A Hydroelectric Scheduling Optimization Method and System Considering the Coupling of Cascade Reservoir Groups

Abstract

The present invention relates to the technical field of reservoir operation, and solves the technical problems in the prior art that the hydropower operation of reservoir groups has high energy consumption and the analysis results are not optimal strategies. In particular, the present invention relates to a hydropower operation optimization method and system considering the coupling of cascade reservoir groups. The method includes the following steps: S1. Obtain the power generation function E of each hydropower station in the cascade reservoir group max , and obtain a set of constraint conditions by constraining the relevant variables related to the power generation function E max ; S2. Take each hydropower station as a sample and divide multiple samples. The present invention comprehensively obtains the power generation situation of the cascade reservoir group through multiple restrictive conditions, more comprehensively measures the reservoir water level and power generation, can not only obtain the optimal power generation strategy, but also monitor the impact of the reservoir group water level on the ecology, ensure the survival of animals and plants around the reservoir, and improve the efficiency of reservoir group operation.

Description

Technical Field

[0001] The present invention relates to the technical field of reservoir operation, and particularly to a hydropower operation optimization method and system considering the coupling of cascade reservoir groups. Background Art

[0002] Cascade reservoir groups generally refer to a series of stepped reservoirs and hydropower stations built from the upstream to the downstream of a river or river section in the water conservancy and hydropower development plan of a river in order to make full use of water conservancy and hydropower resources. The formed group is the cascade reservoir group. The hydropower operation of the stepped reservoir group not only affects the power generation amount and power generation benefit, but also has an important impact on the ecology of the environment along both sides of the river. If the operation is unbalanced, it is easy to cause damage to the ecological environment, seriously affecting the lives of residents and the growth and survival of surrounding animals and plants. The existing hydropower operation optimization methods considering the coupling of cascade reservoir groups generally use a variety of sensors. However, when this method is used for the operation of multiple reservoirs, it is easy to have problems such as complex detection equipment, difficult maintenance and huge energy consumption. For the intelligent analysis model using the population algorithm, since only the optimization strategies in the same iteration are considered in the analysis process, and all optimization strategies are not referred to, the selected optimization strategies are not the optimal strategies. Summary of the Invention

[0003] Aiming at the deficiencies of the prior art, the present invention provides a hydropower operation optimization method and system considering the coupling of cascade reservoir groups, solves the technical problems of large energy consumption in the hydropower operation of reservoir groups and non-optimal analysis results in the prior art, and achieves the purpose of obtaining the optimal operation strategy through diversified operations and reducing energy consumption.

[0004] To solve the above technical problems, the present invention provides the following technical solutions: A hydropower operation optimization method considering the coupling of cascade reservoir groups, the method includes the following steps:

[0005] S1. Obtain the power generation function E of each hydropower station in the cascade reservoir group max and perform constraints on the relevant variables related to the power generation function E max to obtain a set of constraint conditions;

[0006] S2. Take each hydropower station as a sample, divide multiple samples and perform initialization respectively to obtain a first sample group and a second sample group

[0007] S3. Calculate the balance degree Ph of the samples according to the set of constraint conditions g ;

[0008] S4. Calculate the weight coefficient w of each sample according to the balance degree Ph g and respectively determine a first initial preferred set j ​ and the second initial preference set

[0009] S5. For the first initial preference set and the second initial preference set perform neighborhood diversification operations on each sample and obtain a diversified sample set

[0010] S6. Calculate the preference degree of the diversified sample set and based on the preference degree replace the samples in the first sample group and the second sample group to obtain the optimal set Z;

[0011] S7. Obtain the iteration threshold X through the maximum likelihood estimation method based on the historical data of the hydropower station and perform iteration on the optimal set Z;

[0012] If k + 1 = X, end the iteration and obtain the optimal scheduling strategy based on the optimal set Z;

[0013] If k + 1 < X, return to step S3.

[0014] Preferably, in step S1, the specific implementation steps are as follows:

[0015] S11. Obtain the relevant variables of each hydropower station in the cascade reservoir group at different time periods t after preprocessing. The relevant variables include the output coefficient A a , the generated flow rate L a，t , the water head H a，t and the number of hours t j ;

[0016] S12. Calculate the power generation function E max , and the calculation formula is:

[0017]

[0018] where A a represents the output coefficient of the a-th hydropower station, L a，t represents the generated flow rate of the a-th hydropower station in the t-th time period, H a，t represents the water head of the a-th hydropower station in the t-th time period, and t j represents the j-th number of hours;

[0019] S13. Define the constraint conditions for the generated flow rate L a，t and the water head H a，t , and the expression is:

[0020] ​

[0021] Among them, and L a，t respectively represent the upper and lower limits of the power generation flow of the a-th hydropower station in the t-th time period, and H a，t respectively represent the upper and lower limits of the water head of the a-th hydropower station in the t-th time period;

[0022] S14. The constraint conditions of related variables also include the water level constraint on the water level SW a，t and the bandwidth constraint on the hydropower bandwidth DK a，t , and a constraint condition set is generated, and the expression is:

[0023]

[0024] Among them, and SW a，t respectively represent the upper and lower limits of the water level of the a-th hydropower station in the t-th time period, and DK a，t respectively represent the upper and lower limits of the hydropower bandwidth of the a-th hydropower station in the t-th time period.

[0025] Preferably, in step S2, the specific implementation steps are as follows:

[0026] S21. Obtain the power generation benefit DY b of each sample, and calculate the average value b of the power generation benefit DY . The calculation formula is:

[0027]

[0028] Among them, represents the average value of the power generation benefit DY b , DY b represents the b-th power generation benefit, and n represents the number b of the power generation benefit DY;

[0029] S22. Divide multiple samples according to the average value and the sample DY b ;

[0030] If then it is divided into the first sample set;

[0031] If then it is divided into the second sample set;

[0032] S23. Obtain the data quantity E c of the first sample set and the data quantity E d of the second sample set;

[0033] S24. Calculate the decision values of the upper and lower limits of the data in the relevant variables respectively and Jc f , and the calculation formula is as follows:

[0034]

[0035] where and Jc f represent the decision values of the upper and lower limits of the data in the relevant variables respectively;

[0036] S25. Perform an initialization operation on the first sample set and the second sample set to obtain the first sample population and the second sample population The expression is:

[0037]

[0038] where represents the e-th sample in the k-th iteration of the first sample population, represents the f-th sample in the k-th iteration of the second sample population, and β1 and β2 represent random numbers uniformly distributed in the interval [0, 1].

[0039] Preferably, in step S3, the specific implementation steps are as follows:

[0040] S31. Arbitrarily select a certain time period t d , and obtain the values within the constraint condition set during the time period t d , including the generated power flow the water head the water level and the hydropower bandwidth

[0041] S32. Obtain different balance factors B that exceed the constraint condition set through the regularization method cm ;

[0042] If or then obtain the balance factor B c1 ;

[0043] If or then obtain the balance factor B c2 ;

[0044] If or then obtain the balance factor B c3 ;

[0045] If or then obtain the balance factor Bc4 ;

[0046] S33. Take the union of the first sample group and the second sample group as the overall sample The expression is:

[0047]

[0048] where represents the overall sample.

[0049] S34. Calculate the balance degree Ph of each sample in the overall sample max according to the power generation function E and multiple balance factors. The calculation formula is: g The calculation formula is:

[0050]

[0051] where Ph g represents the balance degree of the sample, B cm represents the m-th balance factor, the relevant variable value of the sample exceeding the constraint condition, and M represents the number of balance factors B cm .

[0052] Preferably, in step S4, the specific implementation steps are as follows:

[0053] S41. Calculate the weight coefficient w of each sample in the overall sample g according to the balance degree Ph . The calculation formula is as follows: j The calculation formula is as follows:

[0054]

[0055] where w j represents the weight coefficient of each sample in the overall sample , represents the balance degree of the h-th sample in the k-th iteration in the overall sample, and min and max represent selecting the minimum value and the maximum value respectively among multiple balance degrees;

[0056] S42. Determine the optimal sample in the overall sample according to the balance degree Ph g . The expression is: The expression is:

[0057]

[0058] where represents the overall sample with the balance degree Phg The largest sample;

[0059] S43. Arbitrarily select a sample from the overall sample as the central sample

[0060] S44. Determine a reference sample closest to the central sample near the central sample

[0061] S45. For the first sample group and the second sample group sort them respectively in descending order according to the balance degree Ph g and take the first 50% with larger balance degree Ph g as the first initial preferred set and the second initial preferred set

[0062] Preferably, in step S5, the specific implementation steps are as follows:

[0063] S51. Calculate the perception coefficients V of the optimal sample and the central sample a and V b , and the calculation formula is:

[0064]

[0065] where V a and V b respectively represent the perception coefficients of the optimal sample and the central sample , and respectively represent the Euclidean distances between the optimal sample and the reference sample , and the Euclidean distances between the central sample and the reference sample ;

[0066] S52. Obtain a uniformly distributed random threshold Q f and random numbers g1, g2, g3, and g4 between [0, 1] through a random number generator;

[0067] S53. Perform diversification operations on the first initial preferred set ;

[0068] If g4 ≤ Q f , then the calculation formula is: ​

[0069]

[0070] Among them, represents the first initial optimal set after diversification operations;

[0071] If g4 > Q f , then the calculation formula is:

[0072]

[0073] Among them, g1, g2, g3, and g4 respectively represent random numbers between [0, 1];

[0074] S54. Perform diversification operations on the second initial priority set .

[0075] If g4 ≤ Q f , then the calculation formula is:

[0076]

[0077] Among them, represents the second initial optimal set after diversification operations;

[0078] If g4 > Q f , then the calculation formula is:

[0079]

[0080] Among them, V a and V b respectively represent the perception coefficients of the optimal sample and the central sample ;

[0081] S55. Obtain the diversified sample set from the first initial optimal set after diversification operations and the second initial optimal set

[0082] Preferably, in step S6, the specific implementation steps are as follows:

[0083] S61. Calculate the preference degree g of each sample in the diversified sample set according to the calculation formula of the balance degree Ph The calculation formula is:

[0084]

[0085] Among them, represents the preference degree of the diversified sample set Edy Represents the power generation function corresponding to a diverse sample set B dy Represents the balance factor corresponding to a diverse sample set u represents the balance factor B dy The number of;

[0086] S62. According to the preference degree Replace the samples of the overall sample ;

[0087] If Then Is not the optimal value, replace the corresponding overall sample

[0088] If Then Is the optimal value, do not replace the overall sample ;

[0089] S63. Repeat step S62 until all the samples in the overall sample Are replaced and the optimal set Z is obtained.

[0090] Preferably, the specific content of step S7 is: Arrange the samples in the optimal set Z in descending order according to the preference degree And take the sample with the largest preference degree As the optimal scheduling strategy.

[0091] This technical solution also provides a system for a hydropower scheduling optimization method, and the system includes:

[0092] A power generation constraint module for obtaining the power generation function E of each hydropower station in the cascade reservoir group max And constraining the relevant variables related to the power generation function E max To obtain a constraint condition set;

[0093] An initialization module for taking each hydropower station as a sample, dividing multiple samples and initializing them respectively to obtain a first sample group And a second sample group

[0094] A balance degree module for calculating the balance degree Ph of the samples according to the constraint condition set g ;

[0095] An initial preference module for calculating the weight coefficient w of each sample according to the balance degree Ph g And respectively determining a first initial preference set j And a second initial preference set ​

[0096] A diversification module for performing diversification operations on the neighborhood of each sample in the first initial preference set and the second initial preference set to obtain a diversified sample set

[0097] An optimal selection module for calculating the preference degree of the diversified sample set and replacing the samples in the first sample group and the second sample group according to the preference degree to obtain the optimal set Z; An iteration module for obtaining an iteration threshold X by the maximum likelihood estimation method based on the historical data of the hydropower station and iterating the optimal set Z.

[0098] With the above technical solutions, the present invention provides a hydropower scheduling optimization method and system considering the coupling of cascade reservoir groups, which at least have the following beneficial effects:

[0099]

[0100]

[0101] 1. The present invention calculates the balance degree through multiple constraint conditions, can comprehensively obtain the power generation situation of the cascade reservoir group, more comprehensively measure the reservoir water level and power generation amount, and is convenient for controlling the power generation amount and water level, thereby controlling the ecological environment around the reservoir group. It can not only obtain the optimal power generation strategy, but also monitor the impact of the reservoir group water level on the ecology, ensure the survival of animals and plants around the reservoir, and improve the efficiency of reservoir group scheduling.

[0102] 2. By performing diversification operations on the reservoir group, the present invention can make the balance degree of the reservoir group not only refer to the data of samples within the same set, but also refer to the data of all samples, accelerating the convergence of the model towards the optimal solution. While improving the data diversification and globality of the reservoir group, it can also improve the acquisition speed of the optimal strategy and shorten the time for obtaining the optimal strategy.

[0102] 3. Through the calculation and replacement of the preference degree of diversified samples, the present invention can screen and optimize samples with lower preference degrees, replace the optimization strategies with lower preference degrees after each iteration, so as to quickly obtain the optimal set, and improve the working efficiency of obtaining the optimal strategy. Description of the Drawings

[0103] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The schematic embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation to the present application. In the drawings:

[0104] Figure 1Flow chart of a hydropower scheduling optimization method considering the coupling of cascade reservoir groups according to the present invention;

[0105] Figure 2 Block diagram of the structure of a hydropower scheduling optimization system considering the coupling of cascade reservoir groups according to the present invention. Detailed implementation manners

[0106] To make the above objects, features, and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific implementation manners. Thus, the implementation process of how the present application uses technical means to solve technical problems and achieve technical effects can be fully understood and implemented accordingly.

[0107] Due to the technical problems in the prior art that the hydropower scheduling of reservoir groups has high energy consumption and the analysis results are not the optimal strategies, please refer to Figure 1 and Figure 2 , this embodiment provides a hydropower scheduling optimization method considering the coupling of cascade reservoir groups, which can obtain the optimal scheduling strategy through diversified operations and reduce energy consumption. The method includes the following steps:

[0108] S1. Obtain the power generation function E of each hydropower station in the cascade reservoir group max , and obtain a set of constraint conditions by constraining the relevant variables related to the power generation function E max ; In the calculation of power generation, it is also necessary to consider factors such as water level and flow rate related to power generation and ecological factors. In step S1, the specific implementation steps are as follows:

[0109] S11. Obtain the relevant variables of each hydropower station in the cascade reservoir group at different time periods t after preprocessing. The relevant variables include the output coefficient A a , the power generation flow rate L a，t , the water head H a，t , and the number of hours t j ; The output coefficient refers to the ratio of the average total output power of a unit generator set or a power station to the rated power of the generator set or the power station during the statistical period. The power generation flow rate refers to the amount of water used for power generation per unit time. The water head refers to the mechanical energy of unit weight of water at any cross-section of the water flow. The number of hours is generally one hour as a measurement unit to measure the relevant variables, and multiple physical quantities in different time periods can be generated during the measurement.

[0110] S12. Calculate the power generation function E according to the relevant variables max , and the calculation formula is:

[0111]

[0112] Among them, A a represents the output coefficient of the a-th hydropower station, and L a，tDenote the power generation flow rate in the t-th time period of the a-th hydropower station, H a，t Denote the water head in the t-th time period of the a-th hydropower station, t j Denote the j-th hour number;

[0113] S13. For the power generation flow rate L a，t and the water head H a，t impose the constraint conditions, and the expression is:

[0114]

[0115]

[0116] Among them, and L a，t respectively represent the upper and lower limits of the power generation flow rate of the a-th hydropower station in the t-th time period, and H a，t respectively represent the upper and lower limits of the water head of the a-th hydropower station in the t-th time period; impose condition restrictions on the power generation flow rate L a，t and the water head H a，t of the hydropower stations in multiple reservoir groups. Once these conditions are exceeded, operations will be carried out through balance factors in subsequent calculation steps. The setting of the condition range can be based on the historical power generation flow rate L a，t and the water head H a，t of the hydropower station as a reference, and the range is determined by the empirical method of local staff.

[0117] S14. The constraint conditions of related variables also include the water level constraint on the water level SW a，t and the bandwidth constraint on the hydropower bandwidth DK a，t , and generate a set of constraint conditions, and the expression is:

[0118]

[0119] Among them, and SW a，t respectively represent the upper and lower limits of the water level of the a-th hydropower station in the t-th time period, and DK a，t respectively represent the upper and lower limits of the hydropower bandwidth of the a-th hydropower station in the t-th time period. The constraint on the water level SW a，t and the hydropower bandwidth DK a，t generally can affect the power generation situation through the water level and the hydropower bandwidth. Therefore, it is also necessary to impose constraints on the water level SW a，t and the hydropower bandwidth DK a，t .

[0120] In this embodiment, by constraining various physical quantities around the hydropower station, it is convenient to comprehensively detect the overall situation of the reservoir group, and it is also convenient for the implementation of subsequent steps. By calculating the balance degree through multiple limiting conditions, the power generation situation of the cascade reservoir group can be obtained comprehensively, and the reservoir water level and power generation can be measured more comprehensively. Moreover, by controlling the power generation and water level, the ecological environment around the reservoir group can be controlled, so that the optimal power generation strategy can be obtained, and the impact of the reservoir group water level on the ecology can be monitored, ensuring the survival of animals and plants around the reservoir and improving the efficiency of reservoir group scheduling.

[0121] S2. Take each hydropower station as a sample, divide multiple samples and initialize them respectively to obtain the first sample group and the second sample group In order to enable flexible calculation and analysis of hydropower dispatching, multiple hydropower stations in the reservoir group are taken as samples. In step S2, the specific implementation steps are as follows:

[0122] S21. Obtain the power generation benefit DY of each sample b , calculate the average value of the power generation benefit DY b The formula is: The calculation formula is:

[0123]

[0124] Among them, represents the average value of the power generation benefit DY b DY b represents the b-th power generation benefit, and n represents the number of power generation benefits DY b The power generation benefit generally refers to the economic value and social benefits created by the hydropower station providing electricity, power and playing other functions for society. The samples can be divided according to the situation of the power generation benefit.

[0125] S22. Divide multiple samples according to the average value of DY b ;

[0126] If then it is divided into the first sample set;

[0127] If then it is divided into the second sample set;

[0128] S23. Obtain the data quantity E c of the first sample set and the data quantity F d of the second sample set;

[0129] S24. Calculate the decision values and Jc f of the upper and lower limits of the data in the relevant variables respectively. The calculation formula is as follows:

[0130]

[0131] Among them, and Jc f respectively represent the decision values of the upper and lower limits of the data in the relevant variables; the decision values of the upper and lower limits of the data are the average values of the upper or lower limits of multiple relevant variables in the constraint set.

[0132] S25. Perform initialization operations on the first sample set and the second sample set to obtain the first sample population and the second sample population The expression is:

[0133]

[0134] Among them, represents the e-th sample in the k-th iteration of the first sample population, represents the f-th sample in the k-th iteration of the second sample population, and β1 and β2 respectively represent random numbers uniformly distributed in the interval [0, 1]. The random numbers are generally generated by a random generator and meet the uniform distribution conditions.

[0135] Through this step in this embodiment, the samples can be partitioned and the sample populations can be initialized, facilitating the operations in subsequent steps and improving the accuracy and operation efficiency of the model.

[0136] S3. Calculate the balance degree Ph of the samples according to the constraint set g ; The balance degree is the power generation situation of the hydropower station under the constraints of the constraint set. The regulation calculation operation is completed by the water-determined power method for each hydropower station in turn from upstream to downstream according to the hydraulic connection. During the processing, it is forced to meet the constraints with the constraint conditions. In step S3, the specific implementation steps are as follows:

[0137] S31. Arbitrarily select a certain time period t d , and obtain the values in the constraint set within the time period t d , including the power generation flow head water level and the hydropower bandwidth

[0138] S32. Obtain different balance factors B that exceed the constraint set through the regularization method cm ;

[0139] If or then obtain the balance factor B c1 ;

[0140] If or Then obtain the balance factor B c2 ;

[0141] If or Then obtain the balance factor B c3 ;

[0142] If or Then obtain the balance factor B c4 ; After exceeding the condition, it indicates that the quantity of this item does not meet the requirements and needs to be adjusted and calculated through the balance factor. The balance factor is the penalty factor after the constraint is violated.

[0143] S33. Use the union of the first sample group and the second sample group as the overall sample The expression is:

[0144]

[0145] Among them, means that the overall sample combines the two sample groups for the calculation of the balance degree.

[0146] S34. Calculate the balance degree Ph of each sample in the overall sample max according to the power generation function E and multiple balance factors g , and the calculation formula is:

[0147]

[0148] Among them, Ph g represents the balance degree of the sample, B cm represents the mth balance factor, the relevant variable value of the sample that exceeds the constraint condition, M represents the number of balance factors B cm .

[0149] In this embodiment, through the calculation of the balance degree Ph g , multiple constraint conditions can be related, and a more reasonable analysis of hydropower scheduling can be carried out. By calculating the balance degree through multiple limiting conditions, the power generation situation of the cascade reservoir group can be obtained comprehensively, the reservoir water level and power generation can be measured more comprehensively, and it is convenient to control the power generation and water level, and then control the ecological environment around the reservoir group. It can not only obtain the optimal power generation strategy, but also monitor the impact of the reservoir group water level on the ecology, ensure the survival of animals and plants around the reservoir, and improve the efficiency of reservoir group scheduling.

[0150] S4. Calculate the weight coefficient w of each sample according to the balance degree Ph g ​j , and respectively determine the first initial preferred set and the second initial preferred set Since the balance degree is not sufficient for a more in-depth analysis of the hydropower station scheduling, in step S4, the specific implementation steps are as follows:

[0151] S41. Calculate the weight coefficient w of each sample in the overall sample g according to the balance degree Ph , and the calculation formula is as follows: j

[0152]

[0153] where w j represents the weight coefficient of each sample in the overall sample , represents the balance degree of the h-th sample in the k-th iteration in the overall sample, and min and max represent respectively selecting the minimum value and the maximum value among multiple balance degrees; calculating the weight coefficient of each sample through the balance degree is helpful for the operations in the subsequent steps.

[0154] S42. Determine the optimal sample in the overall sample according to the balance degree Ph g , and the expression is:

[0155]

[0156] where represents the sample with the maximum balance degree Ph in the overall sample; g

[0157] S43. Arbitrarily select a sample as the central sample in the overall sample

[0158] S44. Determine a reference sample nearest to the central sample near the central sample

[0159] S45. Sort the first sample group and the second sample group respectively in descending order according to the balance degree Ph g , and take the first 50% with larger balance degree Ph g as the first initial preferred set and the second initial preferred set ​​​Perform a secondary division of the overall sample here to avoid overlooking some relatively excellent samples. Analyzing the selected samples can obtain better strategies.

[0160] S5. For the first initial selected set and the second initial selected set perform a diversification operation on each sample in them, and obtain a diversified sample set Since the analysis results of the selected samples are not extensive enough, it is necessary to perform a diversification operation on the samples to improve the diversity of the samples. In step S5, the specific implementation steps are as follows:

[0161] S51. Calculate the perception coefficients V and V of the optimal sample a and the central sample b , and the calculation formula is:

[0162]

[0163] where V a and V b represent the perception coefficients of the optimal sample and the central sample respectively, and represent the Euclidean distances between the optimal sample and the reference sample , and the Euclidean distances between the central sample and the reference sample respectively; the Euclidean distance generally refers to the Euclidean metric or Euclidean distance, which is the actual distance between two points in an m-dimensional space. The desired result can be directly obtained through the Euclidean distance, and the Euclidean distance is a common distance calculation method, which will not be elaborated here.

[0164] S52. Obtain a uniformly distributed random threshold Q f and random numbers g1, g2, g3, and g4 between [0, 1] through a random number generator respectively; in order to increase the diversity of the samples, the random numbers can assist in achieving this effect. The random number generator is a method that can obtain random numbers within a certain range and can be implemented without additional steps, which will not be elaborated here.

[0165] S53. Perform a diversification operation on the first initial selected set ;

[0166] If g4 ≤ Q f , then the calculation formula is:

[0167]

[0168] Among them, represents the first initial preferred set after diversification operations;

[0169] If g4 > Q f , then the calculation formula is:

[0170]

[0171] Among them, g1, g2, g3, and g4 respectively represent random numbers between [0, 1];

[0172] S54. Perform diversification operations on the second initial preferred set ;

[0173] If g4 ≤ Q f , then the calculation formula is:

[0174]

[0175] Among them, represents the second initial preferred set after diversification operations;

[0176] If g4 > Q f , then the calculation formula is:

[0177]

[0178] Among them, V a and V b respectively represent the perception coefficients of the optimal sample and the central sample ; In the diversification operations, it includes both the current iterative samples and all samples, and can comprehensively perform diversification operations.

[0179] S55. Obtain the diversified sample set from the first initial preferred set after diversification operations and the second initial preferred set

[0180] In this embodiment, by performing diversification operations on the reservoir group, the balance degree of the reservoir group can not only refer to the data of samples within the same set, but also refer to the data of all samples, accelerating the convergence of the model towards the optimal solution. While improving the data diversification and global nature of the reservoir group, it can also improve the acquisition speed of the optimal strategy and shorten the time for obtaining the optimal strategy.

[0181] S6. Calculate the preference degree of the diversified sample set and, according to the preference degree of the first sample group and the second sample group Replace the samples of and obtain the optimal set Z; Since the situation of each sample has changed after the diversification operation, in step S6, the specific implementation steps are as follows:

[0182] S61. Calculate the preference degree of each sample in the diversification sample set g according to the calculation formula of the balance degree Ph The calculation formula is:

[0183]

[0184] Among them, represents the preference degree of the diversification sample set E dy represents the power generation function corresponding to the diversification sample set B dy represents the balance factor corresponding to the diversification sample set u represents the number of the balance factor B dy Calculate the preference degree of the diversification sample set by the formula of the balance degree and use the preference degree as a reference.

[0185] S62. Replace the samples of the overall sample according to the preference degree ;

[0186] If then is not the optimal value, replace the corresponding overall sample

[0187] If then is the optimal value, do not replace the overall sample ;

[0188] S63. Repeat step S62 until all the samples in the overall sample are replaced and the optimal set Z is obtained. Through the comparison and replacement of the preference degrees, the optimal selection of the samples can be completed, and the excellent samples can be left. By this method, the screening and optimization of the samples with lower preference degrees can be realized, so that the optimization strategy after each iteration replaces the strategy with lower preference degrees to quickly obtain the optimal set and improve the working efficiency of obtaining the optimal strategy.

[0189] S7. Obtain the iteration threshold X by the maximum likelihood estimation method according to the historical data of the hydropower station and perform iteration on the optimal set Z;

[0190] ​If k + 1 = X, terminate the iteration and obtain the optimal scheduling strategy based on the optimal set Z;

[0191] If k + 1 < X, return to step S3. The specific content of step S7 is: Arrange the samples in the optimal set Z in descending order according to the preference degree and take the sample with the highest preference degree as the optimal scheduling strategy. Here, since the number of iterations will not be greater than the iteration threshold X, there is no case of k + 1 > X in the judgment step.

[0192] Through this method, the optimal sample can be selected in this embodiment. The samples in this method represent the set of various physical quantities of hydropower stations in the reservoir group, and the optimal sample represents the optimal scheduling strategy of the hydropower station. By calculating and replacing the preference degrees of diverse samples, the screening and optimization of samples with lower preference degrees can be achieved, enabling the optimization strategy after each iteration to replace the strategy with lower preference degrees to quickly obtain the optimal set and improving the working efficiency of obtaining the optimal strategy.

[0193] Please refer to Figure 2 , which shows the structural block diagram of the hydropower scheduling optimization system provided in this embodiment. The hydropower scheduling optimization system includes a power generation constraint module, an initialization module, a balance degree module, an initial preference module, a diversification module, an optimal preference module, and an iteration module.

[0194] The power generation constraint module is used to obtain the power generation function E max of each hydropower station in the cascade reservoir group, max and perform constraints on the related variables of the power generation function E to obtain a set of constraint conditions. The initialization module is used to take each hydropower station as a sample, divide multiple samples, and initialize them separately to obtain the first sample group and the second sample group g The balance degree module is used to calculate the balance degree Ph g of the sample according to the set of constraint conditions. The initial preference module is used to calculate the weight coefficient w j of each sample according to the balance degree Ph and respectively determine the first initial preference set and the second initial preference set The diversification module is used to perform diversification operations on the neighborhood of each sample in the first initial preference set and the second initial preference set to obtain a diversified sample set The optimal preference module is used to calculate the preference degree of the diversified sample set and according to the preference degree update the first sample group Replace the samples and obtain the optimal set Z. An iterative module is used to obtain the iterative threshold X by the maximum likelihood estimation method based on the historical data of the hydropower station and iterate on the optimal set Z.

[0195] Those of ordinary skill in the art can understand that all or part of the steps in implementing the above-described embodiment methods can be completed by instructing relevant hardware through a program. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0196] The above embodiments have introduced the present invention in detail. Specific examples are used in this article to elaborate on the principles and embodiments of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A hydropower scheduling optimization method considering the coupling of cascade reservoir groups, characterized in that, The method comprises the following steps: S1. Obtain the power generation function of each hydropower station in the cascade reservoir group , and perform constraints on the relevant variables related to the power generation function to obtain a set of constraint conditions. The specific implementation steps are as follows: S11. Obtain the relevant variables of each hydropower station in the cascade reservoir group at different time periods after preprocessing The relevant variables include the output coefficient , the power generation flow , the water head and the number of hours ; S12. Calculate the power generation function based on relevant variables , and the calculation formula is as follows: ; among them, represents the output coefficient of the th hydropower station, represents the power generation flow rate of the th hydropower station during the th time period, represents the water head of the th hydropower station during the th time period, represents the th number of hours; S13. Define the constraint conditions for the power generation flow rate and the water head as follows: ; ; wherein, and respectively represent the upper and lower limits of the power generation flow of the th hydropower station in the time period, and respectively represent the upper and lower limits of the water head of the th hydropower station in the time period; S14. The constraint conditions of relevant variables also include the water level constraint on the water level and the bandwidth constraint on the hydropower bandwidth , and a set of constraint conditions is generated, and the expression is: ; ; wherein, and respectively represent the upper and lower limits of the water level of the th hydropower station in the time period, and respectively represent the upper and lower limits of the hydropower bandwidth of the th hydropower station in the time period; S2. Take each hydropower station as a sample, divide multiple samples, and initialize them separately to obtain the first sample group and the second sample group ; S3. Based on the first sample group and the second sample group to determine the overall sample , and calculate the balance degree of each sample in the overall sample according to the set of constraint conditions ; S4. According to the balance degree Calculate the weight coefficient of each sample , and respectively determine the first initial preferred set and the second initial preferred set ; S5. Perform a diversification operation on the neighborhood of each sample in the first initial preferred set and the second initial preferred set to obtain a diversified sample set ; S6. Calculate the diversity sample set of the preference degree , and according to the preference degree replace the samples of the first sample group and the second sample group , and obtain the optimal set ; S7. Obtain the iterative threshold by the maximum likelihood estimation method based on the historical data of the hydropower station , and perform iteration on the optimal set ; If the number of iterations , then end the iteration and obtain the optimal scheduling strategy based on the optimal set . If the number of iterations , return to step S3.

2. The optimized method for hydropower scheduling according to claim 1, wherein In step S2, the specific implementation steps are as follows: S21. Obtain the power generation benefits of each sample , and calculate the average value of the power generation benefits . The calculation formula is as follows: ​ ; wherein, represents the average value of the power generation benefit , represents the -th power generation benefit, represents the power generation benefit quantity; S22. Divide multiple samples according to the average value and the samples for division; If , it is divided into the first sample set; If , it is divided into the second sample set; S23. Obtain the data quantity of the first sample set by the counting method and the data quantity of the second sample set ; S24. Calculate the decision values of the upper and lower limits of the data in the relevant variables respectively and , and the calculation formula is as follows: ； ; wherein, and respectively represent the decision values of the upper and lower limits of the data in the relevant variables; S25. Initialize the first sample set and the second sample set to obtain the first sample population and the second sample population , the expression is: ; ; where, represents the th iteration and the th sample in the first sample group, represents the th iteration and the th sample in the second sample group, and respectively represent random numbers uniformly distributed within the interval.

3. The optimized method for hydropower dispatching according to claim 1, characterized in that In step S3, the specific implementation steps are as follows: S31. Arbitrarily select a certain time period , and obtain the values within the constraint condition set during the time period , including the power generation flow , water head , water level and the hydropower bandwidth ; S32. Obtain different balance factors outside the constraint condition set by the regularization method ; If or , then obtain the balance factor ; If or , then obtain the balance factor ; If or , then obtain the balance factor ; If or , then obtain the balance factor ; S33. Take the union of the first sample group and the second sample group as the overall sample . The expression is: ; wherein, represents the overall sample; S34. Calculate the balance degree of each sample in the overall sample according to the power generation function and multiple balance factors, and the calculation formula is: ; wherein, represents the balance of the sample, represents the th balance factor, the relevant variable value of the sample exceeding the constraint condition, represents the balance factor quantity.

4. The hydropower dispatching optimization method according to claim 1, characterized in that In step S4, the specific implementation steps are as follows: S41. Calculate the weight coefficient of each sample in the overall sample according to the balance degree, and the calculation formula is as follows: ​ ; wherein, represents the weight coefficient of each sample in the overall sample ; represents the balance degree of the th iteration and the th sample in the overall sample, and represent respectively selecting the minimum value and the maximum value among multiple balance degrees; S42. Among the overall samples determine the optimal sample according to the balance degree , and the expression is: ​ ; among them, represents the overall sample with the highest balance in the sample; S43. Among the overall samples arbitrarily select a sample as the central sample first ; S44. Determine a reference sample closest to the central sample near the central sample ; ; S45. For the first sample group and the second sample group sort them separately according to the balance degree in the order from large to small, and take the top 50% of the balance degree as the first initial preferred set and the second initial preferred set .

5. The hydroelectric power dispatch optimization method according to claim 1, wherein In step S5, the specific implementation steps are as follows: S51. Calculate the perception coefficients of the optimal sample and the central sample respectively, and the calculation formula is: and ​ ; ; wherein, and respectively represent the perception coefficients of the optimal sample and the central sample ; and respectively represent the Euclidean distances between the optimal sample and the reference sample , and the Euclidean distances between the central sample and the reference sample ; S52. Between respectively obtain a uniformly distributed random threshold value and a random number ; S53. Diversify the first initial preferred set for diversification operations; If , the calculation formula is as follows: ; wherein, represents the first initial preferred set after diversified operations; If , the calculation formula is as follows: ; wherein, respectively represent random numbers between ; S54, perform diversification operations on the second initial priority set ; If , the calculation formula is as follows: ; wherein, represents the second initial optimal set after diversified operations; If , the calculation formula is as follows: ; wherein, and respectively represent the perception coefficients of the optimal sample and the central sample ; S55. Obtain a diversified sample set based on the first initial preferred set and the second initial preferred set after diversified operations. .

6. The optimized method for hydropower dispatching according to claim 1, characterized in that In step S6, the specific implementation steps are as follows: S61. Calculate the preference degree of each sample in the diversified sample set according to the calculation formula of the balance degree , and the calculation formula is: ​ ; wherein, represents the preference degree of the diversified sample set ; represents the power generation function corresponding to the diversified sample set ; represents the balance factor corresponding to the diversified sample set ; represents the number of balance factors ; S62. Replace the samples of the overall sample according to the preference degree for the overall sample ; If ≥ , then is not the optimal value, replace the corresponding overall sample ; If , then is the optimal value and does not replace the overall sample . S63. Repeat step S62 until all samples in the overall sample are replaced and the optimal set is obtained .

7. The optimized method for hydropower scheduling according to claim 1, characterized in that The specific content of step S7 is: arrange the samples in the optimal set in descending order according to the preference degree , and take the sample with the highest preference degree as the optimal scheduling strategy.

Images