A mid- and long-term optimal dispatching method for cascade hydropower stations taking into account the decomposition of contracted electricity

By adopting the goal of the minimum risk of power generation returns with a given confidence in the cascade hydropower group, the market share trade-off in the medium- and long-term trading plan is solved by using the random incoming process and electricity price simulation, combined with step-by-step optimization and successive approximation algorithms. The problem improves the rationality and profit stability of the trading plan.

CN114676888BActive Publication Date: 2024-11-08CHINA THREE GORGES UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210213000.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-03-04
Publication Date
2024-11-08
Estimated Expiration
2042-03-04

AI Technical Summary

Technical Problem

When formulating medium- and long-term trading plans, it is difficult to weigh market share at different time scales, resulting in annual contract electricity that may damage returns due to low electricity prices, and monthly contract electricity may not be completed due to insufficient power generation capacity. A few days ago or short-term contracts The electricity volume may fail due to fluctuations in electricity prices.

Method used

The goal is to use the minimum risk of the entire cascade power generation with a given confidence level. Java's Random.nextGaussian() method is used to generate a random monthly water incoming process that meets normal distribution. Three electricity price functions are simulated according to the different proportion of hydropower to market. , gradual optimization and successive approximation hybrid algorithms are used to solve the power station output decision and contract power decomposition cycle iteratively.

Benefits of technology

It effectively deals with the problem of insufficient connection between long-term optimization and market transactions in the cascade hydropower station group, reduces excessive and minor power generation, improves water resource utilization efficiency, and reduces profit risks.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114676888B_ABST
    Figure CN114676888B_ABST
Patent Text Reader

Abstract

The present invention provides a medium- and long-term optimization scheduling method for a cascade hydropower station group that takes into account the decomposition of contracted electricity, and adopts the goal of minimizing the risk of the entire cascade power generation revenue under a given confidence level; first, the software is used to generate a random monthly water inflow process that conforms to the normal distribution; then, three electricity price functions are simulated according to the different market proportions of hydropower electricity; different water inflow frequencies and market electricity prices are selected, and a step-by-step optimization and successive approximation hybrid algorithm is used to iteratively solve the power station output decision and contracted electricity decomposition. It has guiding significance for solving the problems of difficulty in determining different market shares due to the uncertainty factors of natural water inflow and transaction electricity prices, insufficient utilization of water resources, and uneven distribution of electricity leading to greater risks in revenue. This achievement provides a solid theoretical foundation for the construction of a monthly transaction electricity decomposition and verification system for large-scale hydropower stations.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of hydropower station dispatch optimization, and in particular relates to a medium- and long-term optimization dispatch method for a cascade hydropower station group taking into account the decomposition of contracted electricity. Background Art

[0002] With the continuous expansion of hydropower in southwest my country and the improvement of relevant policies in the power market, large-scale hydropower participating in market competition is an inevitable trend to achieve optimal resource allocation. Its trading mode is mainly based on medium- and long-term transactions, supplemented by short-term transactions on the day before. Especially in the medium and long term, due to the characteristics of uncertain natural water supply, limited reservoir regulation capacity, close hydropower connection, and large fluctuations in electricity prices, it is difficult for cascade hydropower stations to weigh the market share of different time scales when formulating trading plans. This is particularly evident in the following aspects: too much annual contract electricity may cause revenue loss due to too low electricity prices; too much monthly contract electricity may cause the contract electricity to be unable to be completed due to insufficient power generation capacity of the power station; too much day-ahead or short-term contract electricity may cause transaction failure due to abnormal fluctuations in electricity prices. Therefore, how to formulate reasonable trading electricity for different time scales under the market economy has become a theoretical and practical problem that needs to be solved for cascade hydropower stations to participate in market competition.

[0003] At present, domestic and foreign scholars have conducted a lot of research on the problem of contract power decomposition in the power market environment, mainly focusing on three aspects: ① The problem of power decomposition in non-hydropower systems, such as annual or monthly power decomposition with the goal of minimizing the deviation between the planned scheme and the actual scheme, minimizing the cost, and maximizing the satisfaction with the load curve. The research objects of the above literature are all independent thermal power or wind power systems, which are not very applicable to hydropower systems with close upstream and downstream hydropower connections; ② The problem of power decomposition in the power system with multi-energy complementation of water, thermal, wind and light, such as power decomposition with the goal of minimizing the purchase cost of electricity, minimizing the energy consumption, and maximizing the benefit. The above literature has certain practical significance for the participation of multi-energy complementary power systems in power market transactions, but hydropower is only used as an auxiliary energy or only considers the coordination of the same time scale, and rarely involves multi-time scale coordination problems; ③ The problem of power decomposition in the hydropower system mainly includes determining the contract power decomposition plan with the goal of maximizing power generation efficiency, selecting the annual contract proportion to determine the annual contract power and the decomposition plan within the year, and considering the market electricity price fluctuation to fit the market electricity price. It has practical significance for the problem of annual contract power decomposition of hydropower, but is limited by the complexity of the problem.

[0004] The above research has the following deficiencies: first, the model for maximizing power generation benefits does not take into account the loss of abandoned water and the risk of income caused by uneven electricity distribution; second, the annual time scale electricity decomposition model for the annual contract total electricity only considers the decomposition of electricity, and the coordination problem of annual base electricity and monthly contract electricity decomposition needs further study; finally, the problem of large differences in market electricity prices when hydropower stations participate in different electricity markets is not taken into account when fitting the market electricity price, and different market electricity prices need further analysis. Summary of the invention

[0005] With the continuous expansion of hydropower in Southwest my country, hydropower has occupied a dominant position in the medium- and long-term trading market. Researching a medium- and long-term optimal scheduling method for cascade hydropower stations that takes into account the decomposition of contract electricity has guiding significance for solving the problems of uncertainty factors in natural water supply and trading electricity prices that make it difficult to determine different market shares, insufficient utilization of water resources, and uneven distribution of electricity that lead to greater risks in returns. This result provides a solid theoretical basis for the construction of monthly trading electricity decomposition and verification system for large-scale hydropower station groups.

[0006] In order to realize the above-mentioned technical features, the purpose of the present invention is achieved as follows: a medium- and long-term optimization scheduling method for a group of cascade hydropower stations taking into account the decomposition of contract electricity, characterized in that: the goal is to minimize the risk of the entire cascade power generation revenue under a given confidence level; first, the software is used to generate a random monthly scale water inflow process that conforms to the normal distribution; then three electricity price functions are simulated according to the different market proportions of hydropower electricity; different water inflow frequencies and market electricity prices are selected, and a hybrid algorithm of stepwise optimization and successive approximation is used to iteratively solve the power station output decision and the decomposition of contract electricity.

[0007] The specific steps are as follows:

[0008] Step 1: Determine the water flow process at different frequencies:

[0009] In order to solve the problem of asynchronous water inflow frequency in cascade hydropower stations, the Java Random.nextGaussian() method is used to generate random water inflow that conforms to the normal distribution.

[0010] Step 2: Fitting the medium- and long-term contract electricity price function:

[0011] In order to solve the problem that the different types of cascade hydropower stations affect the electricity price, we selected different proportions of hydropower in the market to fit the electricity price curve;

[0012] Step 3: Contract electricity decomposition and verification:

[0013] In order to solve the problem of over-generation and under-generation of hydropower caused by the loose connection between power generation dispatch and market transactions, a method for synchronous solution of hydropower station output and contract power decomposition is used;

[0014] Step 4: Medium- and long-term power generation dispatch process and trading plan formulation:

[0015] With the goal of minimizing the risk of the entire cascade power generation revenue, a hybrid algorithm of stepwise optimization and successive approximation is used to iteratively solve the power station output decision and contract electricity decomposition.

[0016] The detailed steps of determining the water inflow process with different frequencies in step 1 are as follows:

[0017] Assume that the forecast deviation of water inflow in any month during the dispatching period follows the normal distribution N(μ1,σ1 2 ), where μ1 is the mathematical expectation and σ1 is the standard deviation. The multi-year average flow of natural water inflow to each reservoir in the basin during the same period is selected as the expected value. The standard deviation of the natural water inflow flow to each reservoir in the basin during the same period represents the degree of water inflow fluctuation. The Random.nextGaussian() method in Java is used to generate random numbers that conform to the normal distribution to simulate the process of generating water inflows with different frequencies.

[0018] In order to fully reflect the random variation characteristics of the incoming water, multiple groups of frequency incoming water are generated by the above method. The incoming water in each scene is arranged in descending order, numbered in sequence (X = 1, 2, ..., Y), and the empirical frequency is calculated according to formula (1):

[0019]

[0020] Where: P represents the empirical frequency; X represents the different water inflow numbers, X = 1, 2, ..., Y; Y represents the number of generated water inflow scenarios.

[0021] The detailed steps of fitting the medium- and long-term contract electricity price function in step 2 are:

[0022] According to different market situations, there are three scenarios for fitting the medium- and long-term transaction electricity prices of cascade hydropower stations:

[0023] Electricity price simulation method 1: Trading situation where cascade hydropower stations account for a large proportion of the market; When large cascade hydropower stations account for a large proportion of the market transaction share, their transaction electricity price is generally linearly negatively correlated with the transaction electricity volume. At the same time, too high and too low electricity prices are likely to lead to speculative behavior and drastic market fluctuations. Therefore, it is necessary to add upper and lower limits to the electricity price relationship function in accordance with the actual situation for correction. According to the actual production law, the annual contract electricity price is generally between the maximum and minimum values ​​of the monthly contract electricity price and slightly higher than the minimum value of the monthly contract electricity price. In addition, as the annual base electricity volume ratio increases, the annual contract electricity price tends to decrease. Therefore, the annual contract electricity price is fitted with the maximum and minimum values ​​of the monthly contract electricity price and the annual base electricity volume ratio. The corresponding monthly contract electricity price and annual contract electricity price are shown in the following formulas (2) and (3):

[0024]

[0025] In the formula, p y,1 , They represent the annual contract electricity price of electricity price simulation method 1 and the monthly contract electricity price of the tth month, respectively, in units of RMB / kWh; i represents the hydropower station number; M represents the total number of hydropower stations; t represents the month; N i,t , P i,t Respectively represent the output and installed capacity of the i-th hydropower station in the t-th month, in MW; r ave It represents the average value of the upper and lower limits of the annual base electricity ratio, that is, The unit is: %; r i,t They represent the upper and lower limits of the annual base electricity volume of the ith hydropower station in the tth month, respectively, in %; p Indicates the maximum and minimum monthly contract electricity price, in units of RMB / kWh. In order to fit the electricity price function that conforms to market rules, the upper limit of the monthly contract electricity price and the monthly contract price floor p Take 0.25 and 0.1 yuan / kWh respectively; Δp m It represents the difference between the maximum and minimum monthly contract electricity price, that is, β represents an empirical constant. The constant β is increased to fit the actual annual electricity price pattern. When the empirical constant β = 0.05, the annual contract electricity price is more consistent with the historical annual contract electricity price pattern. Therefore, the empirical constant β is taken as 0.05;

[0026] Electricity price simulation method 2: The trading situation where cascade hydropower stations account for a small market share; When the trading share of cascade hydropower stations in the market is relatively small, their trading electricity prices are greatly affected by natural water inflow, and show a "stable-decline-stable-increase" pattern with the water inflow process in different seasons. The electricity prices in the dry season and flood season are relatively stable, and the electricity prices before and after the flood season show a downward and upward trend respectively. The electricity price curve is generated according to the actual electricity price and the electricity price function is fitted. It is assumed that the rising and falling parts of the electricity price curve are linearly positively correlated and linearly negatively correlated. The average of the maximum and minimum electricity prices in the monthly electricity price curve is used as the monthly contract electricity price. The annual contract electricity price is fitted with the maximum and minimum values ​​of the monthly contract electricity price and the annual base electricity ratio. The corresponding monthly contract electricity price and annual contract electricity price are shown in the following formulas (4) and (5):

[0027]

[0028] p y,2 =p1-r ave (p1-p2)+β (5)

[0029] In the formula, p y,2 , They represent the annual contract electricity price of electricity price simulation method 2 and the monthly contract electricity price of the tth month, respectively, in yuan / kWh; p1 and p2 represent the stable electricity prices during the dry season and flood season, respectively, in yuan / kWh. In order to fit the electricity price function that conforms to market rules, p1 and p2 are taken as 0.25 and 0.1 yuan / kWh respectively; t represents the month; r ave It represents the average value of the upper and lower limits of the annual base electricity ratio, that is, The unit is: %; r i,t They represent the upper and lower limits of the annual base electricity ratio of the ith hydropower station in the tth month, respectively, in %; β represents an empirical constant, and the constant β is increased to fit the actual annual electricity price law. When the empirical constant β = 0.05, the annual electricity price is more in line with the historical annual electricity price law, so the empirical constant β is taken as 0.05;

[0030] Electricity price simulation method 3: The transaction situation where cascade hydropower stations account for a moderate proportion of the market; Electricity price simulation method 1 and electricity price simulation method 2 are relatively idealized, while the actual market is more complex. It is also important to study the medium- and long-term market transactions between electricity price simulation method 1 and electricity price simulation method 2; For the convenience of understanding, the average annual contract electricity price and the average monthly contract electricity price of electricity price simulation method 1 and electricity price simulation method 2 are selected as the annual contract electricity price and monthly contract electricity price of electricity price simulation method 3, respectively, to explore the medium- and long-term market transaction results under this electricity price simulation method. The corresponding monthly contract electricity price and annual contract electricity price are shown in the following formulas (6) and (7):

[0031]

[0032] In the formula, p y,3 , They represent the annual contract electricity price of electricity price simulation method 3 and the monthly contract electricity price of the tth month, respectively, in yuan / kWh; Respectively represent the monthly contract electricity price of the tth month of electricity price simulation method 1 and electricity price simulation method 2, the unit is: yuan / kWh; p y,1 、p y,2 They represent the annual contract electricity prices of electricity price simulation method 1 and electricity price simulation method 2 respectively, in units of RMB / kWh.

[0033] The detailed steps of contract quantity decomposition and verification in step 3 are as follows:

[0034] The idea of ​​decomposing the contracted electricity of the i-th hydropower station in the t-th month to solve the maximum power generation benefit is:

[0035] The annual base electricity and monthly contract electricity allocation ratio of each power station in other months remain unchanged. The AΔE electricity of the i-th hydropower station in the t-th month is taken as the annual base electricity, and the power generation after deducting the annual base electricity is taken as the monthly contract electricity. If the AΔE electricity exceeds the maximum annual base electricity The maximum annual base electricity is taken as the annual base electricity, and the power generation after deducting the maximum annual base electricity is taken as the monthly contract electricity. The power generation benefits are calculated for different allocation schemes. The allocation scheme with the maximum power generation benefit is the optimal solution for decomposing the contract electricity of the i-th hydropower station in the t-th month. ΔE represents the step size of the power generation allocated to the annual base electricity, and the unit is: kWh; A represents the allocation degree of the annual base electricity.

[0036] Among them, the upper and lower limits of the annual base electricity ratio will directly affect the final power generation efficiency. The specific ideas for selecting the upper and lower limits of the annual base electricity ratio are:

[0037] The annual base electricity ratio will be capped and lower limit r i,t Discretize with Δr as the step size, take the discretized ratio range as the annual base electricity ratio range, select different ratio ranges to calculate the power generation efficiency, and select the upper and lower limits of the ratio when the power generation efficiency is the largest as the upper and lower limits of the annual base electricity ratio;

[0038] When the contract electricity of the i-th hydropower station in the t-th month is decomposed, the annual base electricity ratio r under different discrete combinations i,t for:

[0039]

[0040] In the formula, r i,t It represents the annual base electricity proportion of the i-th hydropower station in the t-th month, in %; Indicates the lower and upper limits of the annual base electricity, in kWh; ΔE indicates the step size of the power generation allocated to the annual base electricity, in kWh; t indicates the month; i indicates the hydropower station number; N i,t represents the output of the i-th hydropower station in the t-th month, in MW; Δt represents the total number of seconds in the t-th month, that is, Δt = T1 × 24 × 60 × 60, in seconds; T1 represents the number of days in the t-th month; A max , A min Indicates the maximum and minimum values ​​of the distribution degree; A indicates the selected discrete degree value, A min ≤A≤A max ;in:

[0041]

[0042] In the formula, Indicates the lower and upper limits of the annual base electricity quantity, in kWh; r i,t They represent the upper and lower limits of the annual base electricity volume of the i-th hydropower station in the t-th month, respectively, in %; t represents the month; i represents the hydropower station number; N i,t represents the output of the i-th hydropower station in the t-th month, in MW; Δt represents the total number of seconds in the t-th month, that is, Δt = T1 × 24 × 60 × 60, in seconds; T1 represents the number of days in the t-th month; A max , A min Indicates the maximum and minimum allocation levels; ΔE indicates the step size of the allocation of power generation to the annual base electricity, in kWh; Respectively represent the ratio The smallest largest integer.

[0043] The detailed steps of the medium- and long-term power generation dispatching process and trading plan formulation in step 4 are as follows:

[0044] The stepwise optimization-successive approximation algorithm is used to solve the medium- and long-term optimal scheduling and contract power decomposition problems of cascade hydropower stations. The minimum benefit risk model is used with a scheduling period of 1 year and a time interval of 1 month to calculate the minimum benefit R under a given confidence level. VaR The maximum dispatching scheme, the medium- and long-term power generation dispatching process and the trading plan formulation steps are as follows:

[0045] Step a: Determine the water supply conditions at different frequencies; randomly generate multiple groups of frequency water supply according to the historical water supply data, determine the empirical frequency by step 1, and select some frequency water supply with a step size of p=10 groups;

[0046] Step b: Discrete flow; the state variables and decision variables of each power station in other periods remain unchanged, the water levels of the (t-1)th month and the (t-1)th month are fixed, and the outflow flow of each power station in the tth month is discretized, which can generate 3 M discrete schemes, M is the total number of hydropower stations;

[0047] Step c: Select the w=1th discrete scheme in step b, and calculate the power generation and abandoned hydropower of each power station in each period under the discrete scheme;

[0048] Step d: Let i = 1, t = 1, and use the discretized power generation of the i-th hydropower station in the t-th month as the monthly contract power, and the discrete value as the annual base power. Calculate the power generation benefits of each discrete combination, select the maximum power generation benefit as the optimized value, save the optimized value of the hydropower station and assign it to the initial value;

[0049] Step e: Let t = t + 1, and optimize the i-th hydropower station in the t-th month, with the same calculation method as step d, until the condition t = T is met, where T represents the total number of months;

[0050] Step f: Let i = i + 1, select the i-th hydropower station, and perform steps d and e until the condition i = M is met. Use the maximum power generation benefit as the expected value and calculate the minimum benefit R of the discrete scheme under a given confidence level. VaR , select R VaR The maximum value is taken as the optimized value and saved;

[0051] Step g: Let w = w + 1, select the wth discrete solution in step b, and calculate it in the same way as steps c to f until the condition w = 3 is met. M ;

[0052] Step h: Let t = t + 1, and optimize the tth month. The calculation method is the same as steps b to g, and all the optimized decision variable values ​​and state variable values ​​are saved;

[0053] Step i: Minimum benefit R obtained from the previous optimization VaR Compare the two results. If the two results are different, start a new round of optimization calculation until the two results are the same.

[0054] Step k: The optimization is completed and the results are output.

[0055] The present invention has the following beneficial effects:

[0056] 1. The present invention fits a normal distribution curve based on water data over the years and generates random water inflow using Java's Random.nextGaussian() method to handle the problem of asynchronous water inflow frequency between flood and drought periods in cascade hydropower stations, and to a certain extent reflects the impact of natural water inflows of different frequencies on the medium- and long-term optimal scheduling of Yunnan hydropower.

[0057] 2. The present invention selects different market proportions of hydropower stations to fit the electricity price curve, and deals with the problem that the different types of market proportions of cascade hydropower stations affect the electricity price law.

[0058] 3. The present invention establishes a medium- and long-term optimal dispatching model for cascade hydropower stations with the goal of minimizing risk in terms of revenue, and adopts a hybrid algorithm of step-by-step optimization and successive approximation to iteratively solve the power station output decision and contract power decomposition cycle, in order to deal with the problem of over-generation or under-generation of hydropower caused by the lack of close connection between power generation dispatching and market transactions. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0060] Figure 1 This is a relationship diagram between power generation efficiency and annual electricity volume ratio.

[0061] Figure 2 It is an analysis diagram of the amount of abandoned water electricity with and without considering the abandoned water model.

[0062] Figure 3 It is the partial frequency hydropower decomposition and water level process diagram (electricity price simulation method 1).

[0063] Figure 4 This is the analysis chart of power generation efficiency-annual electricity ratio results.

[0064] Figure 5 This is a diagram showing the optimization results of power generation efficiency and annual electricity ratio.

[0065] Table 1 shows the electricity consumption results of each power station with partial frequency water inflow (electricity price simulation method 1). DETAILED DESCRIPTION

[0066] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0067] With the continuous expansion of hydropower in southwest my country and the improvement of relevant policies in the power market, large-scale hydropower participating in market competition is an inevitable trend to achieve optimal resource allocation. Its trading mode is mainly based on medium- and long-term transactions, supplemented by short-term transactions on the day before. Especially in the medium and long term, due to the characteristics of uncertain natural water supply, limited reservoir regulation capacity, close hydropower connection, and large fluctuations in electricity prices, it is difficult for cascade hydropower stations to weigh the market share of different time scales when formulating trading plans. This is particularly evident in the following aspects: too much annual contract electricity may cause revenue loss due to too low electricity prices; too much monthly contract electricity may cause the contract electricity to be unable to be completed due to insufficient power generation capacity of the power station; too much day-ahead or short-term contract electricity may cause transaction failure due to abnormal fluctuations in electricity prices. Therefore, how to formulate reasonable trading electricity for different time scales under the market economy has become a theoretical and practical problem that needs to be solved for cascade hydropower stations to participate in market competition.

[0068] The technical problem to be solved by the present invention is to propose a medium- and long-term optimization scheduling method for a cascade hydropower station group taking into account the decomposition of contract electricity. The goal is to minimize the risk of the entire cascade power generation revenue under a given confidence level. First, the Java Random.nextGaussian() method is used to generate a random monthly water inflow process that conforms to the normal distribution. Then, three electricity price functions are simulated according to the different market proportions of hydropower electricity. Different water inflow frequencies and market electricity prices are selected. A hybrid algorithm of stepwise optimization and successive approximation is used to iteratively solve the power station output decision and the decomposition of contract electricity. The following steps 1-4 are implemented:

[0069] Step 1, determine the water inflow process with different frequencies: To solve the problem of asynchronous water inflow frequency between flood and dry seasons of cascade hydropower stations, it is proposed to use Java's Random.nextGaussian() method to generate random water inflow that conforms to the normal distribution;

[0070] Step 2, fitting the medium- and long-term contract electricity price function: In view of the problem that the different types of cascade hydropower stations in the market affect the electricity price law, the electricity price curve is fitted when the hydropower electricity accounts for different proportions in the market;

[0071] Step 3: Contract power decomposition and verification: To address the problem of over-generation and under-generation of hydropower due to the lack of close connection between power generation dispatch and market transactions, a method for synchronous solution of hydropower station output and contract power decomposition is proposed;

[0072] Step 4, medium- and long-term power generation dispatching process and trading plan formulation: With the goal of minimizing the risk of the entire cascade power generation income, a hybrid algorithm of stepwise optimization and successive approximation is used to iteratively solve the power station output decision and contract power decomposition cycle. The objective function is:

[0073] MaxR VaR (F,1-r*)

[0074]

[0075] In the formula, R VaR It represents the minimum return under a given confidence level, in yuan; 1-r* represents the return volatility level, It represents the proportion of the total annual base electricity, in %; F, F1, F2, F3 represent the maximum power generation benefit, annual contract benefit, monthly contract benefit, and water abandonment loss (water abandonment penalty item), respectively, in RMB; M represents the total number of hydropower stations; T represents the total number of months; t represents the month; i represents the hydropower station number; N i,t , They represent the output and the corresponding output of the abandoned water of the i-th hydropower station in the t-th month, in MW; r i,t 、(1-r i,t ) represent the annual base electricity ratio and monthly contract electricity ratio of the ith hydropower station in the tth month, respectively, in %; represents the monthly contract electricity price of the electricity price simulation method ε in the tth month, ε=1,2,3, and the monthly electricity price of each hydropower station in the tth month is the same, the unit is: yuan / kWh; p y,ε It represents the annual contract electricity price of the electricity price simulation method ε, ε=1,2,3, and the annual contract electricity price of each power station is the same throughout the year, and the unit is: yuan / kWh; Δt represents the total number of seconds in the tth month, that is, Δt=T1×24×60×60, and the unit is: seconds; T1 represents the number of days in the tth month.

[0076] The following constraints need to be met:

[0077] (1) Water balance constraints:

[0078]

[0079] Where V i,i+1 、V i,t Respectively represent the storage capacity of the i-th hydropower station in the t+1th month and the tth month, in m 3 / s; They represent the inflow, power generation and abandoned water flow of the i-th hydropower station in the t-th month, in m 3 / s;

[0080] (2) Hydraulic constraints of upstream and downstream reservoirs:

[0081]

[0082] In the formula, I i+1,t It represents the interval flow of the i+1th hydropower station in the tth month, in m 3 / s;

[0083] (3) Outbound flow constraints:

[0084]

[0085] In the formula, It represents the outflow of the i-th hydropower station in the t-th month, in m 3 / s; They represent the minimum and maximum outflow of the ith hydropower station in the tth month, in m 3 / s;

[0086] (4) Output constraints:

[0087]

[0088] Where N i,t It represents the output of the ith hydropower station in the tth month, in units; N i,t , They represent the minimum and maximum output of the ith hydropower station in the tth month, in MW;

[0089] (5) Reservoir water level constraints:

[0090]

[0091] In the formula, Z i,t represents the upper reservoir water level of the i-th hydropower station in the t-th month, in m; Z i,t , They represent the lowest and highest values ​​of the upper reservoir water level of the ith hydropower station in the tth month, in m;

[0092] (6) End-of-period water level control constraints:

[0093] Z i,t =Z i,end

[0094] In the formula, Z i,end It indicates the controlled water level of the reservoir at the end of the dispatching period, in m;

[0095] (7) Electricity price upper and lower limits:

[0096]

[0097] In order to avoid the negative impact of huge fluctuations in electricity prices on the market, it is necessary to regulate the fluctuation range of electricity prices. This is of great practical significance to ensure the safe and stable operation of the market. represents the monthly contract electricity price of the electricity price simulation method ε in the tth month, ε=1,2,3, unit: yuan / kWh; p y,s represents the annual contract electricity price of the electricity price simulation method ε, ε=1,2,3, in yuan / kWh; p Indicates the maximum and minimum monthly contract electricity price, in yuan / kWh;

[0098] (8) Annual base electricity ratio range constraints:

[0099]

[0100] When hydropower participates in medium- and long-term transactions, signing too many annual contracts may result in loss of revenue due to too low electricity prices, and signing too many monthly contracts may result in a deviation between the actual delivered electricity and the contracted electricity due to insufficient power generation capacity of the power station. Therefore, it is necessary to set upper and lower limits for the annual and monthly contract electricity ratios. i,t It represents the annual base electricity proportion of the i-th hydropower station in the t-th month, in %; r i,t Indicates the upper and lower limits of the annual base electricity ratio, in %.

[0101] Furthermore, the detailed steps of determining the water inflow process with different frequencies in step 1 are as follows:

[0102] Assume that the forecast deviation of water inflow in any month during the dispatching period follows a normal distribution Among them, μ1 is the mathematical expectation, σ1 is the standard deviation, and the multi-year average flow of natural water inflow to each reservoir in the basin during the same period is selected as the expected value. The standard deviation of the natural water inflow flow of each reservoir in the basin during the same period represents the degree of water inflow fluctuation. The Random.nextGaussian() method in Java is used to generate random numbers that conform to the normal distribution to simulate the process of generating water inflows with different frequencies;

[0103] In order to fully reflect the random variation characteristics of the incoming water, multiple groups of frequency incoming water are generated by the above method. The incoming water in each scene is arranged in descending order, numbered in sequence (X = 1, 2, ..., Y), and the empirical frequency is calculated according to formula (1):

[0104]

[0105] Where: P represents the empirical frequency; X represents the different water inflow numbers, X = 1, 2, ..., Y; Y represents the number of generated water inflow scenarios.

[0106] Furthermore, the detailed steps of fitting the medium- and long-term contract electricity price function in step 2 are:

[0107] According to different market situations, there are three scenarios for fitting the medium- and long-term transaction electricity prices of cascade hydropower stations:

[0108] Electricity price simulation method 1: Trading situation where cascade hydropower stations account for a large proportion of the market; When large cascade hydropower stations account for a large proportion of the market transaction share, their transaction electricity price is generally linearly negatively correlated with the transaction electricity volume. At the same time, too high and too low electricity prices are likely to lead to speculative behavior and drastic market fluctuations. Therefore, it is necessary to add upper and lower limits to the electricity price relationship function in accordance with the actual situation for correction. According to the actual production law, the annual contract electricity price is generally between the maximum and minimum values ​​of the monthly contract electricity price and slightly higher than the minimum value of the monthly contract electricity price. In addition, as the annual base electricity volume ratio increases, the annual contract electricity price tends to decrease. Therefore, the annual contract electricity price is fitted with the maximum and minimum values ​​of the monthly contract electricity price and the annual base electricity volume ratio. The corresponding monthly contract electricity price and annual contract electricity price are shown in the following formulas (2) and (3):

[0109]

[0110] In the formula, p y,1 , They represent the annual contract electricity price of electricity price simulation method 1 and the monthly contract electricity price of the tth month, respectively, in units of RMB / kWh; i represents the hydropower station number; M represents the total number of hydropower stations; t represents the month; N i,t , P i,t Respectively represent the output and installed capacity of the i-th hydropower station in the t-th month, in MW; r ave It represents the average value of the upper and lower limits of the annual base electricity ratio, that is, The unit is: %; r i,t The upper and lower limits of the annual base electricity volume of the ith hydropower station in the tth month, respectively, in %; p Indicates the maximum and minimum monthly contract electricity price, in units of RMB / kWh. In order to fit the electricity price function that conforms to market rules, the upper limit of the monthly contract electricity price and the monthly contract price floor p Take 0.25 and 0.1 yuan / kWh respectively; Δp m It represents the difference between the maximum and minimum monthly contract electricity price, that is, β represents an empirical constant. The constant β is increased to fit the actual annual electricity price pattern. When the empirical constant β = 0.05, the annual contract electricity price is more consistent with the historical annual contract electricity price pattern. Therefore, the empirical constant β is taken as 0.05;

[0111] Electricity price simulation method 2: The trading situation where cascade hydropower stations account for a small market share; When the trading share of cascade hydropower stations in the market is relatively small, their trading electricity prices are greatly affected by natural water inflow, and show a "stable-decline-stable-increase" pattern with the water inflow process in different seasons. The electricity prices in the dry season and flood season are relatively stable, and the electricity prices before and after the flood season show a downward and upward trend respectively. The electricity price curve is generated according to the actual electricity price and the electricity price function is fitted. It is assumed that the rising and falling parts of the electricity price curve are linearly positively correlated and linearly negatively correlated. The average of the maximum and minimum electricity prices in the monthly electricity price curve is used as the monthly contract electricity price. The annual contract electricity price is fitted with the maximum and minimum values ​​of the monthly contract electricity price and the annual base electricity ratio. The corresponding monthly contract electricity price and annual contract electricity price are shown in the following formulas (4) and (5):

[0112]

[0113] p y,2 =p1-r ave (p1-p2)+β (5)

[0114] In the formula, p y,2 , They represent the annual contract electricity price of electricity price simulation method 2 and the monthly contract electricity price of the tth month, respectively, in units of RMB / kWh; p1 and p2 represent the stable electricity prices during the dry season and flood season, respectively, in units of RMB / kWh. To fit the electricity price function that conforms to market rules, p1 and p2 are taken as 0.25 and 0.1 RMB / kWh, respectively; t represents the month; r ave It represents the average value of the upper and lower limits of the annual base electricity ratio, that is, The unit is: %; r i,t They represent the upper and lower limits of the annual base electricity ratio of the ith hydropower station in the tth month, respectively, in units of %; β represents an empirical constant, and the constant β is increased to fit the actual annual electricity price law. When the empirical constant β = 0.05, the annual contract electricity price is more in line with the historical annual contract electricity price law, so the empirical constant β is taken as 0.05;

[0115] Electricity price simulation method 3: The transaction situation where cascade hydropower stations account for a moderate proportion of the market; Electricity price simulation method 1 and electricity price simulation method 2 are relatively idealized, while the actual market is more complex. It is also important to study the medium- and long-term market transactions between electricity price simulation method 1 and electricity price simulation method 2; For the convenience of understanding, the average annual contract electricity price and the average monthly contract electricity price of electricity price simulation method 1 and electricity price simulation method 2 are selected as the annual contract electricity price and monthly contract electricity price of electricity price simulation method 3, respectively, to explore the medium- and long-term market transaction results under this electricity price simulation method. The corresponding monthly contract electricity price and annual contract electricity price are shown in the following formulas (6) and (7):

[0116]

[0117] In the formula, p y,3 , They represent the annual contract electricity price of electricity price simulation method 3 and the monthly contract electricity price of the tth month, respectively, in yuan / kWh; Respectively represent the monthly contract electricity price of the tth month of electricity price simulation method 1 and electricity price simulation method 2, the unit is: yuan / kWh; p y,1 、p y,2 They represent the annual contract electricity prices of electricity price simulation method 1 and electricity price simulation method 2 respectively, in units of RMB / kWh.

[0118] Furthermore, the detailed steps of decomposing and verifying the contract electricity in step 3 are as follows:

[0119] The idea of ​​decomposing the contracted electricity of the i-th hydropower station in the t-th month to solve the maximum power generation benefit is:

[0120] The annual base electricity and monthly contract electricity allocation ratio of each power station in other months remain unchanged. The AΔE electricity of the i-th hydropower station in the t-th month is taken as the annual base electricity, and the power generation after deducting the annual base electricity is taken as the monthly contract electricity. If the AΔE electricity exceeds the maximum annual base electricity The maximum annual base electricity is taken as the annual base electricity, and the power generation after deducting the maximum annual base electricity is taken as the monthly contract electricity. The power generation benefits are calculated for different allocation schemes. The allocation scheme with the maximum power generation benefit is the optimal solution for decomposing the contract electricity of the i-th hydropower station in the t-th month. ΔE represents the step size of the power generation allocated to the annual base electricity, and the unit is: kWh; A represents the allocation degree of the annual base electricity.

[0121] Among them, the upper and lower limits of the annual base electricity ratio will directly affect the final power generation efficiency. The specific ideas for selecting the upper and lower limits of the annual base electricity ratio are:

[0122] The annual base electricity ratio will be capped and lower limit r i,t Discretize with Δr as the step size, take the discretized ratio range as the annual base electricity ratio range, select different ratio ranges to calculate the power generation efficiency, and select the upper and lower limits of the ratio when the power generation efficiency is the largest as the upper and lower limits of the annual base electricity ratio;

[0123] When the contract electricity of the i-th hydropower station in the t-th month is decomposed, the annual base electricity ratio r under different discrete combinations i,t for:

[0124]

[0125] In the formula, r i,t It represents the annual base electricity proportion of the i-th hydropower station in the t-th month, in %; Indicates the lower and upper limits of the annual base electricity, in kWh; ΔE indicates the step size of the power generation allocated to the annual base electricity, in kWh; t indicates the month; i indicates the hydropower station number; N i,t represents the output of the i-th hydropower station in the t-th month, in MW; Δt represents the total number of seconds in the t-th month, that is, Δt = T1 × 24 × 60 × 60, in seconds; T1 represents the number of days in the t-th month; A max , A min Indicates the maximum and minimum values ​​of the distribution degree; A indicates the selected discrete degree value, A min ≤A≤A max ;in:

[0126]

[0127] In the formula, Indicates the lower and upper limits of the annual base electricity quantity, in kWh; r i,t They represent the upper and lower limits of the annual base electricity volume of the i-th hydropower station in the t-th month, respectively, in %; t represents the month; i represents the hydropower station number; N i,t represents the output of the i-th hydropower station in the t-th month, in MW; Δt represents the total number of seconds in the t-th month, that is, Δt = T1 × 24 × 60 × 60, in seconds; T1 represents the number of days in the t-th month; A max , A min Indicates the maximum and minimum allocation levels; ΔE indicates the step size of the allocation of power generation to the annual base electricity, in kWh; Respectively represent the ratio The smallest largest integer.

[0128] Furthermore, the detailed steps of the medium- and long-term power generation dispatching process and trading plan formulation in step 4 are as follows:

[0129] The stepwise optimization-successive approximation algorithm is used to solve the medium- and long-term optimal scheduling and contract power decomposition problems of cascade hydropower stations. The minimum benefit risk model is used with a scheduling period of 1 year and a time interval of 1 month to calculate the minimum benefit R under a given confidence level. VaR The maximum dispatching scheme, the medium- and long-term power generation dispatching process and the trading plan formulation steps are as follows:

[0130] Step a: Determine the water supply conditions at different frequencies; randomly generate multiple groups of frequency water supply according to the historical water supply data, determine the empirical frequency by step 1, and select some frequency water supply with a step size of p=10 groups;

[0131] Step b: Discrete flow; the state variables and decision variables of each power station in other periods remain unchanged, the water levels of the (t-1)th month and the (t-1)th month are fixed, and the outflow flow of each power station in the tth month is discretized, which can generate 3 M discrete schemes, M is the total number of hydropower stations;

[0132] Step c: Select the w=1th discrete scheme in step b, and calculate the power generation and abandoned hydropower of each power station in each period under the discrete scheme;

[0133] Step d: Let i = 1, t = 1, and use the discretized power generation of the i-th hydropower station in the t-th month as the monthly contract power, and the discrete value as the annual base power. Calculate the power generation benefits of each discrete combination, select the maximum power generation benefit as the optimized value, save the optimized value of the hydropower station and assign it to the initial value;

[0134] Step e: Let t = t + 1, and optimize the i-th hydropower station in the t-th month, with the same calculation method as step d, until the condition t = T is met, where T represents the total number of months;

[0135] Step f: Let i = i + 1, select the i-th hydropower station, and perform steps d and e until the condition i = M is met. Use the maximum power generation benefit as the expected value and calculate the minimum benefit R of the discrete scheme under a given confidence level. VaR , select R VaR The maximum value is taken as the optimized value and saved;

[0136] Step g: Let w = w + 1, select the wth discrete solution in step b, and calculate it in the same way as steps c to f until the condition w = 3 is met. M ;

[0137] Step h: Let t = t + 1, and optimize the tth month. The calculation method is the same as steps b to g, and all the optimized decision variable values ​​and state variable values ​​are saved;

[0138] Step i: Minimum benefit R obtained from the previous optimization VaRCompare the two results. If the two results are different, start a new round of optimization calculation until the two results are the same.

[0139] Step k: The optimization is completed and the results are output.

[0140] Embodiment 2:

[0141] Taking the annual and monthly power generation plan preparation of the Lancang River Basin in Yunnan as an example, a group of six hydropower stations in the middle and lower reaches of the Lancang River are selected to form a cascade hydropower station. The scheduling period is 1 year and the time interval is 1 month. According to the historical water inflow data, 100 groups of frequency water inflows are randomly generated and arranged in descending order according to the water inflow frequency. 10 groups of frequency water inflows are selected with a step size of p = 10. Given a confidence level of α = 95%, the effectiveness and practicality of the proposed method are verified and analyzed under three electricity price simulation methods. From 2017 to 2020, the annual and monthly market transaction electricity prices in Yunnan Province fluctuated in the range of 0.1 to 0.25 yuan / kWh. In order to fit the electricity price function that conforms to market laws, the monthly contract electricity price cap of step (2) electricity price simulation method 1 is and electricity price floor p The dry season stable electricity price p1 and flood season stable electricity price p2 of electricity price simulation method 2 in step 2 are respectively 0.25 yuan / kWh and 0.1 yuan / kWh. After a large number of tests, the annual electricity price is more in line with the historical annual electricity price law when the empirical constant β=0.05, so the empirical constant β is taken as 0.05.

[0142] The following is a detailed analysis of the upper and lower limits of the annual base electricity ratio, the water abandonment situation under different water inflow frequencies and market electricity prices, the dispatch process under different water inflow frequencies and market electricity prices, and the impact of the annual base electricity ratio on benefits. The analysis results are as follows:

[0143] First, the upper and lower limits of the annual base electricity ratio are analyzed:

[0144] Figure 1 Taking one group of water inflow frequencies as an example, the confidence level α=95% is selected to calculate the relationship between power generation efficiency and annual electricity volume ratio under three electricity price simulation methods.

[0145] In the process of formulating the medium- and long-term trading plan for hydropower, it is necessary to set upper and lower limits for the annual base electricity ratio and the monthly contract electricity ratio to avoid the loss of revenue due to excessive annual base electricity contracts and the deviation between the actual delivered electricity and the contracted electricity due to excessive monthly contract electricity contracts. In order to explore the best upper and lower limit ratio combination, the upper limit of the annual base electricity ratio of the i-th hydropower station in the t-th month is and lower limit r i,t The results show that: ① With the increase of the upper and lower limits of the annual base electricity ratio, the power generation efficiency shows a parabolic trend of first increasing and then decreasing; ② For the same annual base electricity ratio range, the minimum benefit R under 95% confidence level VaR The trend is that electricity price simulation method 1> electricity price simulation method 3> electricity price simulation method 2; ③ Under the same water inflow frequency, the trend of the best range of annual base electricity ratio is that electricity price simulation method 1< electricity price simulation method 3< electricity price simulation method 2. Figure 1 The optimal ranges of annual electricity volume ratio for electricity price simulation method 1, electricity price simulation method 2, and electricity price simulation method 3 are 20% to 40%, 35% to 55%, and 25% to 45%, respectively. Due to the large number of combinations of different electricity prices, only the optimal range of annual base electricity volume ratio is selected for verification.

[0146] Secondly, the water abandonment situation is analyzed under different water inflow frequency and electricity price simulation methods:

[0147] In order to further compare and analyze the water abandonment situation, this section adds a comparison scheme for the simulation methods of water supply and electricity prices with different frequencies. The comparison model only takes the minimum benefit risk as the goal and solves the minimum benefit value R under 95% confidence level. VaR When the water abandonment loss penalty F3 is not considered, determine R VaR After the value is reached, the water abandonment penalty will be imposed. Figure 2 The comparison results of water abandonment under three electricity price simulation methods for 10 groups of water inflow frequencies are shown. The results show that: ① Under the same electricity price simulation method, water abandonment decreases with the increase of water inflow frequency. As shown in electricity price simulation method 1 in the water abandonment model, the amount of water abandonment decreases as the water inflow frequency increases from small to large; ② Under each electricity price simulation method in the same model, there is little difference in water abandonment caused by water inflow in normal water years and dry water years. For example, in the water abandonment model, electricity price simulation method 1 has 0 water abandonment when the water inflow frequency is 89.11% and 99.01%, and the amount of water abandonment by electricity price simulation method 2 and electricity price simulation method 3 is 0.001 and 0 billion kWh and 0.005 and 0 billion kWh respectively. There is little difference between the three. The main reason is that there is less water in dry and normal water years, which makes the regulating space of each power station reservoir large, resulting in less water abandonment; ③ Under the same electricity price simulation method, the water abandonment generated by the model considering water abandonment is less than the water abandonment generated by the model not considering water abandonment. Figure 2 As shown in , the differences are all greater than zero, indicating that the water abandonment model produces less water abandonment than the model without considering water abandonment. Therefore, considering the water abandonment model can alleviate the pressure of water abandonment in the optimal scheduling process of cascade hydropower stations and reduce resource waste.

[0148] Then, the dispatching process is analyzed under different frequency water and electricity price simulation methods:

[0149] Taking electricity price simulation method 1 as an example, we selected a group of water in the years of high water, low water and normal water for analysis. Figure 3The power decomposition and water level changes of the three groups of water in the electricity price simulation method 1 are shown in Table 1. The power results of the three groups of water in the three electricity price simulation methods are shown in Table 1. The specific analysis is as follows.

[0150] From the analysis of total power generation, the more water comes in at the same market electricity price, the more total power generation; under the same water inflow frequency, the annual base electricity ratio of electricity price simulation method 2 is higher than that of electricity price simulation method 3; under the same electricity price simulation method, the annual base electricity ratio increases with the increase of water inflow. As shown in Table 1, the total underwater power generation of electricity price simulation method 1 in dry, normal and wet years is 65.481 billion kWh, 76.853 billion kWh and 87.471 billion kWh respectively, and the total underwater power generation of electricity price simulation method 2 and electricity price simulation method 3 in dry, normal and wet years is 66.381 billion kWh, 76.852 billion kWh and 87.531 billion kWh respectively; from the perspective of annual base electricity ratio, the underwater power generation of electricity price simulation method 1, electricity price simulation method 2 and electricity price simulation method 3 in dry, normal and wet years is 24.00%, 42.22% and 34.00% respectively. 0%, the annual base electricity proportion of electricity price simulation method 1 is the smallest, followed by electricity price simulation method 3, and electricity price simulation method 2 is the largest. The main reason is that electricity price 2 is a medium- and long-term transaction in which cascade hydropower stations account for a small market proportion. Compared with electricity price simulation methods 1 and 3, the market transaction risk is relatively large. In order to avoid the risk of income as much as possible, more electricity is allocated to more stable annual contracts. In normal water years and flood years, the annual base electricity proportion of electricity price simulation method 1 is relatively large. The main reason is that after the minimum risk model considers the annual base electricity proportion risk, it reduces the risk by increasing the annual base electricity proportion and achieves the minimum benefit R VaR Maximum; Electricity price simulation method 1 In dry, normal and abundant water years, the proportion of annual base electricity under water is 24.00%, 48.14% and 48.34% respectively. The water inflow degree is positively proportional to the proportion of annual base electricity. The main reason is that when the water inflow is abundant, the water abandonment is large, and the penalty for abandonment is large. Compared with dry and normal water years, the water revenue fluctuates greatly. In order to achieve transaction stability as much as possible, more electricity is allocated to more stable annual contracts. When formulating the medium- and long-term trading plan of cascade hydropower stations, the corresponding annual base electricity proportion can be referred to according to the market proportion of hydropower and the water inflow frequency of cascade hydropower stations, which can be used as the basis for the allocation of annual and monthly contract electricity.

[0151] From the monthly power generation and water level analysis of each power station, such as Figure 3As shown, taking Longtou Power Station in a flood year as an example, the reservoir water level of each power station shows a downward trend during the dry season, the power generation remains at a low level and the proportion of annual base electricity is small; during the normal water season, each power station empties the reservoir to ensure sufficient reservoir capacity for flood storage before the flood season; with the arrival of the flood season, the water level begins to rise due to the increase in water inflow, and the reservoir water storage leads to less power generation; when the water inflow increases to a certain value during the flood season, each hydropower station is close to full load power generation, the proportion of annual base electricity increases, and the water level of each power station rises rapidly to the flood season water level; then the water inflow decreases during the normal water season, the water level of each power station drops to the normal water storage level and the power generation decreases. In the process of decomposing contract electricity, the proportion of annual base electricity in the flood season is large, while the proportion of monthly contract electricity in the dry season is small. The main reason is that the large power generation in the flood season leads to a low monthly electricity price. In order to achieve the minimum benefit R under 95% confidence level VaR The maximum goal is to allocate most of the power generation to the annual contract with higher electricity prices. The results can determine the annual and monthly power decomposition curves under different electricity price simulation methods and water inflow frequencies. When formulating medium- and long-term trading plans for cascade hydropower stations, the corresponding decomposition curves can be referred to according to the proportion of cascade hydropower stations in the market and the water inflow frequency, and a more reasonable medium- and long-term trading plan can be formulated.

[0152] Finally, the impact of annual base electricity ratio on benefits is analyzed:

[0153] This section adds a control group to only take the maximum power generation benefit F as the target for comparative analysis with the present invention, and the control group is called the maximum power generation benefit model. Figure 4 This is a comparison of the benefit curves of the maximum power generation benefit model and the minimum profit risk model with the annual base electricity ratio under the three electricity price simulation methods for partial frequency water inflow. The annual base electricity ratio and benefit after optimization under each frequency water inflow are shown in Figure 2. Figure 5 , the specific analysis is as follows.

[0154] like Figure 4 As shown in the figure, there are certain differences in the power generation benefits under different electricity price simulation methods with the same frequency of water inflow, mainly manifested in that electricity price simulation method 1 has the largest power generation benefit, followed by electricity price simulation method 3, and electricity price simulation method 2 has the smallest power generation benefit. The main reason is that electricity price simulation method 1 is a situation where cascade hydropower stations account for a large proportion of the market and dominate the market, so the risk in the transaction process is relatively small. Therefore, most of the electricity is allocated to monthly contracts with higher electricity prices. Electricity price simulation method 2 is a medium- and long-term transaction in which cascade hydropower stations account for a small proportion of the market. The risk in the transaction process is relatively large. In order to avoid the risk of income and obtain more stable power generation benefits, most of the electricity is allocated to annual contracts with lower electricity prices but more stable. Electricity price simulation method 3 is a situation in which cascade hydropower stations account for a moderate proportion of the market. Compared with electricity price simulation method 1, it has a greater risk, but is more proactive than electricity price simulation method 2. Therefore, the power generation benefits of electricity price simulation method 1, electricity price simulation method 3, and electricity price simulation method 2 at the same water inflow frequency decrease.

[0155] Comparative analysis of the maximum power generation benefit model and the minimum profit risk model. Figure 4 It can be seen that the maximum power generation benefit model can achieve the maximum benefit when the benefit risk caused by the decomposition of the annual base electricity is not considered. When the benefit risk is considered, the minimum benefit risk model has the minimum benefit R at a 95% confidence level. VaR It is larger than the maximum power generation benefit model and can better avoid risks. For scenario 1 of Rulai Water, in the electricity price simulation method 1, the maximum benefits calculated by the maximum power generation benefit model and the minimum benefit risk model are 11.821 billion yuan and 11.781 billion yuan respectively. The former has higher benefits, but at a 95% confidence level, the minimum benefit value R of the maximum power generation benefit model and the minimum benefit risk model after considering risks is VaR 11.092 billion and 11.143 billion respectively, and the minimum benefit value R of the model with the minimum return risk is obvious. VaR The larger the annual base electricity ratio, the smaller the risk; from the perspective of the annual base electricity ratio, the optimal range of the annual base electricity ratio of the model with the minimum benefit risk is not lower than that of the model with the maximum power generation benefit. For example, in the water scenario 1, when using the electricity price simulation method 1, the optimal ranges of the annual base electricity ratio of the model with the maximum power generation benefit and the model with the minimum benefit risk are 0-20% and 10%-30% respectively. The main reason is that the annual base electricity share directly affects the power generation benefit. Considering that the greater the power generation benefit, the greater the risk, the annual electricity market share is increased to avoid risks. Figure 5 It can be seen that the minimum benefit risk model increases the minimum benefit R by increasing the annual base electricity proportion VaR In order to reduce the risk, compared with the maximum power generation benefit model, the minimum benefit R of the model with the minimum risk of profit under the three electricity price simulation methods is VaR The average growth values ​​were 294 million, 241 million and 176 million yuan respectively, which reduced the return risk to a certain extent.

[0156] Table 1: Electricity generation results of each power station with partial frequency water inflow (electricity price simulation method 1)

[0157]

Claims

1. A mid- to long-term optimal dispatching method for a cascade hydropower station group taking into account the decomposition of contracted electricity, characterized by: The goal is to minimize the risk of the entire cascade power generation revenue under a given confidence level. First, the software is used to generate a random monthly water inflow process that conforms to the normal distribution. Then, three electricity price functions are simulated according to the different proportions of hydropower in the market. Different water inflow frequencies and market electricity prices are selected, and a hybrid algorithm of stepwise optimization and successive approximation is used to iteratively solve the power station output decision and contract power decomposition. The specific steps are as follows: Step 1: Determine the water flow process at different frequencies: In order to solve the problem of asynchronous water inflow frequency in cascade hydropower stations, the Java Random.nextGaussian() method is used to generate random water inflow that conforms to the normal distribution. Step 2: Fitting the medium- and long-term contract electricity price function: In order to solve the problem that the different types of cascade hydropower stations affect the electricity price, we selected different proportions of hydropower in the market to fit the electricity price curve; Step 3: Contract electricity decomposition and verification: In order to solve the problem of over-generation and under-generation of hydropower caused by the loose connection between power generation dispatch and market transactions, a method for synchronous solution of hydropower station output and contract power decomposition is used; Step 4: Medium- and long-term power generation dispatch process and trading plan formulation: With the goal of minimizing the risk of the entire cascade power generation revenue, a hybrid algorithm of stepwise optimization and successive approximation is used to iteratively solve the power station output decision and contract power decomposition cycle; The detailed steps of fitting the medium- and long-term contract electricity price function in step 2 are: According to different market situations, there are three scenarios for fitting the medium- and long-term transaction electricity prices of cascade hydropower stations: Electricity price simulation method 1: Trading situation where cascade hydropower stations account for a large proportion of the market; When large cascade hydropower stations account for a large proportion of the market transaction share, their transaction electricity price is generally linearly negatively correlated with the transaction electricity volume. At the same time, too high and too low electricity prices are likely to lead to speculative behavior and drastic market fluctuations. Therefore, it is necessary to add upper and lower limits to the electricity price relationship function in accordance with the actual situation for correction. According to the actual production law, the annual contract electricity price is generally between the maximum and minimum values ​​of the monthly contract electricity price and slightly higher than the minimum value of the monthly contract electricity price. In addition, as the annual base electricity volume ratio increases, the annual contract electricity price tends to decrease. Therefore, the annual contract electricity price is fitted with the maximum and minimum values ​​of the monthly contract electricity price and the annual base electricity volume ratio. The corresponding monthly contract electricity price and annual contract electricity price are shown in the following formulas (2) and (3): In the formula, p y,1 , They represent the annual contract electricity price of electricity price simulation method 1 and the monthly contract electricity price of the tth month, respectively, in units of RMB / kWh; i represents the hydropower station number; M represents the total number of hydropower stations; t represents the month; N i,t , P i,t Respectively represent the output and installed capacity of the i-th hydropower station in the t-th month, in MW; r ave It represents the average value of the upper and lower limits of the annual base electricity ratio, that is, The unit is: %; r i,t They represent the upper and lower limits of the annual base electricity volume of the ith hydropower station in the tth month, respectively, in %; p Indicates the maximum and minimum monthly contract electricity price, in units of RMB / kWh. In order to fit the electricity price function that conforms to market rules, the upper limit of the monthly contract electricity price and the monthly contract price floor p Take 0.25 and 0.1 yuan / kWh respectively; Δp m It represents the difference between the maximum and minimum monthly contract electricity price, that is, β represents an empirical constant. The constant β is increased to fit the actual annual electricity price pattern. When the empirical constant β = 0.05, the annual contract electricity price is more consistent with the historical annual contract electricity price pattern. Therefore, the empirical constant β is taken as 0.05; Electricity price simulation method 2: The trading situation where cascade hydropower stations account for a small proportion of the market; when the trading share of cascade hydropower stations in the market is relatively small, their trading electricity prices are greatly affected by natural water inflow, and show a "stable-decline-stable-increase" pattern with the water inflow process in different seasons. The electricity prices in the dry season and flood season are relatively stable, and the electricity prices before and after the flood season show a downward and upward trend respectively. The electricity price curve is generated according to the actual electricity price and the electricity price function is fitted. It is assumed that the rising and falling parts of the electricity price curve are linearly positively correlated and linearly negatively correlated. The average of the maximum and minimum electricity prices in the monthly electricity price curve is used as the monthly contract electricity price. The annual contract electricity price is fitted with the maximum and minimum values ​​of the monthly contract electricity price and the annual base electricity ratio. The corresponding monthly contract electricity price and annual contract electricity price are shown in the following formulas (4) and (5): p y,2 =p1-r ave (p1-p2)+β (5) In the formula, p y,2 , They represent the annual contract electricity price of electricity price simulation method 2 and the monthly contract electricity price of the tth month, respectively, in yuan / kWh; p1 and p2 represent the stable electricity prices during the dry season and flood season, respectively, in yuan / kWh. In order to fit the electricity price function that conforms to market rules, p1 and p2 are taken as 0.25 and 0.1 yuan / kWh respectively; t represents the month; r ave It represents the average value of the upper and lower limits of the annual base electricity ratio, that is, The unit is: %; r i,t They represent the upper and lower limits of the annual base electricity ratio of the ith hydropower station in the tth month, respectively, in %; β represents an empirical constant, and the constant β is increased to fit the actual annual electricity price law. When the empirical constant β = 0.05, the annual electricity price is more in line with the historical annual electricity price law, so the empirical constant β is taken as 0.05; Electricity price simulation method 3: The transaction situation where cascade hydropower stations account for a moderate proportion of the market; Electricity price simulation method 1 and electricity price simulation method 2 are relatively idealized, while the actual market is more complex. It is also important to study the medium- and long-term market transactions between electricity price simulation method 1 and electricity price simulation method 2. For ease of understanding, the average annual contract electricity price and the average monthly contract electricity price of electricity price simulation method 1 and electricity price simulation method 2 are selected as the annual contract electricity price and monthly contract electricity price of electricity price simulation method 3, respectively, to explore the medium- and long-term market transaction results under this electricity price simulation method. The corresponding monthly contract electricity price and annual contract electricity price are shown in the following formulas (6) and (7): In the formula, p y,3 , They represent the annual contract electricity price of electricity price simulation method 3 and the monthly contract electricity price of the tth month, respectively, in yuan / kWh; Respectively represent the monthly contract electricity price of the tth month of electricity price simulation method 1 and electricity price simulation method 2, the unit is: yuan / kWh; p y,1 、p y,2 They represent the annual contract electricity prices of electricity price simulation method 1 and electricity price simulation method 2 respectively, in units of RMB / kWh.

2. According to claim 1, a method for mid- and long-term optimal dispatching of a cascade hydropower station group taking into account the decomposition of contracted electricity, characterized in that: The detailed steps of determining the water inflow process with different frequencies in step 1 are as follows: Assume that the forecast deviation of water inflow in any month during the dispatching period follows a normal distribution Among them, μ1 is the mathematical expectation, σ1 is the standard deviation, and the multi-year average flow of natural water inflow to each reservoir in the basin during the same period is selected as the expected value. The standard deviation of the natural water inflow flow of each reservoir in the basin during the same period represents the degree of water inflow fluctuation. The Random.nextGaussian() method in Java is used to generate random numbers that conform to the normal distribution to simulate the process of generating water inflows with different frequencies; In order to fully reflect the random variation characteristics of the incoming water, multiple groups of frequency incoming water are generated by the above method. The incoming water in each scene is arranged in descending order, numbered in sequence (X = 1, 2, ..., Y), and the empirical frequency is calculated according to formula (1): Where: P represents the empirical frequency; X represents the different water inflow numbers, X = 1, 2, ..., Y; Y represents the number of generated water inflow scenarios.

3. According to claim 1, a method for mid- and long-term optimal dispatching of a cascade hydropower station group taking into account the decomposition of contracted electricity, characterized in that: The detailed steps of contract quantity decomposition and verification in step 3 are as follows: The idea of ​​decomposing the contracted electricity of the i-th hydropower station in the t-th month to solve the maximum power generation benefit is: The annual base electricity and monthly contract electricity allocation ratio of each power station in other months remain unchanged. The AΔE electricity of the i-th hydropower station in the t-th month is taken as the annual base electricity, and the power generation after deducting the annual base electricity is taken as the monthly contract electricity. If the AΔE electricity exceeds the maximum annual base electricity The maximum annual base electricity is taken as the annual base electricity, and the power generation after deducting the maximum annual base electricity is taken as the monthly contract electricity. The power generation benefits are calculated for different allocation schemes. The allocation scheme with the maximum power generation benefit is the optimal solution for decomposing the contract electricity of the i-th hydropower station in the t-th month. ΔE represents the step size of the power generation allocated to the annual base electricity, and the unit is: kWh; A represents the allocation degree of the annual base electricity. Among them, the upper and lower limits of the annual base electricity ratio will directly affect the final power generation efficiency. The specific ideas for selecting the upper and lower limits of the annual base electricity ratio are: The annual base electricity ratio will be capped and lower limit r i,t Discretize with Δr as the step size, take the discretized ratio range as the annual base electricity ratio range, select different ratio ranges to calculate the power generation efficiency, and select the upper and lower limits of the ratio when the power generation efficiency is the largest as the upper and lower limits of the annual base electricity ratio; When the contract electricity of the i-th hydropower station in the t-th month is decomposed, the annual base electricity ratio r under different discrete combinations i,t for: In the formula, r i,t It represents the annual base electricity proportion of the i-th hydropower station in the t-th month, in %; Indicates the lower and upper limits of the annual base electricity, in kWh; ΔE indicates the step size of the distribution of power generation to the annual base electricity, in kWh; t indicates the month; i indicates the hydropower station number; N i,t represents the output of the i-th hydropower station in the t-th month, in MW; Δt represents the total number of seconds in the t-th month, that is, Δt = T1 × 24 × 60 × 60, in seconds; T1 represents the number of days in the t-th month; A max , A min Indicates the maximum and minimum values ​​of the distribution degree; A indicates the selected discrete degree value, A min ≤A≤A max ;in: In the formula, Indicates the lower and upper limits of the annual base electricity quantity, in kWh; r i,t They represent the upper and lower limits of the annual base electricity volume of the i-th hydropower station in the t-th month, respectively, in %; t represents the month; i represents the hydropower station number; N i,t represents the output of the i-th hydropower station in the t-th month, in MW; Δt represents the total number of seconds in the t-th month, that is, Δt = T1 × 24 × 60 × 60, in seconds; T1 represents the number of days in the t-th month; A max , A min Indicates the maximum and minimum allocation levels; ΔE indicates the step size of the allocation of power generation to the annual base electricity, in kWh; Respectively represent the ratio The smallest largest integer.

4. According to claim 1, a method for mid- and long-term optimal dispatching of a cascade hydropower station group taking into account the decomposition of contracted electricity, characterized in that: The detailed steps of the medium- and long-term power generation dispatching process and trading plan formulation in step 4 are as follows: The stepwise optimization-successive approximation algorithm is used to solve the medium- and long-term optimal scheduling and contract power decomposition problems of cascade hydropower stations. The minimum benefit risk model is used with a scheduling period of 1 year and a time interval of 1 month to calculate the minimum benefit R under a given confidence level. VaR The maximum dispatching scheme, the medium- and long-term power generation dispatching process and the trading plan formulation steps are as follows: Step a: Determine the water supply conditions at different frequencies; randomly generate multiple groups of frequency water supply according to the historical water supply data, determine the empirical frequency by step 1, and select some frequency water supply with a step size of p=10 groups; Step b: Discrete flow; the state variables and decision variables of each power station in other periods remain unchanged, the water levels of the (t-1)th month and the (t-1)th month are fixed, and the outflow flow of each power station in the tth month is discretized, which can generate 3 M discrete schemes, M is the total number of hydropower stations; Step c: Select the w=1th discrete scheme in step b, and calculate the power generation and abandoned hydropower of each power station in each period under the discrete scheme; Step d: Let i = 1, t = 1, and use the discretized power generation of the i-th hydropower station in the t-th month as the monthly contract power, and the discrete value as the annual base power. Calculate the power generation benefits of each discrete combination, select the maximum power generation benefit as the optimized value, save the optimized value of the hydropower station and assign it to the initial value; Step e: Let t = t + 1, and optimize the i-th hydropower station in the t-th month, with the same calculation method as step d, until the condition t = T is met, where T represents the total number of months; Step f: Let i = i + 1, select the i-th hydropower station, and perform steps d and e until the condition i = M is met. Use the maximum power generation benefit as the expected value and calculate the minimum benefit R of the discrete scheme under a given confidence level. VaR , select R VaR The maximum value is taken as the optimized value and saved; Step g: Let w = w + 1, select the wth discrete solution in step b, and calculate it in the same way as steps c to f until the condition w = 3 is met. M ; Step h: Let t = t + 1, and optimize the tth month. The calculation method is the same as steps b to g, and all the optimized decision variable values ​​and state variable values ​​are saved; Step i: Minimum benefit R obtained from the previous optimization VaR Compare the two results. If the two results are different, start a new round of optimization calculation until the two results are the same. Step k: The optimization is completed and the results are output.