A method for making short-term power generation plan of cascade hydropower stations considering peak load depth

By using different incoming period models and heuristic algorithms in cascade hydropower stations to optimize peak regulating depth, the problem of insufficient flexibility in production of short-term power generation plans in the existing technology is solved, and the matching of power generation plans and power consumption needs of the power grid is achieved, reducing the The amount of water abandoned improves the efficiency of power generation.

CN114548666BActive Publication Date: 2024-11-05FUJIAN HUADIAN FURUI ENERGY DEV CO LTD CHITAN HYDROPOWER PLANT +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210026188.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-01-11
Publication Date
2024-11-05
Estimated Expiration
2042-01-11

AI Technical Summary

Technical Problem

The existing short-term power generation plan production methods for cascade hydropower stations are not flexible enough during the scheduling process, making it difficult to effectively track the power grid electricity demand, resulting in the power generation plan not having the characteristics of frequent electricity generation during peak electricity use and fewer electricity generation during low electricity use, and requires manual intervention and adjustment.

Method used

Different incoming water period models (flood, dry and flat water period models) are adopted as the goal, and the peak regulating depth and water discarded amount are optimized through heuristic algorithms, combined with water balance and reservoir water level constraints, short-term power generation plans are formulated to limit peak regulating of leading power stations. The amplitude is used to control the rungage output to ensure that the power generation plan matches the power consumption requirements of the power grid.

Benefits of technology

It improves the flexibility and practicality of the short-term power generation plan of cascade hydropower stations, can adapt to different incoming water conditions, reduce the amount of water abandoned in downstream power stations, effectively match the power consumption needs of power grids, and improves the power generation efficiency of power stations with limited regulation capabilities.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure BDA0003464757470000021
    Figure BDA0003464757470000021
  • Figure BDA0003464757470000031
    Figure BDA0003464757470000031
  • Figure BDA0003464757470000033
    Figure BDA0003464757470000033
Patent Text Reader

Abstract

The invention discloses a method for making a short-term power generation plan of a cascade hydropower station considering the peak-shaving depth, and relates to the field of short-term optimization scheduling of cascade hydropower stations in the power industry. The steps are as follows: Step 1: Determine a short-term power generation plan making idea considering the peak-shaving depth; Step 2: According to the characteristics of different peak-shaving amplitudes and abandoned water volumes in different periods, respectively establish flood season, dry season and normal water season models, take peak-shaving of different depths and reduction of abandoned water volume as the goals, combine the cascade energy storage control strategy at key time nodes in the actual scheduling process and the peak capacity of the cascade power station that meets the power demand of the power grid to determine the model constraints, the constraints include water balance, reservoir water level constraints, end water level control, peak-shaving depth constraints, power generation flow constraints, outflow flow constraints, power station output constraints and other constraints, and determine the model solving algorithm; Step 3: Use a heuristic algorithm to solve the short-term optimization scheduling.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of short-term optimization scheduling of cascade hydropower stations in the electric power industry, and in particular to a method for making a short-term power generation plan for a cascade hydropower station. Background Art

[0002] The short-term power generation plan of cascade hydropower stations is based on the scheduling cycle of the day and the scheduling period of the hour or 15 minutes. The power generation plan of each cascade power station is arranged on the subsequent date according to the actual situation of water inflow forecast, unit maintenance arrangement, power grid peak regulation, etc., and reported to the power grid for approval and execution. For the basin where the cascade master plan is reported to the power grid for approval, the power generation plan of the entire cascade should match the power demand of the power grid as much as possible, with more power generation during the peak period of daytime and less power generation during the trough period of night. Therefore, it is necessary to deeply tap the peak regulation capacity of cascade hydropower stations. For cascade hydropower stations in the basin with certain regulation capacity, the peak regulation depth should be increased as much as possible during the dry season, while ensuring that the downstream power stations with poor regulation capacity do not abandon water; the peak regulation depth can be appropriately reduced during the normal water period, while meeting the requirement of not abandoning water; in the flood season, due to the large water inflow and the problem of abandoning water, it is difficult to avoid it. The peak regulation demand can be appropriately considered or not considered, and the units can be arranged to be fully generated as much as possible to increase the power generation efficiency. Taking the peak regulation depth as a factor to meet the needs of short-term power generation plan production in different periods has a high practical application value.

[0003] The research content of the existing cascade hydropower short-term power generation plan production problem: The first direction is to use the optimization scheduling theory to study, by setting the maximum power generation, maximum power generation efficiency, minimum water consumption rate and other goals, to meet the constraints of water balance, unit maintenance, water level limit, flow limit, etc., and use dynamic programming, genetic algorithm or a series of bionic algorithms and other optimization algorithms to find the optimal solution or local optimal solution that meets the conditions. The second direction is the practical short-term scheduling plan research, which does not seek the optimal solution, but seeks a feasible solution that meets the basic constraints, and then corrects it through artificial experience to meet the practical requirements. This method sets goals such as power generation of the power station or reservoir water level at the end of the scheduling period, and seeks a feasible solution that meets the initial constraints through the water balance algorithm.

[0004] Problems in current research content: The theoretical results sought through optimized dispatching methods often do not match practical applications. Due to rigid requirements such as power grid dispatching, group assessment, and ecological constraints, the actual arrangement of short-term power generation plans does not necessarily aim at a single goal such as maximizing power generation benefits, but rather meets comprehensive needs such as peak regulation, avoiding water abandonment, and high water level operation. It must also be consistent with the experience of dispatchers and match the recent dispatching plan. It is not a single extreme value problem, and the results obtained are not very practical. Although the dispatching results obtained by conventional methods can achieve the preset dispatching goals, the flexibility of the entire dispatching process is not high, and it cannot track the power demand situation of the power grid well. The preliminary results do not have the characteristics of more power generation during peak periods and less power generation during trough periods. It is necessary to intervene manually and adjust the power generation plan for certain periods manually, and then repeatedly calculate by using electricity to determine water until a short-term power generation plan that meets both the power or water level targets and the dispatching process can meet the power demand of the power grid is derived. Summary of the invention

[0005] The present invention aims to overcome the problem of insufficient practicability of optimized scheduling methods and conventional scheduling methods in making short-term power generation plans for cascade hydropower stations, and provides a method for making short-term power generation plans for cascade hydropower stations that can take into account the peak-shaving depth.

[0006] The technical solution adopted by the present invention to solve the above problem is: a method for making a short-term power generation plan of a cascade hydropower station taking into account the peak-shaving depth, which is characterized by comprising the following steps:

[0007] (1) Ideas for making short-term power generation plans considering peak load depth

[0008] The leading power station of the cascade power stations in the basin has a strong regulation capacity, and its main task is to undertake the peak-shaving and frequency-regulating tasks of the power grid. The main task of the downstream power station is to generate electricity in the high-efficiency area and abandon as little water as possible. Considering that different power stations have different generating heads, unit types and installed capacities, the amount of abandoned water represented by the same amount of abandoned water is also different. Therefore, short-term optimization scheduling is carried out with peak regulation as the goal. According to different water inflow conditions, flood season models, dry season models and normal water season models can be selected. The difference between these models lies in the peak regulation amplitude and the amount of abandoned water.

[0009] There is a lot of water coming in during the flood season. Due to the limited regulation performance of the downstream power station group, in order to avoid water abandonment, the output of the leading power station is not prone to large fluctuations. Therefore, the scheduling of the cascade hydropower station group is optimized with the goal of maximizing and minimizing the amount of abandoned water.

[0010] During the dry season, there is less water coming in, so the regulation and storage capacity of the leading power station can be fully utilized to reduce the amount of water abandoned in the cascade. Priority can be given to alleviating the peak-to-valley pressure of the power grid, with the goal of minimizing the maximum peak-to-valley difference of the power grid's residual load.

[0011] When the cascade water inflow is equal to the annual average water inflow, due to the limited regulation performance of the downstream cascade hydropower station group, excessive peak-shaving of the leading power station may cause the downstream cascade power stations to be unable to use their own storage capacity to absorb the unbalanced water volume. Therefore, it is planned to control the entire cascade output process by limiting the peak-shaving amplitude of the leading power station.

[0012] (2) Construction of a model for short-term power generation planning for cascade power stations in a river basin

[0013] S1. Determine the objective function.

[0014] S11. Flood season target: Optimize the dispatch of cascade hydropower stations by maximizing and minimizing the amount of abandoned water, as follows:

[0015]

[0016] Where: Q spill is the maximum amount of water abandoned at the cascade; It represents the water discharge flow of power station m in the tth period, in m 3 / s, Δt is the number of seconds in the tth time period, T is the total number of time periods; M represents the power station collection.

[0017] S12. Dry season target: The target is to minimize the maximum value of the peak-to-valley difference of the residual load of the power grid, as follows:

[0018]

[0019] Where: D represents the peak-to-valley difference of the residual load of the power grid, in MW; C t and X t They represent the original load and residual load values ​​in the tth period, in MW; It represents the output of the mth power station in the tth time period, in MW, where M is the number of hydropower stations participating in the optimization; T is the total number of time periods.

[0020] S13. Target during normal water period: Control the entire cascade output process by limiting the peak load amplitude of the leading power station, as follows:

[0021]

[0022] Where: and They represent the minimum and maximum output constraints of power station m at time period t, represents the output limit bandwidth of power station m in period t, which can be controlled by setting parameter δ The size of the bandwidth can be used to limit the peak load. In a scheduling cycle, the maximum and minimum output constraints may be different in different time periods, resulting in multiple The final optimized output result The difference between the maximum and minimum values ​​of The minimum value in is used to control the peak regulation depth.

[0023] S2. Determine model constraints: The constraints for making short-term power generation plans for cascade power stations in the basin that need to be considered include: water balance, water level constraints at the end of the dispatch period, power generation constraints during the dispatch period, power generation flow constraints, outflow flow constraints, power station output constraints, reservoir water level constraints and other non-negative constraints. When determining each constraint, it is necessary to combine the study of the cascade energy storage control strategy at key time nodes in the actual dispatch process of the basin and the peak-shaving capacity of the cascade power stations that meet the dispatch needs of the power grid.

[0024] 1) Water balance equation: ensure water balance in the time period of a single station and in the spatial dimension of upstream and downstream power stations

[0025]

[0026] In the formula, They are the inflow, power generation and abandoned water flow of power station m at period t, in m 3 / s; when m = 1, it means that power station m is the leading power station, and the inflow flow is equal to the interval flow When m>1, it means that power station m is a non-leading power station, and the inflow flow is equal to the interval flow. and the outflow flow of upstream power station m-1 The sum of the units is m 3 / s; and is the initial and final storage capacity of power station m at time period t, in m 3 ;Δt=15×60, unit is s.

[0027] 2) Final water level constraint: Control the final water level of the scheduling period

[0028]

[0029] Where: and They respectively represent the upper reservoir water level and control target value at the end of the dispatching period of power station m.

[0030] 3) Power generation constraint: Control the power generation during the dispatch period

[0031]

[0032] Where: E m and They represent the power generation and control target value of power station m during the dispatch period respectively. It is sufficient if one of the constraint conditions and formula (5) is satisfied.

[0033] 4) Power generation flow constraints: Generally, the upper limit depends on the maximum flow capacity of the turbine, the unit maintenance plan, etc., and the lower limit depends on the minimum output and ecological flow;

[0034]

[0035] Where: They represent the upper and lower limits of the power generation flow of power station m in the t period respectively.

[0036] 5) Power station output constraints: limit the power station's power output, depending on indicators such as the minimum technical output of the generator set and the unit maintenance capacity;

[0037]

[0038] Where: They are the upper and lower limits of the average output of the power station.

[0039] 6) Reservoir water level constraints: Set upper and lower limits of reservoir water level based on actual reservoir dispatching methods to ensure that the reservoir operates within a safe and reasonable water level range;

[0040]

[0041] Where: They represent the upper and lower limits of the water level of power station m on day t respectively.

[0042] S3. Determine the solution algorithm: When optimizing the leading power station, it is necessary to respond to the peak-shaving demand of the power grid. Under the premise of given initial and final water levels or daily power consumption, the maximum peak-to-valley difference of the power grid's residual load on that day should be minimized; when optimizing the downstream power station with poor regulation ability, try to ensure that the power station operates in the economic water level zone and try not to abandon water. At the same time, it is also necessary to ensure that the water consumption rate of the unit is as small as possible when distributing loads. Peak shaving refers to the fact that the power station greatly increases or decreases its output as the load of the power system increases or decreases. Therefore, the best time for the power station to generate electricity should be the highest load point. Based on this idea, a heuristic algorithm is used to solve the short-term optimization scheduling of cascade hydropower stations. The specific steps are as follows:

[0043] Step 1: Global iteration number k = 1;

[0044] Step 2: Determine power station m. If it is the leading power station, jump to Step 3; if it is other downstream power stations, jump to Step 7;

[0045] Step 3: Use the water balance check algorithm to determine whether the water level at the end of the dispatch period of the Longtou Power Station meets the preset target. If it does, the process jumps to Step 7.

[0046] Step 4: If the water level at the end of the dispatch period is higher than the predetermined target, search for the highest load point high on the load curve t, and including high t Find the point with the largest sum of v consecutive loads, and increase the power generation output of these v periods by ΔP. If the power station has reached the maximum output in these v periods, search for the second highest load point of the secondary load curve, and increase the power generation output of these v periods by ΔP, and so on.

[0047] Step 5: If the water level at the end of the dispatch period is lower than the predetermined target, search for the lowest load point of the load curve. t , and including low t The point where the sum of the continuous v loads including the point is the smallest, the power generation output of these v time periods is reduced by ΔP. If the power station output has reached the minimum value in these v time periods, search for the second lowest load point of the secondary load curve, and reduce the power generation output of these v time periods by ΔP, and so on;

[0048] Step 6: Process the output duration constraint, output ramp constraint and water balance constraint, and set the power station m=m+1;

[0049] Step 7: Search for the downstream power station m and generate the initial solution using the fixed water level method;

[0050] Step 8: Processing duration constraints, output ramp constraints, and water balance constraints;

[0051] Step 9: According to the relationship between the power generation flow of the power station and the maximum flow of the unit, adjust the power station output to the economic operation area;

[0052] Step 10: When the final water level is not within the economic zone constraint, set power station m = m + 1 and jump to Step 7 until all power stations are traversed;

[0053] Step 11: Determine whether the water level at the end of the dispatch period of the leading power station meets the preset target. If not, set k=k+1 and jump to Step 2. If the water level at the end of the dispatch period of the leading power station meets the preset target, exit.

[0054] Compared with the prior art, the present invention has the following advantages and effects:

[0055] 1. Aiming at the different peak-shaving amplitudes and water abandonment volumes of cascade hydropower stations in the basin at different times, the present invention establishes models for flood season, dry season and normal water season respectively, takes peak-shaving at different depths and reducing water abandonment as the goal, adopts a heuristic algorithm, and formulates a short-term power generation plan. This method is highly practical for short-term scheduling of cascade hydropower stations with limited regulation capacity, can adapt to the needs of power generation plan making under different water inflow conditions, and is more flexible than the optimization scheduling method and the conventional scheduling method.

[0056] 2. As the regulation performance of the downstream cascade hydropower station group is limited, excessive peak regulation of the leading power station may cause the downstream cascade power station to be unable to use its own storage capacity to absorb the unbalanced water volume and thus abandon water. By variably adjusting the peak regulation depth of the leading power station to control the entire cascade output process, the power generation plan of the entire cascade power station can be well matched to the electricity demand of the power grid, while at the same time reducing the amount of water abandoned by the power stations with poor regulation capabilities in the cascade power stations, achieving a good balancing effect. DETAILED DESCRIPTION

[0057] The present invention will be further described in detail below by way of examples. The following examples are provided to explain the present invention but the present invention is not limited to the following examples.

[0058] Example

[0059] In this embodiment, a method for preparing a short-term power generation plan of a cascade hydropower station taking into account the peak load depth includes the following steps:

[0060] Step 1: Determine the idea of ​​making a short-term power generation plan considering the peak load depth;

[0061] Step 2: According to the different characteristics of peak-shaving amplitude and water abandonment in different periods, models for flood season, dry season and normal water season are established respectively. With the goal of peak-shaving at different depths and reducing water abandonment, the model constraints are determined in combination with the cascade energy storage control strategy at key time nodes in the actual dispatching process and the peak capacity of cascade power stations that meet the power demand of the power grid. The constraints include water balance, reservoir water level constraints, final water level control, peak-shaving depth constraints, power generation flow constraints, outflow flow constraints, power station output constraints, etc., and the model solving algorithm is determined;

[0062] Step 3: Use heuristic algorithm to solve short-term optimal scheduling.

[0063] In step 2, the model construction considering the peak regulation depth in different water inflow periods includes determining the objective function, determining the constraints of the model and the solution algorithm.

[0064] S1. Determine the objective function.

[0065] S11 flood season target: Optimize the dispatch of cascade hydropower stations by maximizing and minimizing the amount of abandoned water, as follows:

[0066]

[0067] Where: Q spill is the maximum amount of water abandoned at the cascade; It represents the water discharge flow of power station m in the tth period, in m 3 / s, Δt is the number of seconds in the tth time period, T is the total number of time periods; M represents the power station collection.

[0068] S12 dry season target: The target is to minimize the maximum value of the peak-to-valley difference of the grid residual load, as follows:

[0069]

[0070] Where: D represents the peak-to-valley difference of the residual load of the power grid, in MW; Ct and Xt represent the original load and residual load values ​​in the tth period, in MW; It represents the output of the mth power station in the tth time period, in MW, M is the number of hydropower stations participating in the optimization; T is the total number of time periods.

[0071] S13 normal water period target: Control the entire cascade output process by limiting the peak regulation amplitude of the leading power station, as follows:

[0072]

[0073] Where: They represent the minimum and maximum output constraints of power station m at time period t, represents the output limit bandwidth of power station m in period t, which can be controlled by setting parameter δ The size of the bandwidth can be used to limit the peak load. In a scheduling cycle, the maximum and minimum output constraints in different periods may be different, resulting in multiple The final optimized output result The difference between the maximum and minimum values ​​of The minimum value in is used to control the peak regulation depth.

[0074] S2. Determine model constraints: The constraints for making short-term power generation plans for cascade power stations in the basin that need to be considered include: water balance, water level constraints at the end of the dispatch period, power generation constraints during the dispatch period, power generation flow constraints, outflow flow constraints, power station output constraints, reservoir water level constraints and other non-negative constraints. When determining each constraint, it is necessary to combine the study of the cascade energy storage control strategy at key time nodes in the actual dispatch process of the basin and the peak-shaving capacity of the cascade power stations that meet the dispatch needs of the power grid.

[0075] 1) Water balance equation: ensure water balance in the time period of a single station and in the spatial dimension of upstream and downstream power stations

[0076]

[0077] In the formula, They are the inflow, power generation and abandoned water flow of power station m at period t, in m 3 / s; when m = 1, it means that power station m is the leading power station, and the inflow flow is equal to the interval flow When m>1, it means that power station m is a non-leading power station, and the inflow flow is equal to the interval flow. and the outflow flow of upstream power station m-1 The sum of the units is m 3 / s; and is the initial and final storage capacity of power station m at time period t, in m 3 ;Δt=15×60, unit is s.

[0078] 2) Final water level constraint: Control the final water level of the scheduling period

[0079]

[0080] Where: and They respectively represent the upper reservoir water level and control target value at the end of the dispatching period of power station m.

[0081] 3) Power generation constraint: Control the power generation during the dispatch period

[0082]

[0083] Where: E m and They represent the power generation and control target value of power station m during the dispatch period respectively. It is sufficient if one of the constraint conditions and formula (5) is satisfied.

[0084] 4) Power generation flow constraints: Generally, the upper limit depends on the maximum flow capacity of the turbine and the unit maintenance plan, and the lower limit depends on the minimum output and ecological flow;

[0085]

[0086] Where: They represent the upper and lower limits of the power generation flow of power station m in the t period respectively.

[0087] 5) Power station output constraint: Limiting the power station's power output depends on indicators such as the minimum technical output of the generator set and the unit maintenance capacity;

[0088]

[0089] Where: They are the upper and lower limits of the average output of the power station.

[0090] 6) Reservoir water level constraints: Set upper and lower limits of reservoir water level based on actual reservoir dispatching methods to ensure that the reservoir operates within a safe and reasonable water level range;

[0091]

[0092] Where: They represent the upper and lower limits of the water level of power station m on day t respectively.

[0093] S3. Determine the solution algorithm: When optimizing the leading power station, it is necessary to respond to the peak-shaving demand of the power grid. Under the premise of given initial and final water levels or daily power consumption, the maximum peak-to-valley difference of the power grid's residual load on that day should be minimized; when optimizing the downstream power station with poor regulation ability, try to ensure that the power station operates in the economic water level zone and try not to abandon water. At the same time, it is also necessary to ensure that the water consumption rate of the unit is as small as possible when distributing loads. Peak shaving refers to the fact that the power station greatly increases or decreases its output as the load of the power system increases or decreases. Therefore, the best time for the power station to generate electricity should be the highest load point. Based on this idea, a heuristic algorithm is used to solve the short-term optimization scheduling of cascade hydropower stations. The specific steps are as follows:

[0094] Step 1: Global iteration number k = 1;

[0095] Step 2: Determine power station m. If it is the leading power station, jump to Step 3; if it is other downstream power stations, jump to Step 7;

[0096] Step 3: Use the water balance algorithm to check the water balance and determine whether the water level at the end of the dispatch period of the Longtou Power Station meets the preset target. If it does, the direct current jumps to Step 7.

[0097] Step 4: If the water level at the end of the dispatch period is higher than the predetermined target, search for the highest load point high on the load curve t , and package high t The point with the largest sum of v consecutive loads including the point, increase the power generation output of these v periods by ΔP. If the power station has reached the maximum output in these v periods, search for the second highest load point of the secondary load curve, and increase the power generation output of these v periods by ΔP, and so on;

[0098] Step 5: If the water level at the end of the dispatch period is lower than the predetermined target, search for the lowest load point of the load curve. t , and low t Find the point where the sum of the consecutive v loads including the point is the smallest, reduce the power generation output of these v periods by ΔP. If the power station output has reached the minimum value in these v periods, search for the second lowest load point of the secondary load curve, and reduce the power generation output of these v periods by ΔP, and so on;

[0099] Step 6: Process the output duration constraint, output ramp constraint and water balance constraint, and set the power station m=m+1;

[0100] Step 7: Search for the downstream power station m and generate the initial solution using the fixed water level method;

[0101] Step 8: Processing duration constraints, output ramp constraints, and water balance constraints;

[0102] Step 9: According to the relationship between the power generation flow of the power station and the maximum flow of the unit, adjust the power station output to the economic operation area;

[0103] Step 10: When the final water level is not within the economic zone constraint, set power station m = m + 1 and jump to Step 7 until all power stations are traversed;

[0104] Step 11: Determine whether the water level at the end of the dispatch period of the leading power station meets the preset target. If not, set k=k+1 and jump to Step 2. If the water level at the end of the dispatch period of the leading power station meets the preset target, exit.

[0105] In step 2, the model construction considering the peak regulation depth during the normal water period and the control method of the peak regulation depth are carried out.

[0106] Target during normal water season: Control the entire cascade output process by limiting the peak load amplitude of the leading power station, as follows:

[0107]

[0108] Where: and They represent the minimum and maximum output constraints of power station m at time period t, represents the output limit bandwidth of power station m in period t, which can be controlled by setting parameter δ The size of the bandwidth can be used to limit the peak shaving depth.

[0109] In a scheduling cycle, the maximum and minimum output constraints may be different in different time periods, resulting in multiple The final optimized output result The difference between the maximum and minimum values ​​of The minimum value in is used to control the peak load depth. By manually setting the peak load coefficient, the peak load depth of the short-term power generation plan can be controlled. Through trial calculation, the peak load effect and the abandoned water volume target of the cascade power generation plan can be balanced, making the calculation result more in line with the actual dispatching situation.

[0110] The contents not described in detail in this specification belong to the prior art known to professional and technical personnel in this field.

[0111] Although the present invention has been disclosed as above by way of embodiments, it is not intended to limit the protection scope of the present invention. Any changes and modifications made by any technician familiar with the technology without departing from the concept and scope of the present invention should fall within the protection scope of the present invention.

Claims

1. A method for making a short-term power generation plan for a cascade hydropower station taking into account the peak-shaving depth, characterized in that: The following steps are involved: Step 1: Determine the idea of ​​making a short-term power generation plan considering the peak load depth; Step 2: According to the different characteristics of peak-shaving amplitude and water abandonment in different periods, models for flood season, dry season and normal water season are established respectively. With the goal of peak-shaving at different depths and reducing water abandonment, the model constraints are determined in combination with the cascade energy storage control strategy at key time nodes in the actual dispatching process and the peak capacity of cascade power stations that meet the power demand of the power grid. The constraints include water balance, reservoir water level constraints, final water level control, peak-shaving depth constraints, power generation flow constraints, outflow flow constraints, and power station output constraints, and the model solution algorithm is determined; Step 3: Use heuristic algorithm to solve short-term optimal scheduling; The model construction considering the peak regulation depth in different water inflow periods in step 2 includes determining the objective function, determining the constraints of the model and the solution algorithm; S1. Determine the objective function; S11 flood season target: Optimize the dispatch of cascade hydropower stations by maximizing and minimizing the amount of abandoned water, as follows: Where: Qspill is the maximum amount of water discarded in the cascade; It represents the water discharge flow of power station m in the tth period, in m 3 / s, Δt is the number of seconds in the tth period, T is the total number of periods; M represents the power station set; S12 dry season target: The target is to minimize the maximum value of the peak-to-valley difference of the grid residual load, as follows: Where: D represents the peak-to-valley difference of the residual load of the power grid, in MW; Ct and Xt represent the original load and residual load values ​​in the tth period, in MW respectively; It represents the output of the mth power station in the tth period, in MW, M is the number of hydropower stations participating in the optimization; T is the total number of periods; S13 normal water period target: Control the entire cascade output process by limiting the peak regulation amplitude of the leading power station, as follows: Where: They represent the minimum and maximum output constraints of power station m at time period t, represents the output limit bandwidth of power station m in period t, which is controlled by setting parameter δ The size of the bandwidth is used to limit the peak load. In a scheduling cycle, the maximum and minimum output constraints in different periods are different, resulting in multiple The final optimized output result The difference between the maximum and minimum values ​​of The minimum value in is used to control the peak regulation depth; S2. Determine model constraints: The constraints for making short-term power generation plans for cascade power stations in the basin that need to be considered include: water balance, water level constraints at the end of the dispatch period, power generation constraints during the dispatch period, power generation flow constraints, outflow flow constraints, power station output constraints, and reservoir water level constraints. When determining each constraint, combine the study of the cascade energy storage control strategy at key time nodes in the actual dispatch process of the basin and the peak-shaving capacity of the cascade power stations that meet the dispatch needs of the power grid; 1) Water balance equation: ensure water balance in the time period of a single station and in the spatial dimension of upstream and downstream power stations in, In the formula, They are the inflow, power generation and abandoned water flow of power station m at period t, in m 3 / s; when m = 1, it means that power station m is the leading power station, and the inflow flow is equal to the interval flow When m>1, it means that power station m is a non-leading power station, and the inflow flow is equal to the interval flow. and the outflow flow of upstream power station m-1 The sum of the units is m 3 / s; and is the initial and final storage capacity of power station m at time period t, in units of m 3 ; Δt = 15 × 60, the unit is s; 2) Final water level constraint: Control the final water level of the scheduling period Where: and They represent the upper reservoir water level and control target value at the end of the dispatch period of power station m respectively; 3) Power generation constraint: Control the power generation during the dispatch period Where: E m and They represent the power generation and control target value of power station m during the dispatch period respectively. This constraint condition and formula (5) can satisfy one of them; 4) Power generation flow constraints: The upper limit depends on the maximum flow capacity of the turbine and the unit maintenance plan, and the lower limit depends on the minimum output and ecological flow; Where: They represent the upper and lower limits of power generation flow of power station m in period t respectively; 5) Power station output constraint: Limiting the power station's power output depends on the minimum technical output of the generator set and the unit maintenance capacity index; Where: They are the upper and lower limits of the average power output of the power station; 6) Reservoir water level constraints: Set upper and lower limits of reservoir water level based on actual reservoir dispatching methods to ensure that the reservoir operates within a safe and reasonable water level range; Where: They represent the upper and lower limits of the water level of power station m on day t respectively; S3. Determine the solution algorithm: When optimizing the leading power station, respond to the peak-shaving demand of the power grid. Under the premise of given initial and final water levels or daily power, minimize the maximum peak-to-valley difference of the power grid's residual load on that day; when optimizing the downstream power station with poor regulation ability, ensure that the power station operates in the economic water level zone and does not abandon water, while ensuring that the water consumption rate of the unit is as small as possible when distributing loads; peak-shaving refers to the power station's substantial increase or decrease in output as the power system load increases or decreases, and the best time for the power station to generate electricity is the highest load point. Based on this idea, a heuristic algorithm is used to solve the short-term optimal scheduling of cascade hydropower stations. The specific steps are as follows: Step 1: Global iteration number k = 1; Step 2: Determine power station m. If it is the leading power station, jump to Step 3; if it is other downstream power stations, jump to Step 7; Step 3: Use the water balance algorithm to check the water balance and determine whether the water level at the end of the dispatch period of the Longtou Power Station meets the preset target. If it does, the direct current jumps to Step 7. Step 4: If the water level at the end of the dispatch period is higher than the predetermined target, search for the highest load point high on the load curve t , and package high t The point with the largest sum of v consecutive loads including the point, increase the power generation output of these v periods by ΔP. If the power station has reached the maximum output in these v periods, search for the second highest load point of the secondary load curve, and increase the power generation output of these v periods by ΔP, and so on; Step 5: If the water level at the end of the dispatch period is lower than the predetermined target, search for the lowest load point of the load curve. t , and low t Find the point where the sum of the consecutive v loads including the point is the smallest, reduce the power generation output of these v periods by ΔP. If the power station output has reached the minimum value in these v periods, search for the second lowest load point of the secondary load curve, and reduce the power generation output of these v periods by ΔP, and so on; Step 6: Process the output duration constraint, output ramp constraint and water balance constraint, and set the power station m=m+1; Step 7: Search for the downstream power station m and generate the initial solution using the fixed water level method; Step 8: Processing duration constraints, output ramp constraints, and water balance constraints; Step 9: According to the relationship between the power generation flow of the power station and the maximum flow of the unit, adjust the power station output to the economic operation area; Step 10: When the final water level is not within the economic zone constraint, set power station m = m + 1 and jump to Step 7 until all power stations are traversed; Step 11: Determine whether the water level at the end of the dispatch period of the leading power station meets the preset target. If not, set k=k+1 and jump to Step 2. If the water level at the end of the dispatch period of the leading power station meets the preset target, exit.