluctuation processing method of the invention is further described by taking the dam front water level data of one year from 2019, 1/0 to 2019, 12/31/24 as an example, wherein the sampling frequency of the dam front water level data of the downstream target reservoir is 5min, and the second column in table 1 lists the original dam front water level historical data of the downstream target reservoir part.
Firstly, processing coarse data in an original dam water level historical data sequence, namely calculating the selected dam water level data of one year to obtain a mean value mu =552.226 and a standard deviation sigma =1.043, wherein the value of mu-3 sigma is 549.097, and the value of mu +3 sigma is 555.355, so that data exceeding a range [549.097 and 555.355] in the original dam water level historical data are removed, for example, 20 minutes of dam water level data of 18 days of 1 month and 10 months in 2019 exceeds the range when the dam water level data of 20 minutes exceeds the range, and the data are removed. The data of the water level before the dam after processing the coarse data of the original data are shown in the third column of table 1.
And performing data missing processing on dam front water level data after coarse data processing, wherein the data missing processing mainly comes from two aspects, namely missing existing in original data, such as data missing at 17: 35 in 2019, 1/month, 10/day, and data missing caused by removing coarse data, such as data missing at 17: 50: 17: 1/month, 10/month, and 2019, calculating missing data according to the data missing processing formula, filling the missing data in a data sequence, such as data missing at 35: 17: 2019, 1/month, 10/day, and obtaining 553.16 by calculation, filling the data in a corresponding time position in the data sequence, namely realizing data missing processing at 17: 35: 1/month, 10/month, and 2019, and performing data missing processing on the front dam front water level data sequence, and displaying the data before dam as the water level data in the fourth column of table 1.
TABLE 1 statistical table of water level data before dam of downstream reservoir
And smoothing the dam front water level historical data after the processing of the coarse data and the missing data by adopting an SG filtering mode. When SG filtering smoothing is performed, first, filter parameters k and N need to be determined, in this embodiment, the value of the filter parameter k is 3, that is, a ternary linear equation is used to fit dam front water level data, and the parameter N is obtained by solving the saint-wien equation set. The saint-wien equation set is solved discretely by adopting a Preissmann four-point weighted implicit difference format, after the change of the ex-warehouse flow of the upstream adjacent reservoir is obtained, the time for the changed flow to reach the downstream target reservoir is the arrival time, under the normal condition, when the start-up flow of the upstream adjacent reservoir changes suddenly, the flow change of the target reservoir reaches 99% of the sudden change flow, the arrival time is defined as the arrival of the flow, namely, the moment when the start-up flow of the upstream reservoir changes suddenly and the moment when the flow change of the downstream reservoir reaches 99% of the sudden change flow, the time of the interval between the two moments is the arrival time, the arrival time is divided by the data sampling frequency of 5min, the nearest odd number is taken upwards, and the filtering parameter N is obtained, and the N value is 37 if the calculated value is 35.6 or 36.5.
Suppose that the starting regulation flow rate of the upstream adjacent reservoir is 900m 3 And/s, solving the saint-winan equation set, wherein the flow reaching time is 212min when the downstream target reservoir dam front control water level is 550m, the flow reaching time is 207min when the target reservoir dam front control water level is 551m, the flow reaching time is 209min when the target reservoir dam front control water level is 552m, the flow reaching time is 205min when the target reservoir dam front control water level is 553m, and the flow reaching time is 202min when the target reservoir dam front control water level is 554m, as shown in the second row of table 2. If the flow rate of the upstream adjacent reservoir is adjusted to 900m 3 The reduction is 10 percent on the basis of/s, namely the start regulating flow of the upstream adjacent reservoir is810m 3 At/s, the calculated flow arrival times are shown in the third row of Table 2; if the starting flow rate of the upstream adjacent reservoir is adjusted to 900m 3 The reduction is 20 percent on the basis of/s, namely the start regulating flow of the upstream adjacent reservoir is 720m 3 At/s, the calculated flow arrival times are shown in the fourth row of Table 2; if the starting flow rate of the upstream adjacent reservoir is adjusted to 900m 3 The reduction is 30 percent on the basis of/s, namely the initial regulated flow of an upstream adjacent reservoir is 630m 3 The calculated flow arrival times at/s are shown in the fifth row of table 2. If the initial regulation flow is 3000m 3 At/s, the calculated flow arrival time is shown in the sixth row of Table 2, if the flow rate is adjusted to 3000m 3 The calculated flow arrival times at 10%, 20%, 30% reductions on a/s basis are shown in the seventh to ninth rows of Table 2, in such a way that the adjusted flow rate from the upstream adjacent reservoir is 4100m 3 The flow times are shown in Table 2, lines tenth to thirteen when the flow times are reduced by 10%, 20%, 30% on this basis.
TABLE 2 flow arrival time statistics Table
As can be seen from table 2, when the startup flow of the upstream adjacent reservoir is unchanged, the influence of the control water level of the downstream target reservoir on the water flow evolution time from the upstream adjacent reservoir to the downstream target reservoir is small, and the difference of the flow arrival time of the downstream target reservoir under different control water levels can be controlled to be about 5 minutes; when the upstream starting flow is different, the water flow evolution time from the upstream adjacent reservoir to the downstream target reservoir is obviously different, and the larger the upstream starting flow is, the shorter the flow arrival time is. For example, the flow rate of the upstream adjacent reservoir is 900m 3 In the case of/s, the flow arrival time may be set to be about 207min, and the filter window sequence length N may be set to be 43 (the nearest odd number is taken upward) at the data frequency of 5min, but it may be slightly larger than the calculated value when the filter window sequence length N is determined, and for example, it may be set to be 45 in this case. The starting regulating flow of the upstream adjacent reservoir is 3000m 3 When is/sThe arrival time can be set to be about 143min, the length N of the filtering window sequence can be set to be 29 (taking the nearest odd number upwards) by using the data frequency of 5min, and the flow rate is 4100m from the upstream adjacent reservoir 3 In/s, the flow arrival time can be fixed to about 128min, and the filter window sequence length N can be fixed to 27 (taking the nearest odd number upward) at the 5min data frequency. Therefore, the evolution time of the river water flow mainly depends on the starting adjustment flow of the upstream adjacent reservoir, namely the size of the delivery flow. That is, when SG smoothing filtering is performed, the value of the filtering parameter N needs to be calculated according to the outlet flow of the upstream adjacent reservoir, so that the values of the filtering parameter N in the flood season and the non-flood season have great difference.
Calculating a filtering parameter N value by combining the historical data of the flow of the upstream reservoir out of the reservoir, listing N ternary linear equations in a filtering window, and solving a fitting coefficient a for fitting in the filtering window by adopting a least square method 0 ，a 1 ，a 2 For example, the starting flow rate of the upstream adjacent reservoir is 900m 3 When the parameter is/s, the value of the filtering parameter N is 43, the value of the filtering parameter k is 3, 43 dam front water level data between 17 points 15 minutes in 1/10 months in 2019 and 20 points 45 minutes in 10/10 months in 1/9 are subjected to SG filtering smoothing processing, and a coefficient a can be obtained 0 ＝552.98，a 1 ＝0.0421，a 2 = -0.0018, and then a fitting formula is used for calculating to obtain dam front water level data after SG filtering smoothing as shown in the fifth column of the table 1, and the graph 2 shows that comparison is performed before and after dam front water level data is processed, so that curve fluctuation of the processed dam front water level data along with time is reduced, and the curve is smoother.
And after the fluctuation of the dam front water level data of the selected target reservoir 2019 is processed, judging whether the value of P1 is smaller than the value of P0. Specifically, from 0 in 1/2019 to 24 in 12/31/2019, 525600min is provided, the downstream target reservoir collects one dam front water level data every 5min, and 105121 dam front water level data are provided, that is, 105120 time periods are provided, that is, the total number of time periods is 105120. Calculating the primary outlet flow of the upstream reservoir in each time interval, reversely deducing the inlet flow of the downstream target reservoir by the outlet flow of the downstream target reservoir and the water level data before the dam and combining with a reservoir capacity curve, wherein the outlet flow of the downstream target reservoir comprises the power generation flow and the gate discharge flow as described above, if the reversely deduced inlet flow of the downstream target reservoir is less than the outlet flow of the upstream adjacent reservoir, the runoff inverse hanging phenomenon occurs in the time interval, which needs to be explained. Calculating the total time interval of the runoff hang-up phenomenon in 2019, and dividing the total time interval of the runoff hang-up phenomenon by the total time interval in 2019 to obtain the runoff hang-up rate in 2019; further, if the months from 4 to 9 are defined as flood seasons and the rest of the months are defined as non-flood seasons, the runoff inverse hanging rate corresponding to the flood seasons and the non-flood seasons can be obtained according to the calculation principle.
In the runoff inverse hanging rate calculating process, if the dam front water level data before the water level fluctuation processing is adopted to calculate the runoff inverse hanging rate P0 before the processing, if the dam front water level data after the water level fluctuation processing is adopted to calculate the runoff inverse hanging rate P1 after the processing, if the P1 is less than the P0, the data quality improvement after the water level fluctuation processing is carried out on the dam front water level data is shown, the value of the P1 and the dam front water level data after the processing are output, if the value of the P1 is more than or equal to the P0, the value of the parameter k is adjusted to continue the SG filtering, for example, when the SG filtering smoothing processing is carried out for the first time, the value of the k is 3, but the calculated value of the P1 is more than the value of the P0, the value of the k can be adjusted to 4 to continue the SG filtering smoothing processing until the value of the P1 is less than the P0. Table 3 lists the runoff inverse hanging rate values of the downstream target reservoir dam front water level after primary treatment and before treatment, and fig. 3 shows the corresponding target reservoir warehousing flow before and after dam front water level data treatment by comparing, wherein the horizontal coordinate in the figure represents the time interval, and the vertical coordinate represents the warehousing flow of the target reservoir. It can be seen from table 3 that the runoff inverse-hanging phenomenon between the two reservoirs after the target reservoir is subjected to primary dam-front water level fluctuation treatment is obviously improved, the inverse-hanging phenomenon is greatly reduced after the water level fluctuation treatment is performed on the dam-front water level in terms of the overall inverse-hanging rate, but the overall improvement effect is also greatly different between the flood season and the flood season, the inverse-hanging rate in the flood season is reduced by 7.96 percentage points after the dam-front water level treatment, 55% of the inverse-hanging phenomenon is reduced, 10.81 percentage points of the inverse-hanging rate in the flood season are reduced, 37% of the inverse-hanging phenomenon is reduced, and the effect in the flood season is improved more obviously in the overall view. As can be seen from the attached figure 3, the step phenomenon caused by fluctuation of the target reservoir warehousing flow calculated through the processed dam front water level data is reduced, so that the error caused by the dam front water level fluctuation is reduced, the accuracy of the dam front water level data is improved, the curve of the target reservoir warehousing flow changing along with time is smoother, the curve of the target reservoir warehousing flow changing along with time is more consistent with the curve of the ex-warehouse flow of the upstream adjacent reservoir along with time, the runoff consistency is improved, the data such as the reservoir storage capacity and the like indirectly calculated through the dam front water level data are more accurate, and the scheduling risk is reduced.
TABLE 3 statistical table of runoff inverse hanging rate before and after treatment of front water level of downstream reservoir dam
When the runoff inverse hanging rate P1 of the processed dam front water level data is calculated, the method corresponding to the invention is adopted to process the missing data for the gate opening data and the load data of the downstream target reservoir, so that the accuracy of the data can be further improved when the runoff inverse hanging rate after processing is calculated.
While embodiments of the invention have been shown and described, it will be understood by those skilled in the art that: various changes, modifications, substitutions and alterations can be made to the embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.
Claims (8)
1. A reservoir water level fluctuation processing method is characterized by comprising the following steps: comprises the steps of (a) preparing a substrate,
s1, acquiring dam front water level historical data of a target reservoir;
s2, removing coarse data in the historical data;
s3, filling missing data in the historical data after the operation of the S2;
s4, filtering and smoothing the historical data operated in the S3 by adopting an SG filtering mode;
s5, calculating a runoff inverse hanging rate P0 by using the original dam front water level historical data, calculating a runoff inverse hanging rate P1 by using the dam front water level data operated in the S4, judging whether the P1 is smaller than the P0, if so, outputting the value of the P1 and the dam front water level data processed in the S4, and if not, returning to execute the S4; the runoff inverse hanging rate indicates that the frequency of the phenomenon that the warehousing flow of the target reservoir is smaller than the ex-warehouse flow of the upstream adjacent reservoir.
2. The reservoir level fluctuation processing method according to claim 1, characterized in that:
in the step S2, the manner of removing the coarse data is as follows: and eliminating data exceeding the interval range (mu-3 sigma, mu +3 sigma) in the historical data, wherein mu represents the mean value of the historical data, and sigma represents the standard deviation of the historical data.
3. The reservoir level fluctuation processing method according to claim 2, wherein:
the manner of filling up the missing data in S3 is,
wherein, l represents the filled missing data and the unit is m, l 1 、l 2 Representing dam front water level data adjacent to the missing data in the historical data sequence, wherein the unit is m, t 1 、t 2 And t represents the expected time of the missing data.
4. The reservoir level fluctuation processing method according to claim 3, characterized in that:
the filtering and smoothing process in S4 by using an SG filtering method specifically includes,
and fitting the historical data subjected to the coarse data processing and the missing data processing by adopting the following fitting formula:
wherein x is i Representing the ith dam front water level data to be fitted in the filtering window, y i Represents the ith fitted data, a 0 ，a 1 ，…，a k-1 K-1 represents fitting by using a k-1 polynomial, wherein k is a natural number and is more than or equal to 2;
wherein, a 0 ，a 1 ，…，a k-1 The solution of (c) is as follows:
if the length of the data sequence in the filtering window is N, N =2m +1, m is a natural number, and N is greater than or equal to k, N polynomials can be generated in the filtering window according to the fitting formula, where k variables to be solved, that is, k element linear equations, are determined in the N polynomials, and the least square formula is used:
can solve to obtain a 0 ，a 1 ，…，a k-1 。
5. The reservoir level fluctuation processing method according to claim 4, wherein:
the calculation mode of the filtering window data sequence length N is as follows:
dispersing the saint-winan equation set by adopting a Preissmann four-point weighted implicit differential format, then simplifying the equation set into a linear equation set by ignoring second-order micro quantity, solving the linear equation set, and obtaining the flow arrival time required by the sudden change flow to reach a downstream target reservoir when the flow of the upstream adjacent reservoir is suddenly changed, wherein the flow arrival time is divided by the dam front water level data sampling frequency, and the nearest odd number is taken upwards to be the filtering window data sequence length N.
6. The reservoir level fluctuation processing method according to claim 5, wherein:
the calculation mode of the radial flow hang-up rate in the step S5 is as follows:
wherein P represents the runoff inverse hanging rate, I t' Representing the upstream river inflow of the target reservoir in the period t', I t "represents the warehousing flow of the target reservoir in the time period T 'by the inverse deduction of the storage quantity change and the ex-warehouse flow, T' represents the time period, and T represents the total calculation time period;
wherein, I t' ＝O t'-j ，
Wherein O is t'-j The flow of the upstream river channel control node of the target reservoir is represented, namely the flow of the upstream adjacent reservoir, j represents the number of time periods required for the flow of the upstream river channel control node of the target reservoir to reach the target reservoir, and l t'+1 Representing the dam front water level of the target reservoir in T' +1 time period, f is the mapping relation between the dam front water level of the target reservoir and the reservoir capacity to convert the dam front water level into the reservoir capacity, T is the calculated total time period length, O t "is the flow of the target reservoir, and the calculation mode is
Wherein
The generated flow rate of the target reservoir in the period t' is shown,
the NHQ curve can be inquired according to the load data and the dam water level data for calculation,
representing the gate flow of the target reservoir in the period t',
the calculation can be carried out according to the gate opening and dam front water level data inquiry discharge curve.
7. The reservoir level fluctuation processing method according to claim 6, wherein:
in the P1 calculation process, the missing load data is filled by adopting the following formula:
wherein w represents missing load data in MW, w 1 、w 2 Load data adjacent to each other before and after the missing load data, with the unit of MW, t 1 ”、t' 2 'indicates a time scale corresponding to load data adjacent to the missing load data, and t' indicates a desired time scale of the missing load data.
8. The reservoir level fluctuation processing method according to claim 6, wherein:
and in the P1 calculation process, the missing gate opening data is filled by adopting the following formula:
g＝g 1
wherein g represents the missing gate opening data and the unit is m, g 1 The unit of the gate opening data adjacent to the missing gate opening data is m.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211481284.1A CN115712800B (en) | 2022-11-24 | 2022-11-24 | Reservoir water level fluctuation treatment method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211481284.1A CN115712800B (en) | 2022-11-24 | 2022-11-24 | Reservoir water level fluctuation treatment method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115712800A true CN115712800A (en) | 2023-02-24 |
| CN115712800B CN115712800B (en) | 2023-07-28 |
Family
ID=85234601
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211481284.1A Active CN115712800B (en) | 2022-11-24 | 2022-11-24 | Reservoir water level fluctuation treatment method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115712800B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117787566A (en) * | 2024-02-23 | 2024-03-29 | 国能大渡河流域水电开发有限公司 | Method, system, electronic equipment and storage medium for correcting unit flow characteristic curve |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008184838A (en) * | 2007-01-30 | 2008-08-14 | Hokkaido River Disaster Prevention Research Center | Dam inflow amount predicting device, dam inflow amount predicting method, and dam inflow amount predicting program |
| CN102155938A (en) * | 2011-04-07 | 2011-08-17 | 武汉大学 | Measuring method for inversing reservoir feeding flow procedures |
| CN108108838A (en) * | 2017-12-18 | 2018-06-01 | 华电福新能源股份有限公司福建分公司 | A kind of season balancing reservoir Optimization Scheduling of high water provenance |
| CN108537449A (en) * | 2018-04-12 | 2018-09-14 | 长江勘测规划设计研究有限责任公司 | Meter and river are passed the flood period the reservoir coordinated scheduling strategy acquisition methods of demand |
| CN109447336A (en) * | 2018-10-22 | 2019-03-08 | 南瑞集团有限公司 | Water level optimal control method between a kind of upper pond and its reregulating reservoir dam |
| JP2019094640A (en) * | 2017-11-20 | 2019-06-20 | 日本無線株式会社 | Water level prediction method, water level prediction program and water level prediction device |
| CN111328671A (en) * | 2020-03-05 | 2020-06-26 | 红河哈尼族彝族自治州水利水电工程地质勘察咨询规划研究院 | Reservoir photovoltaic pumping irrigation control system and method for realizing automatic frequency adjustment |
| CN111898253A (en) * | 2020-07-15 | 2020-11-06 | 武汉大学 | Reservoir scheduling and downstream river ecological environment protection cooperation value evaluation method |
| CN114494925A (en) * | 2022-02-18 | 2022-05-13 | 武汉大学 | Reservoir warehousing flow calculation method and device, electronic equipment and storage medium |
| CN115271380A (en) * | 2022-07-06 | 2022-11-01 | 四川华能宝兴河水电有限责任公司 | Self-adaptive hydropower station real-time water affair calculation method |
-
2022
- 2022-11-24 CN CN202211481284.1A patent/CN115712800B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008184838A (en) * | 2007-01-30 | 2008-08-14 | Hokkaido River Disaster Prevention Research Center | Dam inflow amount predicting device, dam inflow amount predicting method, and dam inflow amount predicting program |
| CN102155938A (en) * | 2011-04-07 | 2011-08-17 | 武汉大学 | Measuring method for inversing reservoir feeding flow procedures |
| JP2019094640A (en) * | 2017-11-20 | 2019-06-20 | 日本無線株式会社 | Water level prediction method, water level prediction program and water level prediction device |
| CN108108838A (en) * | 2017-12-18 | 2018-06-01 | 华电福新能源股份有限公司福建分公司 | A kind of season balancing reservoir Optimization Scheduling of high water provenance |
| CN108537449A (en) * | 2018-04-12 | 2018-09-14 | 长江勘测规划设计研究有限责任公司 | Meter and river are passed the flood period the reservoir coordinated scheduling strategy acquisition methods of demand |
| CN109447336A (en) * | 2018-10-22 | 2019-03-08 | 南瑞集团有限公司 | Water level optimal control method between a kind of upper pond and its reregulating reservoir dam |
| CN111328671A (en) * | 2020-03-05 | 2020-06-26 | 红河哈尼族彝族自治州水利水电工程地质勘察咨询规划研究院 | Reservoir photovoltaic pumping irrigation control system and method for realizing automatic frequency adjustment |
| CN111898253A (en) * | 2020-07-15 | 2020-11-06 | 武汉大学 | Reservoir scheduling and downstream river ecological environment protection cooperation value evaluation method |
| CN114494925A (en) * | 2022-02-18 | 2022-05-13 | 武汉大学 | Reservoir warehousing flow calculation method and device, electronic equipment and storage medium |
| CN115271380A (en) * | 2022-07-06 | 2022-11-01 | 四川华能宝兴河水电有限责任公司 | Self-adaptive hydropower station real-time water affair calculation method |
Non-Patent Citations (2)
| Title |
|---|
| 李贤成 等: "水库出库站水位一流量关系不稳定原因分析", vol. 7, no. 27, pages 40 - 41 * |
| 耿芳 等: "水库水位在线高精度测量方法研究", 《浙江水利科技》, no. 243, pages 22 - 25 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117787566A (en) * | 2024-02-23 | 2024-03-29 | 国能大渡河流域水电开发有限公司 | Method, system, electronic equipment and storage medium for correcting unit flow characteristic curve |
| CN117787566B (en) * | 2024-02-23 | 2024-05-07 | 国能大渡河流域水电开发有限公司 | Method, system, electronic equipment and storage medium for correcting unit flow characteristic curve |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115712800B (en) | 2023-07-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP4216117B1 (en) | Method and apparatus for training water-and-sediment prediction model for reservoir and method and apparatus for predicting water-and-sediment in reservoir | |
| CN108519129B (en) | Channel flow monitoring method and system | |
| WO2018026651A1 (en) | Method and apparatus for model-based control of a water distribution system | |
| CN111784536B (en) | Method for estimating water level overrun time in open channel according to actually measured water level change condition | |
| CN115712800A (en) | Reservoir water level fluctuation processing method | |
| CN107358542A (en) | A kind of parameter determination method of excitation system Performance Evaluation Model | |
| CN106598918B (en) | Nonuniformity methods for calculating designed flood based on quantile estimate | |
| CN109783934A (en) | A kind of mean velocity in section fitting rating method based on H-ADCP | |
| CN114580316A (en) | Small reservoir flood level forecasting method based on two-dimensional-zero-dimensional coupling model | |
| CN115640881A (en) | Reservoir water level correction method of reservoir warehousing runoff forecasting model based on LSTM | |
| CN109299830A (en) | A kind of hydrologic forecast system and method based on Soil Moisture Monitoring | |
| CN109523077B (en) | Wind power prediction method | |
| CN113672872A (en) | Plain river network water volume replacement rate calculation method based on remote sensing image and terminal | |
| CN106320257A (en) | Lake and reservoir channel storage curve determining method based on hydrometry | |
| CN117421951A (en) | River pollution tracing method | |
| CN114861571B (en) | Method and device for calculating dynamic boundary of river channel type reservoir and storage medium | |
| CN110059443B (en) | Rapid prediction method for water taking and discharging temperature of stratified reservoir | |
| CN114861473A (en) | Effective storage capacity analysis method and device for river channel type reservoir and electronic equipment | |
| Temeepattanapongsa et al. | Generic unified rating for Cutthroat flumes | |
| CN117010727B (en) | Runoff sequence consistency correction method based on horizontal years | |
| CN117313393B (en) | Calculation method of tidal range gate drainage flow process based on compensation water head difference | |
| CN115309199B (en) | Method and system for controlling basic flow of downstream channel of junction based on channel scale | |
| CELESTE et al. | Stochastic generation of inflow scenarios to be used by optimal reservoir operation models | |
| CN112699521B (en) | Method for analyzing annual daily change data of top overflowing height of water-proof curtain wall | |
| CN114169047B (en) | Method for calculating reservoir backwater range |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |