A successive dimensionality reduction method for economic dispatch of cascade hydropower stations adapting to complex constraints

Abstract

The invention discloses a method for successive dimensionality reduction of cascade hydropower station economic dispatch adapting to complex constraints, which is specifically implemented in the following steps: step 1, establishing an inter-plant economic operation objective function and an intra-plant economic operation objective function; step 2, setting constraint conditions for the inter-plant economic operation objective function and the intra-plant economic operation objective function of step 1; step 3, using the DRS-DSA algorithm to solve the inter-plant economic operation objective function. The method of the invention solves the problems of unstable results, long calculation time and poor effectiveness obtained by the existing method.

Description

Technical Field

[0001] The invention belongs to the technical field of economic dispatching of hydropower stations, and relates to a successive dimensionality reduction method for economic dispatching of cascade hydropower stations adapting to complex constraints. Background Art

[0002] The economic dispatch problem of cascade hydropower stations has a large and complex set of constraints. The complex constraints make the feasible solution space discontinuous, which greatly increases the difficulty of searching for the optimal solution. In order to solve this type of cascade hydropower station economic dispatch problem, domestic and foreign scholars have conducted a lot of research and proposed a variety of linear programming, dynamic programming and intelligent evolution algorithms. However, when faced with the complex hydraulic connections of cascade hydropower stations, a large number of decision variables, and complex constraints, the stability and effectiveness of the calculation results of intelligent evolution algorithms such as genetic algorithms and particle swarm algorithms are difficult to guarantee; dynamic programming algorithms increase the calculation time and storage space on a large scale with the increase of discrete precision of decision variables, the number of power stations, the number of time periods, and other factors, resulting in the problem of "dimensionality curse".

[0003] In this context, "dimensionality reduction" has become one of the hot topics in hydropower system research at home and abroad. In general, the dimensionality reduction strategy can be divided into spatial dimensionality reduction to reduce the number of cascade hydropower stations to be optimized. The dimensionality reduction strategy greatly reduces the computational complexity of the cascade hydropower station optimization scheduling problem, ensuring the stability and effectiveness of the calculation results; at the same time, it improves the computational efficiency and avoids the problem of "dimensionality curse". Summary of the invention

[0004] The purpose of the present invention is to provide a method for economic dispatch of cascade hydropower stations adapting to complex constraints and reducing dimensions step by step, which solves the problems of unstable results, long calculation time and poor effectiveness obtained by existing methods.

[0005] The technical solution adopted by the present invention is a method for economic dispatch of cascade hydropower stations that adapts to complex constraints and reduces the dimension step by step.

[0006] Follow the steps below to implement it:

[0007] Step 1, establishing the inter-factory economic operation objective function and the intra-factory economic operation objective function;

[0008] Step 2, setting constraints on the inter-plant economic operation objective function and the intra-plant economic operation objective function of step 1;

[0009] Step 3: Use the DRS-DSA algorithm to solve the inter-plant economic operation objective function.

[0010] The present invention is also characterized in that

[0011] The expression of the inter-factory economic operation objective function is:

[0012]

[0013] In formula (1), F is the total energy storage of the cascade power station at the end of the period; E s (n) is the amount of electricity that can be generated at power station n by the water storage above the dead water level of power station n and all its upstream power stations; E lag (n) is the amount of electricity generated by the upstream power station due to water lag at power station n;

[0014] in,

[0015]

[0016]

[0017]

[0018] In formula (2), W(n) is the water storage capacity above the dead water level of power station n at the end of the dispatch period; W up (n) is the water storage capacity of all upstream power stations above the dead water level at the end of the dispatch period; W lag (n) is the delayed water volume generated by all power stations upstream of power station n; Q out (k,t) is the outflow flow of power station k in period t (Q out (k,t)=Q qs,n-1 (t-τ)+Q fd,n-1 (t-τ), that is, Q out (k,t)=Q qs,n-1 (t-τ)+Q); τ k is the water flow lag time between power station k and power station k+1; Δt is the time period length; T is the total number of time periods; η n is the average water consumption rate of power station n during the dispatch period;

[0019] The expression of the objective function of the plant's economic operation is:

[0020]

[0021] In formula (3), Q is the water consumption of the power station; Q l (N l ,H) is the output of unit 1 under water head H, N l The power generation flow at that time; L is the number of units under a certain unit combination;

[0022] According to the optimization principle of dynamic programming, formula (3) is recursively derived, then:

[0023]

[0024] In formula (4), is the total output of units 1 to 1; When the water head is H, the total output is Total power generation flow when optimally distributed among units 1 to 1; It is the load at the initial stage of the power station; When the water head is H, the output at the initial stage is The total power generation flow when optimally distributed among units 1 to 1; R l is the load range of unit l; When the water head is H, the output The power generation flow when optimizing the distribution between units 1 to l-1.

[0025] The constraints include: single period constraints, period coupling constraints and water level coordination constraints.

[0026] Single-period constraints include: cascade hydropower station output constraints, water balance constraints, reservoir capacity constraints, reservoir outflow constraints, power station spinning reserve capacity constraints, turbine head limit constraints, and power station restricted area constraints;

[0027] (1) Output constraints of cascade hydropower stations:

[0028]

[0029] In formula (5), P D (t) is the total output of the cascade power stations given by the power grid during period t, MW; P n (t) is the output of power station n during period t, MW; N l (t) is the output of unit 1 during period t, MW;

[0030] (2) Water balance constraints:

[0031]

[0032] In formula (6), V n (t) is the storage capacity of the reservoir where the power station n is located during time period t, m 3 ;I n (t) is the inflow flow of the reservoir where the power station n is located during period t, m 3 / s;Q fd,n-1 (t-τ) is the power generation flow of the reservoir where the power station n-1 is located during the t-τ period, m 3 / s;Q qs,n-1 (t-τ) is the water discharge of the reservoir where the power station n-1 is located during the period t-τ, m 3 / s; τ is the water flow lag time between the reservoir where power station n-1 is located and the reservoir where power station n is located, m 3 / s;Q qj,n (t) is the interval inflow between the reservoir where power station n-1 is located and the reservoir where power station n is located, m 3 / s;Q n(t) is the outflow rate of the reservoir where power station n is located during period t, m 3 / s;Q qs,n-1 (t) is the water discharge of the reservoir where power station n-1 is located during period t; H n (t) is the water head of the reservoir where power station n is located in period t, m; f zv,n (·) is the functional relationship between the water level and storage capacity curve of the reservoir where the power station n is located; f zq,n (·) is the functional relationship between the tailwater level and flow rate curve of the reservoir where the power station n is located; f Δh,n (·) is the curve function relationship of the head loss of the reservoir where the power station n is located; f zv,n (·) and f zq,n (·) can generally be obtained by fitting the corresponding measured data, f Δh,n (·) can generally be considered as the outbound flow Q n (t) is a quadratic relationship or constant;

[0033] (3) Reservoir capacity constraints:

[0034] V n min (t)≤V n (t)≤V n max (t) (7)

[0035] In formula (7), V n max (t) is the upper limit of the storage capacity of reservoir n during period t; V n min (t) is the lower limit of reservoir capacity of reservoir n during period t, m 3 ;

[0036] (4) Reservoir outflow constraints:

[0037]

[0038] In formula (8), Q n max (t) is the upper limit of the outflow of the reservoir during period t, m 3 / s;Q n min (t) is the lower limit of the outflow of the reservoir during period t, m 3 / s;

[0039] (5) Power station spinning reserve capacity constraints:

[0040] P n max ≥P n (t)+P R (t) (9)

[0041] In formula (9): Pn max is the upper limit of the output allowed by power station n, MW; P R (t) is the spinning reserve capacity of power station n during period t, MW;

[0042] (6) Turbine head limitation constraints:

[0043]

[0044] In formula (10), H n max (t) is the upper limit of the turbine head of power station n in period t, m; H n min (t) is the lower limit of turbine head at power station n in period t, m; H n (t) is the turbine head of power station n in time period t, m;

[0045] (7) Power station restricted area constraints:

[0046]

[0047] In formula (11), is the upper limit of the mth restriction zone of power station n in time period t; is the lower limit of the mth restriction zone of power station n in time period t.

[0048] Time period coupling constraints include: power station output ramp constraints, unit minimum start and stop time constraints;

[0049] (1) Power station output ramp constraints:

[0050] |P n (t)-P n (t-1)|≤PC n (12)

[0051] In formula (12), PC n is the maximum output fluctuation of power station n between adjacent time periods; n (t-1) is the output of power station n during period t-1, MW;

[0052] (2) Minimum start-up and shutdown time constraints of the unit:

[0053]

[0054] In formula (13), is the shortest startup time of unit l; is the shortest downtime of unit l; is the continuous operation time of unit l before time period t-1; is the continuous downtime of unit l before period t-1.

[0055] Water level coordination constraint: There are upper and lower water level limits during power station operation. When cascade hydropower stations are operated together, the power stations should be in the economic operation area or the upper water level limit area or the lower water level limit area at the same time;

[0056] Among them, the lower limit water level is less than the economic operation zone and the upper limit water level.

[0057] Step 3 is implemented as follows:

[0058] Step 3.1, initial value determination

[0059] Assume the total output P of n power stations 总 And the output ratio of each power station in several groups, and get the output P of each power station in each group n , the water consumption Q of the power station is obtained by formula (4) and formula (3), and then the total energy storage F of the cascade power station at the end of the period is obtained by formula (6), formula (2), and formula (1) in turn; the total energy storage F of several cascade power stations at the end of the period is compared, and the largest total energy storage F of the cascade power station at the end of the period is selected as the initial total energy storage F of the cascade power station at the end of the period; the output P of each power station corresponding to the total energy storage F of the initial cascade power station at the end of the period is calculated. n Provides initial output for each power station;

[0060] Step 3.2: Dimensionality reduction optimization

[0061] The total number of time periods T is divided into several stages, each stage contains several time periods, the number of time periods in each stage is equal and the number of stages can be divided by the total number of time periods T, and in each time period, the same increase and decrease operation is performed on the initial output of each power station in step 3.1 by the Box-Behnken Design method, and then according to the calculation process in step 3.1, the total energy storage F of the cascade power station at the end of the period under the increase and decrease operation is obtained, and the total energy storage F of the cascade power station at the end of the period under the increase and decrease operation is compared, and the largest total energy storage F of the cascade power station at the end of the period is selected as the optimized total energy storage F of the cascade power station at the end of the period; the output of each power station after the increase and decrease operation corresponding to the optimized total energy storage F of the cascade power station at the end of the period is the optimized output of each power station;

[0062] Step 3.3, step-by-step optimization

[0063] Increase or decrease the optimized output of each power station in step 3.2, with a step size of 5MW to 10MW. Then, according to the calculation process in step 3.1, the optimal total energy storage F of the cascade power stations at the end of the period is obtained. The optimal total energy storage F of the cascade power stations at the end of the period is compared, and the largest total energy storage F of the cascade power station at the end of the period is selected as the optimal total energy storage F of the cascade power station at the end of the period; the output of each power station corresponding to the optimal total energy storage F of the cascade power station at the end of the period is taken as the optimal output of each power station.

[0064] The beneficial effect of the present invention is that the present invention adopts the DRS-DSA algorithm, and through three dimensionality reductions, namely, initial value determination, dimensionality reduction optimization, and step-by-step optimization, obtains the optimal economic dispatch model of cascade hydropower stations, namely, the inter-plant economic operation objective function and the intra-plant economic operation objective function. The result is stable and time-efficient, has good effectiveness and good applicability to constraint conditions, thereby providing a reliable solution and basis for the economic dispatch of cascade hydropower stations. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a comparison diagram between a successive dimensionality reduction method for economic dispatch of cascade hydropower stations adapting to complex constraints and a DSA algorithm of the present invention. DETAILED DESCRIPTION

[0066] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0067] The present invention provides a method for economic dispatch of cascade hydropower stations adapting to complex constraints and reducing dimensions step by step, which is specifically implemented according to the following steps:

[0068] Step 1, establishing the inter-factory economic operation objective function and the intra-factory economic operation objective function;

[0069] In the "power-based water" mode, in order to consider the energy effect of reservoir water storage on different power stations, the maximum total energy storage of cascade power stations at the end of the period is taken as the goal, and the expression of the inter-plant economic operation objective function is:

[0070]

[0071] In formula (1), F is the total energy storage of the cascade power station at the end of the period; E s (n) is the amount of electricity that can be generated at power station n by the water storage above the dead water level of power station n and all its upstream power stations; E lag (n) is the amount of electricity generated by the upstream power station due to water lag at power station n;

[0072] in,

[0073]

[0074]

[0075]

[0076] In formula (2), W(n) is the water storage capacity of power station n above the dead water level at the end of the dispatch period, that is, the reservoir capacity V of the power station at the end of the dispatch period. n Reservoir capacity below dead water level V dead The difference between up (n) is the water storage capacity of all upstream power stations above the dead water level at the end of the dispatch period; W lag(n) is the delayed water volume generated by all power stations upstream of power station n; Q out (k,t) is the outflow flow of power station k in period t (Q out (k,t)=Q qs,n-1 (t-τ)+Q fd,n-1 (t-τ), that is, Q out (k,t)=Q qs,n-1 (t-τ)+Q); τ k is the water flow lag time between power station k and power station k+1; Δt is the time period length; T is the total number of time periods; η n is the average water consumption rate of power station n during the dispatch period;

[0077] The purpose of plant economic operation is to solve the problem of unit output allocation when the whole plant load is constant in each period. Minimum water consumption is the main goal of plant economic operation. In order to improve the solution speed, the results of plant economic operation are presented in the form of offline economic operation tables under different unit combinations. Different unit combinations of power stations are used as classifications. When the internal units have different heads and different outputs, the output and power generation flow of each unit under the goal of minimum water consumption are expressed as follows:

[0078]

[0079] In formula (3), Q is the water consumption of the power station; Q l (N l ,H) is the output of unit 1 under water head H, N l The power generation flow at that time; L is the number of units under a certain unit combination;

[0080] According to the optimization principle of dynamic programming, formula (3) is recursively derived, then:

[0081]

[0082] In formula (4), is the total output of units 1 to 1, where It is also the output P allocated to power station n during the joint dispatch of cascade power stations. n ; When the water head is H, the total output is Total power generation flow when optimally distributed among units 1 to 1; It is the load at the initial stage of the power station; When the water head is H, the output at the initial stage is The total power generation flow when optimally distributed among units 1 to 1; R l is the load range of unit l; When the water head is H, the output The power generation flow rate when optimizing the distribution between units 1 to l-1;

[0083] Step 2, setting constraints for the inter-plant economic operation objective function and the intra-plant economic operation objective function of step 1, the constraints include: single time period constraint, time period coupling constraint and water level coordination constraint;

[0084] Single-period constraints include: cascade hydropower station output constraints, water balance constraints, reservoir capacity constraints, reservoir outflow constraints, power station spinning reserve capacity constraints, turbine head limit constraints, and power station restricted area constraints;

[0085] (1) Output constraints of cascade hydropower stations:

[0086]

[0087] In formula (5), P D (t) is the total output of the cascade power stations given by the power grid during period t, MW; P n (t) is the output of power station n during period t, MW; N l (t) is the output of unit 1 during period t, MW;

[0088] (2) Water balance constraints:

[0089]

[0090] In formula (6), V n (t) is the storage capacity of the reservoir where the power station n is located during time period t, m 3 ;I n (t) is the inflow flow of the reservoir where the power station n is located during period t, m 3 / s;Q fd,n-1 (t-τ) is the power generation flow of the reservoir where the power station n-1 is located during the t-τ period (Q fd,n-1 (t-τ)=Q),m 3 / s;Q qs,n-1 (t-τ) is the water discharge of the reservoir where the power station n-1 is located during the period t-τ, m 3 / s, where the abandoned water flow is the outflow flow Q of the reservoir where the power station n is located during the period t-τ n When (t-τ) is greater than the full power flow of the power station, the difference between the full power flow of the power station, or V n (t+1)>V n max (t), the abandoned water flow Q qs,n-1 (t-τ)=V n (t+1)-V n max (t)-Q n (t); τ is the water lag time between the reservoir where power station n-1 is located and the reservoir where power station n is located, m 3 / s;Q qj,n(t) is the interval inflow between the reservoir where power station n-1 is located and the reservoir where power station n is located, m 3 / s;Q n (t) is the outflow rate of the reservoir where power station n is located during period t, m 3 / s;Q qs,n-1 (t) is the water discharge of the reservoir where power station n-1 is located during period t; H n (t) is the water head of the reservoir where power station n is located in period t, m; f zv,n (·) is the functional relationship between the water level and storage capacity curve of the reservoir where the power station n is located; f zq,n (·) is the functional relationship between the tailwater level and flow rate curve of the reservoir where the power station n is located; f Δh,n (·) is the curve function relationship of the head loss of the reservoir where the power station n is located; f zv,n (·) and f zq,n (·) can generally be obtained by fitting the corresponding measured data, f Δh,n (·) can generally be considered as the outbound flow Q n (t) is a quadratic relationship or constant;

[0091] (3) Reservoir capacity constraints:

[0092] V n min (t)≤V n (t)≤V n max (t) (7)

[0093] In formula (7), V n max (t) is the upper limit of the storage capacity of reservoir n during period t; V n min (t) is the lower limit of reservoir capacity of reservoir n during period t, m 3 ;

[0094] (4) Reservoir outflow constraints:

[0095]

[0096] In formula (8), Q n max (t) is the upper limit of the outflow of the reservoir during period t, m 3 / s;Q n min (t) is the lower limit of the outflow of the reservoir during period t, m 3 / s;

[0097] (5) Power station spinning reserve capacity constraints:

[0098] P n max ≥P n(t)+P R (t) (9)

[0099] In formula (9): P n max is the upper limit of the output allowed by power station n, MW; P R (t) is the spinning reserve capacity of power station n during period t, MW;

[0100] (6) Turbine head limitation constraints:

[0101]

[0102] In formula (10), H n max (t) is the upper limit of the turbine head of power station n in period t, m; H n min (t) is the lower limit of turbine head at power station n in period t, m; H n (t) is the turbine head of power station n in time period t, m;

[0103] (7) Power station restricted area constraints:

[0104]

[0105] In formula (11), is the upper limit of the mth restriction zone of power station n in time period t; is the lower limit of the mth restriction zone of power station n in time period t;

[0106] Time period coupling constraints include: power station output ramp constraints, unit minimum start and stop time constraints;

[0107] (1) Power station output ramp constraints:

[0108] |P n (t)-P n (t-1)|≤PC n (12)

[0109] In formula (12), PC n is the maximum output fluctuation of power station n between adjacent time periods; n (t-1) is the output of power station n during period t-1, MW;

[0110] (2) Minimum start-up and shutdown time constraints of the unit:

[0111]

[0112] In formula (13), is the shortest startup time of unit l; T l down is the shortest downtime of unit l; Tl on (t) is the continuous operation time of unit 1 before the period t-1; T l off (t) is the continuous downtime of unit 1 before the period t-1;

[0113] Water level coordination constraint: The power station is set with an upper limit water level (not greater than the normal water storage level) and a lower limit water level. In order to reserve sufficient time for emergency operations such as "black start" (meaning that after the entire system stops operating due to a fault, the system can recover on its own without relying on other network help) and sudden floods, the power station will set an upper limit water level during operation. In order to prevent the reservoir water level from being too low and affecting the power generation of the power station in the subsequent period, or the water level in front of the dam is too low at the current moment and affecting the water diversion flow of the turbine unit, the power station will also set a lower limit water level. During the operation of the power station, it should avoid operating for a long time above the upper limit water level and below the lower limit water level. When the cascade hydropower station is operated jointly, the power station should be in the economic operation area (lower limit water level < economic operation area < upper limit water level) or in the high limit water level area or the low limit water level area at the same time. The power station output mode under different combinations of reservoir water levels of the cascade hydropower station is shown in the following table:

[0114]

[0115] The joint economic operation of cascade power stations refers to the joint dispatch under the inter-plant economic operation objective function when the water levels of the two reservoirs are in the same economic operation zone;

[0116] Power Station A and Power Station B are any two cascade hydropower stations;

[0117] Step 3, using DRS-DSA (Dimensionality-reduction strategy-based dynamic search algorithm, DRS-DSA) algorithm and traditional dynamic search algorithm (DSA algorithm) to solve the inter-plant economic operation objective function and the intra-plant economic operation objective function;

[0118] That is, the traditional dynamic search algorithm (DSA algorithm) is used to solve the intra-plant economic operation objective function, and the DRS-DSA algorithm is used to solve the inter-plant economic operation objective function;

[0119] Follow the steps below to implement it:

[0120] Step 3.1, initial value determination

[0121] Assume the total output P of n power stations 总 And the output ratio of each power station in several groups, and get the output P of each power station in each group n, the water consumption Q of the power station is obtained by formula (4) and formula (3), and then the total energy storage F of the cascade power station at the end of the period is obtained by formula (6), formula (2), and formula (1) in turn; the total energy storage F of several cascade power stations at the end of the period is compared, and the largest total energy storage F of the cascade power station at the end of the period is selected as the initial total energy storage F of the cascade power station at the end of the period; the output P of each power station corresponding to the total energy storage F of the initial cascade power station at the end of the period is calculated. n Provides initial output for each power station;

[0122] Step 3.2: Dimensionality reduction optimization

[0123] The total number of time periods T is divided into several stages, each stage contains several time periods, the number of time periods in each stage is equal and the number of stages can be divided by the total number of time periods T, and in each time period, the same increase and decrease operation is performed on the initial output of each power station in step 3.1 by the Box-Behnken Design method, and then according to the calculation process in step 3.1, the total energy storage F of the cascade power station at the end of the period under the increase and decrease operation is obtained, and the total energy storage F of the cascade power station at the end of the period under the increase and decrease operation is compared, and the largest total energy storage F of the cascade power station at the end of the period is selected as the optimized total energy storage F of the cascade power station at the end of the period; the output of each power station after the increase and decrease operation corresponding to the optimized total energy storage F of the cascade power station at the end of the period is the optimized output of each power station;

[0124] Taking the power plant executing the day-ahead plan every 15 minutes as an example, the total time periods of 24 hours a day can be divided into 96 time periods. If the day-ahead plan is executed every 60 minutes, the corresponding total calculation time periods are 24 time periods. Taking the 96 time periods divided into 3 stages as an example, three stages are formed, namely, time periods 1-32, 33-64, and 65-96, and each stage contains 32 time periods.

[0125] Step 3.3, step-by-step optimization

[0126] Increase or decrease the optimized output of each power station in step 3.2, with a step size of 5MW to 10MW. Then, according to the calculation process in step 3.1, the optimal total energy storage F of the cascade power stations at the end of the period is obtained. The optimal total energy storage F of the cascade power stations at the end of the period is compared, and the largest total energy storage F of the cascade power station at the end of the period is selected as the optimal total energy storage F of the cascade power station at the end of the period; the output of each power station corresponding to the optimal total energy storage F of the cascade power station at the end of the period is taken as the optimal output of each power station.

[0127] Example

[0128] The installed capacities of the Zangmu-Jiacha cascade power stations in the middle reaches of the Yarlung Zangbo River are 510MW and 360MW respectively, and the number of units are 6 and 3 respectively. A two-day plan for the Zangmu-Jiacha cascade power stations is made, in which: the flow propagation time between the two stations is calculated as 15 minutes; in addition to the traditional single period constraints, the period coupling constraints mainly consider that the output difference between the two power stations is less than 60MW between the time periods; the water level coordination constraints mainly consider the coordinated control of the water level zones of the two stations. The upper and lower limit water levels of the Zangmu Reservoir are 3309m and 3306m respectively, and [3306m, 3309m] is the economic operation zone of the reservoir. The upper and lower limit water levels of the Jiacha Reservoir are 3245m and 3243m respectively, and [3243m, 3245m] is the economic operation zone of the reservoir.

[0129] By analyzing the results obtained using the DRS-DSA and DSA algorithms, it can be concluded that:

[0130] (1) On typical day 1, the end-of-period energy storage values ​​of the cascade power stations obtained by using the DRS-DSA algorithm and the DSA algorithm differ by 65,400 kW·h; on typical day 2, the DRS-DSA algorithm can achieve an additional benefit of 83,700 kW·h compared with the DSA algorithm. Therefore, the DRS-DSA algorithm can achieve the above additional benefit compared with the DSA algorithm.

[0131] (2) Figure 1 As shown in the figure, if the DRS-DSA algorithm stays in the secondary dimensionality reduction stage (dimensionality reduction optimization), the time it consumes will be greatly reduced and is less than that of the DSA algorithm, and a better scheduling result can be achieved.

Claims

1. A method for economic dispatch of cascade hydropower stations adapting to complex constraints, characterized by: Follow the steps below to implement it: Step 1, establishing the inter-factory economic operation objective function and the intra-factory economic operation objective function; The expression of the inter-factory economic operation objective function is: （1） In formula (1), F is the total energy storage of the cascade power stations at the end of the period; E s ( n ) is a power station n And the water storage above the dead water level of all upstream power stations is at the power station n The amount of electricity that can be generated; E lag ( n ) is the upstream power station at the power station n The amount of electricity generated by the water lag; in, （2） In formula (2), W ( n ) is a power station n The amount of water stored above the dead water level at the end of the dispatch period; W up ( n ) is the water storage capacity of all upstream power stations above the dead water level at the end of the dispatch period; W lag ( n ) is a power station n Delayed water volume generated by all upstream power stations; Q out ( k , t ) is a power station k exist t Outbound flow in the time period; τ k For power station k With power station k +1 water lag time between is the time period length; T is the total number of time periods; η n For power station n The average water consumption rate during the dispatch period; The expression of the objective function of the plant's economic operation is: （3） In formula (3), Q The water consumption of the power station; for l Unit No. in water head H The output is N l The power generation flow at that time; L is the number of units under a certain unit combination; According to the optimization principle of dynamic programming, formula (3) is recursively derived, then: （4） In formula (4), 1~ l Total output of unit No. When the water head is H, the total output is Total power generation flow when optimally allocated between units 1 to 1; It is the load at the initial stage of the power station; The water head is H When the initial stage output In 1~ l Total power generation flow when optimizing allocation among units; It's the crew l Load range; When the water head is H, the output In 1~ Power generation flow when optimizing allocation between units; Step 2, setting constraints on the inter-plant economic operation objective function and the intra-plant economic operation objective function of step 1; Constraints include: single period constraints, period coupling constraints and water level coordination constraints; Single-period constraints include: cascade hydropower station output constraints, water balance constraints, reservoir capacity constraints, reservoir outflow constraints, power station spinning reserve capacity constraints, turbine head limit constraints, and power station restricted area constraints; (1) Output constraints of cascade hydropower stations: （5） In formula (5), P D ( t )for t Total output of cascade power stations in a given grid period, MW; P n ( t )for t Time period power station n Output, MW; N l ( t )for t Time unit l Output, MW; (2) Water balance constraints: （6） In formula (6), V n ( t )for t Time period power station n Reservoir capacity, m 3 ; I n ( t )for t Time period power station n Inflow of the reservoir, m 3 / s; Q fd,n-1 ( t-τ )for t-τ Time period power station n -1 The power generation flow of the reservoir, m 3 / s; Q qs,n-1 ( t-τ )for t-τ Time period power station n -1 The amount of water discharged from the reservoir, m 3 / s; τ For power station n -1 Reservoirs and power stations n Water stagnation time of the reservoir, m 3 / s; Q qj,n ( t ) is a power station n -1 Reservoirs and power stations n The interval inflow of the reservoir, m 3 / s; Q n ( t ) is a power station n Reservoir t Outbound flow rate during the period, m 3 / s; for t Time period power station n -1 The amount of water discharged from the reservoir; H n ( t ) is a power station n The reservoir is located in t The water head of the period, m; f zv,n (·) is the functional relationship between the water level and storage capacity curve of the reservoir where the power station n is located; f zq,n (·) is the functional relationship between the tailwater level and flow rate curve of the reservoir where the power station n is located; is the curve function relationship of the head loss of the reservoir where the power station n is located; f zv,n (·)and f zq,n (·) is obtained by fitting the corresponding measured data, Outbound traffic Q n ( t ) is a quadratic relationship or constant; (3) Reservoir capacity constraints: （7） In formula (7), V n max ( t )for t Time Reservoir n Upper limit of storage capacity; V n min ( t )for t Time Reservoir n Storage capacity lower limit, m 3 ; (4) Reservoir outflow constraints: （8） In formula (8), Q n max ( t )for t Upper limit of reservoir outflow during the period, m 3 / s; Q n min ( t )for t Lower limit of reservoir outflow during the period, m 3 / s; (5) Power station spinning reserve capacity constraints: （9） In formula (9): P n max For power station n Allowable output upper limit, MW; P R ( t )for t Time period power station n Spinning reserve capacity, MW; (6) Turbine head limitation constraints: （10） In formula (10), H n max ( t ) is the time period t Power Station n Upper limit of turbine head, m; H n min ( t ) is the time period t Power Station n Lower limit of turbine head, m; H n ( t ) is the time period t Power Station n Turbine head, m; (7) Power station restricted area constraints: （11） In formula (11), For power station n In the period t No. m The upper limit of the restricted area; For power station n In the period t No. m The lower limit of the restricted area; Time period coupling constraints include: power station output ramp constraints, unit minimum start and stop time constraints; (1) Power station output ramp constraints: （12） In formula (12), PC n For power station n Maximum output fluctuation between adjacent time periods; for t- 1-period power station n Output, MW; (2) Minimum start-up and shutdown time constraints of the unit: （13） In formula (13), T l up For the crew l The shortest boot time; T l down For the crew l Minimum downtime; T l on ( t ) is the unit l exist t -Continuous power-on time before 1 period; T l off ( t ) is the unit l exist t - Continuous downtime 1 period ago; Step 3: Use the DRS-DSA algorithm to solve the inter-plant economic operation objective function.

2. A method for economic dispatch of cascade hydropower stations adapting to complex constraints according to claim 1, characterized in that: Water level coordination constraint: There are upper and lower water level limits during power station operation. When cascade hydropower stations are operated together, the power stations should be in the economic operation area or the upper water level limit area or the lower water level limit area at the same time; Among them, the lower limit water level is less than the economic operation zone and the upper limit water level.

3. The method for economic dispatch of cascade hydropower stations adapting to complex constraints according to claim 1 is characterized in that: Step 3 is implemented as follows: Step 3.1, initial value determination Set the total output of n power stations P 总 And the output ratio of each power station in several groups, and get the output of each power station in each group P n , the water consumption of the power station is obtained by formula (4) and formula (3): Q , and then use formula (6), formula (2), and formula (1) to get the total energy storage of the cascade power station at the end of the period F ; The total energy storage of several cascade power stations at the end of the period F For comparison, the total energy storage at the end of the largest cascade power station was selected. F is the total energy storage at the end of the initial cascade power station F ; The total energy storage of the initial cascade power station at the end of the period F The corresponding power station output P n Provides initial output for each power station; Step 3.2: Dimensionality reduction optimization The total number of time periods T Divided into several stages, each stage contains several time periods, the number of time periods in each stage is equal and the number of stages can be divided by the total number of time periods. T In each period, the Box-Behnken Design method is used to perform the same increase and decrease operation on the initial output of each power station in step 3.1, and then the calculation process in step 3.1 is followed to obtain the total energy storage of the cascade power station at the end of the period under the increase and decrease operation. F , the total energy storage of the cascade power station at the end of the period under the increase and decrease operation F For comparison, the total energy storage at the end of the largest cascade power station was selected. F To optimize the total energy storage at the end of the cascade power station F ; The total energy storage of cascade power stations at the end of the period will be optimized F The output of each power station after the corresponding increase or decrease operation is the optimized output of each power station; Step 3.3, step-by-step optimization Increase or decrease the optimized output of each power station in step 3.2, with a step size of 5MW~10MW. Then, according to the calculation process in step 3.1, the total energy storage of the optimized cascade power station at the end of the period is obtained. F , the total energy storage of the optimal cascade power station at the end of the period F For comparison, the total energy storage at the end of the largest cascade power station was selected. F is the total energy storage of the optimal cascade power station at the end of the period F ; The total energy storage of the optimal cascade power station at the end of the period F The corresponding output of each power station is the optimal output of each power station.

Images