Cascade water-light-storage multi-target medium-term scheduling method considering incoming water photovoltaic uncertainty

Abstract

The invention relates to the technical field of optimal operation of a comprehensive energy system, and discloses a cascade water-light-storage multi-target medium-term scheduling method considering incoming water photovoltaic uncertainty. A cascade water-light-storage medium-term complementary power generation model is established by taking maximization of the total generating capacity of the system and minimization of the residual load variance of the system as targets and combining various unit output constraints, so that the economical efficiency, safety and stability of the system are guaranteed; a hydroelectric generating set vibration area constraint and a direct-current tie line power transmission constraint are considered, so that safe operation of the cascade hydroelectric generating set is guaranteed, and a direct-current tie line power optimization effect is fully exerted; aiming at seasonal change of natural incoming water and daily scale fluctuation of photovoltaic output, an information gap decision theory is adopted to construct a scheduling model considering water and light uncertainty, so that the scheduling method can better avoid risks caused by the water and light uncertainty; and aiming at the constructed multi-target model, solving by adopting an epsilon constraint method, and selecting a compromise solution based on a fuzzy decision-making method, so that the comprehensive satisfaction is maximized.

Description

Technical Field

[0001] The present invention relates to the technical field of optimized operation of integrated energy systems, and in particular to a multi-objective medium-term scheduling method for cascade hydro-photovoltaic-storage systems that takes into account the uncertainty of incoming water and photovoltaic power generation. Background Art

[0002] Photovoltaic power stations have the advantages of low carbon, short construction period, and continuously declining investment costs, which have driven the surge in their installed capacity in recent years. However, solar energy is greatly affected by ambient temperature and meteorological conditions, has great fluctuations in space and time, and exhibits intermittent and random characteristics. The increasing popularity of solar energy poses a threat to the stable and safe operation of the power system. In order to accelerate the large-scale integration of renewable energy other than solar energy, multi-energy complementary power generation has emerged in recent years. Hydropower has the characteristics of flexibility and adjustability. The turbine can be fully powered from startup within a few minutes, thus becoming a perfect complementary power source to smooth the volatility of photovoltaic energy.

[0003] In addition to cascade hydropower, pumped storage power stations are another important power source that can safely and stably alleviate the intermittency and volatility of renewable energy. When the amount of renewable energy generated exceeds the regulation capacity of the system, power abandonment will occur. Therefore, exploring the energy time-shifting capabilities of system components with energy storage capabilities (such as pumped storage with medium- and long-term energy storage capabilities) is crucial to reducing the renewable energy abandonment rate and the corresponding economic losses.

[0004] The influence of vibration zone is often ignored in the modeling process of cascade hydropower stations. However, when the turbine runs in the vibration zone for a long time, it is easy to cause damage to the turbine shaft and blades. Therefore, it is necessary to consider the vibration zone constraint in the modeling process of hydropower units.

[0005] China's renewable energy centers are far away from the load centers, which leads to the need for long-distance, large-capacity energy transmission. However, there are few existing studies on the impact of DC interconnection lines on the operation of multi-energy complementary power generation systems. In addition, the single-objective multi-energy complementary power generation model is difficult to take into account both economic benefits and system stability and safety, so it is necessary to establish a multi-objective scheduling model.

[0006] Light intensity and natural water supply are largely affected by natural conditions such as climate and temperature, which inevitably causes uncertainty in scheduling results and poses a considerable challenge to the efficient use of renewable energy. Domestic and foreign scholars have conducted extensive research on uncertainty processing methods. At present, most research results use robust optimization or stochastic optimization to deal with uncertainty problems. However, the results obtained by robust optimization are often conservative, and the computational complexity of stochastic optimization will increase significantly with the increase of uncertain scenarios. In addition, in many cases it is quite difficult to determine the probability distribution and membership relationship of variables.

[0007] Therefore, it is of great significance to consider vibration zone constraints, DC interconnection line power transmission constraints and multiple scheduling objectives based on the cascade hydro-PV-storage complementary power generation model. In the modeling process of uncertain photovoltaic output and natural water supply, the use of non-probabilistic and non-fuzzy information gap decision theory can effectively deal with the hydro-PV-storage complementary power generation system with complex internal mechanisms and less known information. Summary of the invention

[0008] In view of the above problems, the purpose of the present invention is to provide a multi-objective medium-term scheduling method for cascade hydropower, photovoltaic and storage systems that takes into account the uncertainty of incoming water and photovoltaic power, while taking into account the system's economic performance and safety and stability requirements, giving full play to the power optimization role of hydropower units and DC interconnection lines, and reducing the risks of water and photovoltaic uncertainty to the system's economic performance. The technical solution is as follows:

[0009] A multi-objective medium-term scheduling method for cascade hydro-solar-storage considering the uncertainty of incoming water and photovoltaic power generation includes the following steps:

[0010] Step 1: Determine the objective function of the multi-objective medium-term scheduling model of cascade hydropower-solar-storage: maximize the total power generation of the system and minimize the variance of the remaining load;

[0011] Step 2: Determine the constraints of cascade hydropower stations, including output constraints, water balance and reservoir capacity constraints, and vibration zone constraints;

[0012] Step 3: Establish constraints for the pumped storage power station, including reservoir capacity constraints, state constraints, output constraints, and start-stop times constraints;

[0013] Step 4: Establish output constraints for photovoltaic power plants;

[0014] Step 5: Establish power transmission constraints for DC interconnection lines, including upper and lower limit constraints, regulation rate constraints, and regulation number constraints;

[0015] Step 6: Further consider the uncertainty of natural water supply and photovoltaic output, and establish a robust model based on information gap decision theory;

[0016] Step 7: Use the ε constraint method to solve the model and the fuzzy decision method to select the compromise solution;

[0017] Step 8: Input the equipment parameters and operation parameters of the cascade hydro-PV-storage complementary power generation system, use the commercial solver Gurobi to solve the multi-objective model, and obtain the scheduling results.

[0018] Furthermore, the objective function of the multi-objective medium-term scheduling model of cascade hydro-solar-storage power generation described in step 1 is as follows:

[0019]

[0020] C γ,t =L γ,t-P γ,loc,t

[0021]

[0022] Where: I, S, R and T represent the number of cascade hydropower units, photovoltaic power stations, pumped storage power stations and the number of time periods respectively; γ is the typical day index; P γ,i,t and P γ,s,t Represent the output of cascade hydropower unit i and photovoltaic power station s respectively; and Represent the output of the pumped storage unit r in the power generation state and the pumping state respectively; Δt is the length of the time period; N γ is the number of typical days; L γ,t , C γ,t and They represent the local load, residual load and the mean residual load of a typical day γ in period t; P γ,d,t and P γ,loc,t They represent the transmitted power and local power consumption of DC interconnection line d during period t of a typical day γ respectively; D represents the number of DC interconnection lines; W represents the total power generation of the system; and C represents the variance of the residual load.

[0023] Furthermore, the constraints of the cascade hydropower stations in step 2 are as follows:

[0024] (1) Output constraints:

[0025]

[0026] h γ,i,t ·P i min ≤P γ,i,t ≤h γ,i,t ·P i max

[0027] -ΔP i max ≤P γ,i,t -P γ,i,t-1 ≤ΔP i max

[0028] Where: Q γ,i,t is the power generation flow of cascade hydropower station i in period t; is the average water consumption rate of hydropower conversion; c i Indicates the conversion coefficient; P i min and P i max Indicates the minimum and maximum output of cascade hydropower station i; 0-1 variable h γ,i,trepresents the start-up and shutdown state variables of hydropower unit i in period t of a typical day γ; ΔP i max is the maximum climbing capacity of cascade hydropower unit i;

[0029] (2) Water balance and reservoir capacity constraints:

[0030]

[0031] q γ,i,t =Q γ,i,t +S γ,i,t

[0032] V i min ≤V γ,i,t ≤V i max

[0033]

[0034] Where: V γ，i,t represents the storage capacity of cascade hydropower station i during period t on a typical day γ; q γ,i,t and S γ,i,t Each represents the predicted value of natural water flow, downstream flow and abandoned water flow of cascade hydropower unit i during period t on a typical day γ; V i min and V i max They represent the minimum and maximum storage capacity of cascade hydropower station i respectively; and They represent the minimum and maximum power generation flow of cascade hydropower station i respectively;

[0035] (3) Vibration zone constraints:

[0036]

[0037] u i,k ∈{0, 1}

[0038] Where: M is a sufficiently large positive number; and represents the range of vibration zone k of cascade hydropower station i; u i,k It is a 0-1 variable, indicating that the unit output is less than the lower limit of the vibration zone or higher than the upper limit of the vibration zone. If it is less than the lower limit of the vibration zone, it takes 1, and if it is higher than the upper limit of the vibration zone, it takes 0.

[0039] Furthermore, the constraints of the pumped storage power station in step 3 are as follows:

[0040] (1) Reservoir capacity constraints:

[0041]

[0042] Where: V γ,u,t and V γ,d,t They represent the storage capacity of the upper reservoir and the lower reservoir of the pumped storage power station during the period t of a typical day γ respectively; and Represents the lower and upper limits of the upper reservoir capacity; and Represents the lower and upper limits of the reservoir capacity; η p and η g The hydropower conversion coefficients representing the pumping state and the power generation state respectively;

[0043] (2) State constraints:

[0044]

[0045]

[0046] Where: 0-1 variable and are the power generation and pumping state variables of the unit; and Each represents the startup variable of the unit's power generation and pumping status; and Indicates the shutdown variables for power generation and pumping status; and Represents the power generation and pumping state variables of the entire power station;

[0047] (3) Output constraints:

[0048]

[0049] Where: and Each represents the maximum output of unit r in the pumping and generating states; and Each represents the climbing limit of the unit in the power generation and pumping states;

[0050] (4) Start and stop times constraints:

[0051]

[0052] Where: and They each represent the maximum number of startup times of unit r in the power generation state and the pumping state during the scheduling period.

[0053] Furthermore, the output constraints of the photovoltaic power station in step 4 are as follows:

[0054]

[0055] Where: Represents the predicted value of PV plant s in time period t.

[0056] Furthermore, the DC tie line power transmission constraints in step 5 are as follows:

[0057]

[0058]

[0059] Where: P γ,d,t is the power delivered by the DC tie line during period t of a typical day γ; and Represents the lower and upper limits of power transmission; 0-1 variable and Indicates whether power is adjusted upward or downward; and Indicates the power increase limit and decrease limit in adjacent time periods; n d Represents the maximum allowable number of power adjustments.

[0060] Furthermore, the robust model based on information gap decision theory described in step 6 is as follows:

[0061] (1) Uncertainty description

[0062] The relationship between the predicted and actual values ​​of PV output and natural water supply is as follows;

[0063]

[0064] Where: is the actual value of the photovoltaic power station s in time period t; e represents the photovoltaic output prediction deviation, e max and e min represents the upper and lower limits of the photovoltaic output prediction deviation; β is the photovoltaic output deviation coefficient; represents the water inflow forecast deviation, and represents the upper and lower limits of the water inflow forecast deviation; α is the deviation coefficient of the water inflow forecast; U S and U I are the uncertainty set of photovoltaic output and the uncertainty set of water inflow; I γ,i,t is the actual value of the natural water flow of cascade hydropower unit i during period t on a typical day γ;

[0065] (2) Construct a robust model based on information gap decision theory;

[0066] Write the original deterministic model into matrix form;

[0067] min B(X,d0)

[0068] stF(X,d0)=0,G(X,d0)≤0

[0069] Where: X represents the decision variable; d0 represents the uncertain parameters, including natural water supply and photovoltaic output; F and G represent equality and inequality constraints respectively; B represents the original objective function;

[0070] In information gap decision theory, the uncertainty of uncertain parameters is regarded as the deviation between the predicted value and the actual value. The uncertainty set is as follows:

[0071] d0∈U(d * ,D0)

[0072]

[0073] Where: d * represents the predicted value of the uncertain parameter, U(d * ,D0) is the uncertainty set indicating that the uncertain parameters deviate from the predicted value range; D0 is the maximum fluctuation range of the uncertain parameters;

[0074] The relevant constraints after considering water-light uncertainty are modified as follows;

[0075] V γ,i,t =V γ,i,t-1 +I γ,i,t +q γ,i-1,t -q γ,i,t

[0076]

[0077] Where: I γ,i,t and They belong to the uncertain water supply set U I and the PV output uncertainty set U S ;

[0078] The robust model based on information gap decision theory is described as follows;

[0079] maxα,β

[0080] maxC=C(X,d0)≤(1+λ C )·C0

[0081] minW=W(X,d0)≥(1-λ W )·W0

[0082]

[0083] g(x)≤0,f(x)=0

[0084] Where: C and W represent the residual load variance and total power generation; C0 and W0 represent the residual load variance and total power generation obtained under the deterministic model, which are called basic quantities; λ C and λ W represents the deviation coefficient of the remaining load and total power generation; g(x) and f(x) represent the remaining inequality and equality constraints.

[0085] Furthermore, the ε constraint method in step 7 solves the model and the fuzzy decision method selects the compromise solution as follows:

[0086] Step 7.1: Solve using the ε-constraint method

[0087] The uncertain constraints of water supply and photovoltaic power generation in the robust model are simplified as follows:

[0088]

[0089] The simplified information gap decision theory model is as follows:

[0090] maxα,β

[0091] maxC=C(X,d0)≤(1+λ C )·C0

[0092] minW=W(X,d0)≥(1-λ W )·W0

[0093]

[0094] g(x)≤0,f(x)=0

[0095] The specific steps of solving the multi-objective model by the ε constraint method are as follows;

[0096] 1) Calculate the optimal value of each single objective and obtain their optimal solutions, which represent the optimal solutions under the single objective;

[0097] 2) Calculate the payoff matrix: Substitute the optimal solution of a single objective into other objectives to obtain the payoff matrix;

[0098] 3) Select the primary goal: Select the primary goal based on the decision maker’s preference;

[0099] 4) Determine the ε constraint: Taking another objective as a constraint, the original multi-objective model is transformed into the following model;

[0100] maxα(X)

[0101] g(x)≤0,f(x)=0

[0102] C=C(X,d0)≤(1+λ C )·C0

[0103] W=W(X,d0)≥(1-λ W )·W0

[0104]

[0105] β≥ε i

[0106]

[0107] Where: n represents the number of equal parts; ε i is the lower limit of the target placed in the constraint during the i-th solution, β(X1) and β(X2) are the values ​​of β(X) when considering the targets α(X) and β(X) separately;

[0108] 5) Solve the single-objective model n times to obtain the Pareto solution set;

[0109] Step 7.2: Use fuzzy decision-making method to select Pareto compromise solution. The specific steps are as follows;

[0110] 1) Determine the satisfaction of each Pareto solution:

[0111]

[0112] Where: A v represents the satisfaction of the Pareto solution v; represents the value of the Pareto solution; and Indicates upper and lower limits;

[0113] 2) Determine the target weight

[0114] Determine the weight of the target according to the decision maker's preference v ;

[0115] 3) Determine overall satisfaction

[0116]

[0117] Where: A represents comprehensive satisfaction, V represents the number of Pareto solutions;

[0118] According to the fuzzy decision-making method, the solution with the largest comprehensive satisfaction is selected as the compromise solution; the weights of the two objectives, total system power generation W and residual load variance C, are both set to 0.5; the comprehensive satisfaction is calculated by the following formula:

[0119]

[0120] Where: W max and W min is the upper and lower limits of the total power generation of the system; Cmax and C min are the upper and lower limits of the residual load variance.

[0121] Furthermore, the equipment parameters of the cascade hydro-photovoltaic-storage complementary power generation system described in step 8 include the upper and lower limits of unit output, hydro-electricity conversion coefficient, maximum reservoir capacity, maximum climbing grade and upper and lower limits of vibration range, as well as load data, water inflow and photovoltaic output forecast data.

[0122] The beneficial effects of the present invention are:

[0123] 1) The present invention aims to maximize the total power generation of the system and minimize the variance of the system residual load and establishes a medium-term complementary power generation model of cascade hydropower, photovoltaic and storage combined with the output constraints of various units, while ensuring the economy, safety and stability of the system;

[0124] 2) The present invention takes into account the vibration zone constraints of the hydropower units and the power transmission constraints of the DC interconnection lines, thereby ensuring the safe operation of the cascade hydropower units and giving full play to the power optimization effect of the hydropower units and the DC interconnection lines;

[0125] 3) In view of the seasonal changes in natural water supply and the daily fluctuations in photovoltaic output, the present invention adopts the information gap decision theory to construct a scheduling model that takes into account the uncertainty of water and light, so that the scheduling method can better avoid the risks brought by the uncertainty of water and light.

[0126] 4) The present invention adopts the ε constraint method to solve the constructed multi-objective model, and selects a compromise solution based on the fuzzy decision-making method to maximize the comprehensive satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0127] Figure 1 It is a flow chart of the steps of the method of the present invention.

[0128] Figure 2 It is a schematic diagram of the vibration zone of the hydropower unit.

[0129] Figure 3 This is the structural diagram of the example test system.

[0130] Figure 4 is the multi-timescale water-light uncertainty map. DETAILED DESCRIPTION

[0131] In order to fully illustrate the technical solution disclosed by the present invention, the present invention is further described below in conjunction with the accompanying drawings and specific embodiments.

[0132] The present invention discloses a multi-objective medium-term scheduling method for cascade hydro-solar-storage considering the uncertainty of natural water supply and photovoltaic power generation. The specific implementation steps are as follows: Figure 1 As shown, the technical solution of the present invention comprises the following steps:

[0133] Step 1: Determine the objective function of the multi-objective medium-term scheduling model of cascade hydropower, solar power and storage: maximize the total power generation of the system and minimize the variance of the remaining load.

[0134]

[0135] C γ,t =L γ,t -P γ,loc,t

[0136]

[0137] Where: I, S, R and T represent the number of cascade hydropower units, photovoltaic power stations, pumped storage power stations and the number of time periods. γ is the typical day index. P γ,i,t , P γ,s,t , and represents the output of cascade hydropower unit i, photovoltaic power station s, and pumped storage unit r in power generation and pumping states. Δt is the time period length, which is set to 1 hour. N γ is the typical day number. γ,t , C γ,t and Respectively represent the local load, residual load and residual load mean value of the typical day γ in period t. γ,d,t and P γ,loc,t They represent the power delivered and local power consumed by the DC tie line d in period t of a typical day γ. D represents the number of DC tie lines. W represents the total power generation of the system, and C represents the variance of the residual load.

[0138] Step 2: Determine the output constraints of cascade hydropower stations, mainly including output constraints, water balance and reservoir capacity constraints, and vibration zone constraints.

[0139] (2.1) Output constraints:

[0140]

[0141] h γ,i,t ·P i min ≤P γ,i,t ≤h γ,i,t ·P i max

[0142] -ΔP i max ≤P γ,i,t -P γ,i,t-1 ≤ΔP i max

[0143] Where: Q γ,i,tis the power generation flow of cascade hydropower station i in period t; is the average water consumption rate of hydropower conversion; c i Indicates the conversion coefficient; P i min and P i max Indicates the minimum and maximum output of cascade hydropower station i; 0-1 variable h γ,i,t It represents the start and stop state variable of hydropower unit i in the period t of a typical day γ. ΔP i max is the maximum climbing capacity of cascade hydropower unit i.

[0144] (2.2) Water balance and reservoir capacity constraints:

[0145]

[0146] q γ,i,t =Q γ,i,t +S γ,i,t

[0147] V i min ≤V γ,i,t ≤V i max

[0148]

[0149] Where: V γ，i,t represents the storage capacity of cascade hydropower station i during period t on a typical day γ. q γ,i,t and S γ,i,t Each represents the predicted value of natural water flow, downstream flow and abandoned water flow of cascade hydropower unit i during period t on a typical day γ; V i min and V i max They respectively represent the minimum and maximum storage capacity of cascade hydropower station i. and They represent the minimum and maximum power generation flow of cascade hydropower station i respectively.

[0150] (2.3) Vibration zone constraints:

[0151]

[0152] u i,k ∈{0, 1}

[0153] Where: M is a sufficiently large positive number; and Represents the range of vibration zone k of cascade hydropower station i. i,kis a 0-1 variable.

[0154] Schematic diagram of vibration zone of hydropower unit Figure 2 shown.

[0155] Step 3: Establish constraints for the pumped storage power station, mainly including reservoir capacity constraints, state constraints, output constraints, and start-stop times constraints.

[0156] (3.1) Reservoir capacity constraints:

[0157]

[0158] Where: V γ,u,t and V γ,d,t They each represent the upper and lower reservoir capacities of the pumped storage power station during period t on a typical day γ. and Represents the lower and upper limits of the upper reservoir capacity. Similarly, and Represents the lower and upper limits of the reservoir capacity. η p and η g The hydropower conversion coefficients represent the pumping state and the power generation state respectively.

[0159] (3.2) State constraints:

[0160]

[0161] Where: 0-1 variable and are the power generation and pumping state variables of the unit; and They represent the startup variables of the unit's power generation and pumping states. and Indicates the shutdown variable for power generation and pumping status. and Represents the power generation and pumping state variables of the entire power station.

[0162] (3.3) Output constraints:

[0163]

[0164] Where: and They each represent the maximum output of unit r in the pumping and generating states. and They respectively represent the climbing limit of the unit in power generation and pumping states.

[0165] (3.4) Start and stop times constraints:

[0166]

[0167] Where: and They each represent the maximum number of startup times of unit r in the power generation state and the pumping state during the scheduling period.

[0168] Step 4: Establish output constraints for PV power plants.

[0169]

[0170] Where: Represents the predicted value of PV plant s in time period t.

[0171] Step 5: Establish power transmission constraints for DC interconnection lines, including upper and lower limit constraints, regulation rate constraints, and regulation times constraints.

[0172]

[0173]

[0174] Where: P γ,d,t is the transmission power of the DC interconnection line in time period t. and Represents the lower and upper limits of power transfer. 0-1 variable and Indicates whether power is being adjusted up or down. and Indicates the power increase and decrease limits in adjacent time periods. d Represents the maximum allowable number of power adjustments.

[0175] Step 6: Further consider the uncertainty of natural water supply and photovoltaic output, and establish a robust model based on information gap decision theory.

[0176] (6.1) Information Gap Decision Theory

[0177] Information gap decision theory does not require probability distribution function and membership function, and is suitable for uncertainty modeling of systems with complex internal mechanisms and less known information. Information gap decision theory includes three parts: original system model, uncertain set, and preset target. Information refers to the factors that introduce uncertainty parameters, and gap refers to the deviation between the actual value of uncertain parameters and their predicted values. The uncertain set is a set of all possible values ​​of uncertain parameters. The uncertainty brought by uncertain parameters will have two possible effects on the original target: increasing the risk of the target becoming worse or increasing the possibility of achieving a better target. The model constructed based on the former influence is called a risk-averse robust model, which sets a target value in the worst case and finds the maximum fluctuation range of uncertain parameters that is not worse than that in this case; the model constructed based on the latter influence is called an opportunity model, which sets a better target value and finds the minimum fluctuation range of uncertain parameters when the target can be achieved. The choice of which model depends on the decision maker's wishes. In the present invention, a robust model is selected.

[0178] (6.2) Uncertainty description.

[0179] Due to the influence of rainfall and climate conditions, the natural water inflow of cascade hydropower stations has an uncertainty on a seasonal scale. Unlike photovoltaic power stations, since cascade hydropower stations have a certain water storage capacity, the impact of water inflow uncertainty on power generation is not relatively independent in time. Photovoltaic power generation has uncertainty on a daily scale. The relationship between the predicted value and actual value of photovoltaic output and natural water inflow is as follows.

[0180]

[0181] Where: is the actual value of the photovoltaic power station s in time period t; e represents the photovoltaic output prediction deviation, e max and e min represents the upper and lower limits of the photovoltaic output prediction deviation; β is the photovoltaic output deviation coefficient; represents the water inflow forecast deviation, and represents the upper and lower limits of the water inflow forecast deviation; α is the deviation coefficient of the water inflow forecast; U S and U I are the uncertain sets of photovoltaic power output and water inflow respectively; I γ,i,t is the actual value of the natural water flow of cascade hydropower unit i during period t on a typical day γ.

[0182] (6.3) Construct a robust model based on information gap decision theory.

[0183] The original deterministic model can be written in matrix form.

[0184] min B(X,d0)

[0185] stF(X,d0)=0,G(X,d0)≤0

[0186] Where: X represents the decision variable; d0 represents the uncertain parameters, including natural water supply and photovoltaic output. F and G represent equality and inequality constraints respectively.

[0187] In information gap decision theory, the uncertainty of uncertain parameters is regarded as the deviation between the predicted value and the actual value, and the uncertainty set is as follows:

[0188] d0∈U(d * ,D0)

[0189]

[0190] Where: d * represents the predicted value of the uncertain parameter, U(d * ,D0) is the uncertainty set indicating that the uncertain parameters deviate from the predicted value range; D0 is the maximum fluctuation range of the uncertain parameters.

[0191] The relevant constraints after considering the water-light uncertainty are modified as follows.

[0192] V γ,i,t =V γ,i,t-1 +I γ,i,t +q γ,i-1,t -q γ,i,t

[0193]

[0194] Where: I γ,i,t and They belong to the uncertain water supply set U I and the PV output uncertainty set U S .

[0195] The robust model based on information gap decision theory is described as follows.

[0196] maxα,β

[0197] maxC=C(X,d0)≤(1+λ C )·C0

[0198] minW=W(X,d0)≥(1-λ W )·W0

[0199]

[0200] g(x)≤0,f(x)=0

[0201] Where: C and W represent the residual load variance and total power generation; C0 and W0 represent the residual load variance and total power generation obtained under the deterministic model, which are called basic quantities; λ C and λ W represents the deviation coefficient of the remaining load and total power generation; g(x) and f(x) represent the remaining inequality and equality constraints.

[0202] Step 7: Use the ε constraint method to solve the model and the fuzzy decision method to select the compromise solution.

[0203] (7.1) Introduction to the ε constraint method.

[0204] The robust model established in step (6) is a two-level programming model, which is difficult to solve directly. Considering that the power generation and residual load constraints are easier to meet when there is more water inflow. For a water-solar-electric storage system with sufficient regulation capacity, the greater the photovoltaic output, the easier it is to meet the power generation and residual load constraints in the model. Therefore, the water inflow and photovoltaic uncertainty constraints in the robust model are simplified as follows:

[0205]

[0206] The simplified information gap decision theory model is as follows:

[0207] maxα,β

[0208] maxC=C(X,d0)≤(1+λ C )·C0

[0209] minW=W(X,d0)≥(1-λ W )·W0

[0210]

[0211] g(x)≤0,f(x)=0

[0212] The specific steps of solving the multi-objective model using the ε constraint method are as follows:

[0213] 1) Calculate the optimal value of each single objective and obtain their optimal solutions. represents the optimal solution under a single objective;

[0214] 2) Calculate the payoff matrix. Substitute the optimal solution of a single objective into other objectives to obtain the payoff matrix.

[0215] 3) Select the primary objective. Select the primary objective based on the decision maker’s preference.

[0216] 4) Determine the ε constraint. Taking another objective as a constraint, the original multi-objective model is transformed into the following model.

[0217] maxα(X)

[0218] g(x)≤0,f(x)=0

[0219] C=C(X,d0)≤(1+λ C )·C0

[0220] W=W(X,d0)≥(1-λ W )·W0

[0221]

[0222] β≥ε i

[0223]

[0224] Where: n represents the number of equal parts. i is the lower limit of the objective placed in the constraint during the i-th solution, and β(X1) and β(X2) are the values ​​of β(X) when considering the objectives α(X) and β(X) separately.

[0225] 5) Solving the n-times single-objective model yields the Pareto solution set.

[0226] (7.2) The specific steps of selecting the Pareto compromise solution using the fuzzy decision-making method are as follows.

[0227] 1) Determine the satisfaction of each Pareto solution:

[0228]

[0229] Where: A v represents the satisfaction of the Pareto solution v; Represents the value of the Pareto solution. and Indicates upper and lower limits.

[0230] 2) Determine the target weight

[0231] The weight of the target ω v It can be determined according to the decision maker's preferences, and the choice of weights can directly affect the choice of compromise solution.

[0232] 3) Determine overall satisfaction

[0233]

[0234] Where: A represents the comprehensive satisfaction, and V represents the number of Pareto solutions. According to the fuzzy decision-making method, the solution with the largest comprehensive satisfaction is selected as the compromise solution. The present invention considers that power generation and residual load variance are equally important, and the weights of the two objectives are both 0.5. The comprehensive satisfaction is calculated by the following formula:

[0235]

[0236] The effects of the present invention are described in detail below through specific embodiments.

[0237] (1) Example introduction.

[0238] like Figure 3 The structure diagram of the test system is shown in the figure. The test system includes a photovoltaic power station, a cascade hydropower system with two upstream and downstream power stations, an equivalent pumped storage power station and an equivalent high-voltage DC interconnection line. Figure 4 The multi-time-scale water-light uncertainty diagram shown selects three typical days to represent the normal water period (first stage, 24 hours), the flood period (second stage, 24 hours) and the dry season (third stage, 24 hours).

[0239] Considering the complementarity between water and photovoltaic power, the vibration area of ​​cascade hydropower units, the access of pumped storage power stations, the power optimization of DC interconnection lines, and the influence of water and photovoltaic uncertainty are studied. The settings of the examples are shown in Table 1. It is worth noting that the power transmission plan of the DC interconnection line in Examples 2 and 3 is consistent with that in Example 1.

[0240] Table 1: Settings for Examples 1 to 5

[0241]

[0242] (2) Analysis of the results of the embodiments.

[0243] Table 2 gives the solution results of Example 1 to Example 5. It can be seen that after considering the vibration zone constraint, the power generation of Example 2 is reduced by 21.62MWh compared with Example 1, and the abandonment rate increases to 24.59%, indicating that the system economy is slightly deteriorated after considering the vibration zone constraint, but it avoids the unit operating in a low-efficiency and life-damaging vibration range. Example 3 is connected to the pumped storage power station on the basis of Example 2. It can be seen that under the same residual load variance, the total power generation of Example 3 increases to 1703094.81MWh, and the abandonment rate decreases to 21.11%. After considering the power optimization effect of the DC interconnection line, the total power generation of the system further increases to 1882375.42MWh, and the abandonment rate decreases to 14.88%. Both the pumped storage power station and the DC interconnection line help improve the system economy and the utilization rate of renewable energy.

[0244] Table 2 Comparison of results of different examples

[0245]

[0246] Table 3 shows the variance of the output curve of the cascade hydropower units in Examples 1-4. After considering the role of the pumped storage power station, the variance of the output curve of the cascade hydropower units is significantly reduced, the output is smoother, and the peak load pressure of the units is relieved. However, after considering the role of the DC interconnection line, the output variance of the two units increases to 1227.81MW respectively. 2 and 2147.82MW 2 , which, to a certain extent, sacrifices the regulation capacity of cascade hydropower units.

[0247] Table 3 Variance of output curves of cascade hydropower units in Example 1-4 (MW 2 )

[0248]

[0249] Example 5 considers the uncertainty of natural water supply and photovoltaic output on the basis of Example 4, and uses the information gap decision theory for modeling. Table 4 shows the comparison of the results before and after considering the uncertainty of water and light, where the deviation coefficient of power generation and residual load is 0.02. Obviously, after considering the uncertainty of water and light, the total power generation of Example 5 is 1844727.92MWh, which is higher than 1842028.73MWh of Example 4. This shows that uncertainty modeling based on information gap decision theory has significant advantages in dealing with the adverse effects of water and light uncertainty on the system.

[0250] Table 4 Comparison of results before and after considering water-light uncertainty

[0251]

[0252] The above description is a specific embodiment of the present invention, but it does not limit the patent protection scope of the present invention. Any equivalent changes or substitutions made using the contents of the specification and drawings of the present invention, directly or indirectly applied to other related technical fields, should be included in the protection scope of the present invention.

Claims

1. A multi-objective medium-term scheduling method for cascade hydro-solar-storage considering the uncertainty of incoming water and photovoltaic power generation, characterized in that: The following steps are involved: Step 1: Determine the objective function of the multi-objective medium-term scheduling model of cascade hydropower-solar-storage: maximize the total power generation of the system and minimize the variance of the remaining load; Step 2: Determine the constraints of cascade hydropower stations, including output constraints, water balance and reservoir capacity constraints, and vibration zone constraints; Step 3: Establish constraints for the pumped storage power station, including reservoir capacity constraints, state constraints, output constraints, and start-stop times constraints; Step 4: Establish output constraints for photovoltaic power plants; Step 5: Establish power transmission constraints for DC interconnection lines, including upper and lower limit constraints, regulation rate constraints, and regulation number constraints; Step 6: Further consider the uncertainty of natural water supply and photovoltaic output, and establish a robust model based on information gap decision theory; Step 7: Use the ε constraint method to solve the model and the fuzzy decision method to select the compromise solution; Step 8: Input the equipment parameters and operation parameters of the cascade hydro-PV-storage complementary power generation system, use the commercial solver Gurobi to solve the multi-objective model, and obtain the scheduling results.

2. The multi-objective medium-term scheduling method for cascade hydro-solar-storage considering the uncertainty of incoming water and photovoltaic power according to claim 1 is characterized in that: The objective function of the multi-objective medium-term scheduling model of cascade hydro-solar-storage described in step 1 is as follows: C γ,t =L γ,t -P γ,loc,t Where: I, S, R and T represent the number of cascade hydropower units, photovoltaic power stations, pumped storage power stations and the number of time periods respectively; γ is the typical day index; P γ,i,t and P γ,s,t They represent the output of cascade hydropower unit i and photovoltaic power station s respectively; and Represent the output of the pumped storage unit r in the power generation state and the pumping state respectively; Δt is the length of the time period; N γ is the number of typical days; L γ,t , C γ,t and They represent the local load, residual load and the mean residual load of a typical day γ in period t; P γ,d,t and P γ,loc,t They represent the transmitted power and local power consumption of DC interconnection line d during period t of a typical day γ respectively; D represents the number of DC interconnection lines; W represents the total power generation of the system, and C represents the variance of the residual load.

3. The multi-objective medium-term scheduling method for cascade hydro-solar-storage considering the uncertainty of incoming water and photovoltaic power according to claim 2 is characterized in that: The constraints of the cascade hydropower stations in step 2 are as follows: (1) Output constraints: h γ,i,t ·P i min ≤P γ,i,t ≤h γ,i,t ·P i max -ΔP i max ≤P γ,i,t -P γ,i,t-1 ≤ΔP i max Where: Q γ,i,t is the power generation flow of cascade hydropower station i in period t; is the average water consumption rate of hydropower conversion; c i Indicates the conversion factor; P i min and P i max Indicates the minimum and maximum output of cascade hydropower station i; 0-1 variable h γ,i,t represents the start-up and shutdown state variables of hydropower unit i in period t of a typical day γ; ΔP i max is the maximum climbing capacity of cascade hydropower unit i; (2) Water balance and reservoir capacity constraints: q γ,i,t =Q γ,i,t +S γ,i,t In i min ≤V γ,i,t ≤V i max Where: V γ，i,t represents the storage capacity of cascade hydropower station i during period t on a typical day γ; q γ,i,t and S γ,i,t Each represents the predicted value of natural water flow, downstream flow and abandoned water flow of cascade hydropower unit i during period t on a typical day γ; V i min and V i max They represent the minimum and maximum storage capacity of cascade hydropower station i respectively; and They represent the minimum and maximum power generation flow of cascade hydropower station i respectively; (3) Vibration zone constraints: u i,k ∈{0，1} Where: M is a sufficiently large positive number; and represents the range of vibration zone k of cascade hydropower station i; u i,k It is a 0-1 variable, indicating that the unit output is less than the lower limit of the vibration zone or higher than the upper limit of the vibration zone. If it is less than the lower limit of the vibration zone, it takes 1, and if it is higher than the upper limit of the vibration zone, it takes 0.

4. The multi-objective medium-term scheduling method for cascade hydro-solar-storage considering the uncertainty of incoming water and photovoltaic power according to claim 3 is characterized in that: The pumped storage power station constraints described in step 3 are as follows: (1) Reservoir capacity constraints: Where: V γ,u,t and V γ,d,t They represent the storage capacity of the upper reservoir and the lower reservoir of the pumped storage power station during the period t of a typical day γ respectively; and Represents the lower and upper limits of the upper reservoir capacity; and Represents the lower and upper limits of the reservoir capacity; η p and η g The hydropower conversion coefficients representing the pumping state and the power generation state respectively; (2) State constraints: Where: 0-1 variable and are the power generation and pumping state variables of the unit; and Each represents the startup variable of the unit's power generation and pumping status; and Shutdown variables representing power generation and pumping status; and Represents the power generation and pumping state variables of the entire power station; (3) Output constraints: Where: P r pmax and P r gmax Each represents the maximum output of unit r in the pumping and generating states; and Each represents the climbing limit of the unit in the power generation and pumping states; (4) Start and stop times constraints: Where: and They each represent the maximum number of startup times of unit r in the power generation state and the pumping state during the scheduling period.

5. The multi-objective medium-term scheduling method of cascade hydro-solar-storage considering the uncertainty of incoming water and photovoltaic power according to claim 4 is characterized in that: The output constraints of the photovoltaic power station described in step 4 are as follows: Where: Represents the predicted value of PV plant s in time period t.

6. The multi-objective medium-term scheduling method for cascade hydro-solar-storage considering the uncertainty of incoming water and photovoltaic power according to claim 1 is characterized in that: The DC tie line power transmission constraints described in step 5 are as follows: Where: P γ,d,t is the power delivered by the DC tie line during time period t of a typical day γ; and Represents the lower and upper limits of power transfer; 0-1 variables and Indicates whether power is adjusted upward or downward; and Indicates the power increase limit and decrease limit in adjacent time periods; n d Represents the maximum allowable number of power adjustments.

7. The multi-objective medium-term scheduling method for cascade hydro-solar-storage considering the uncertainty of incoming water and photovoltaic power according to claim 6 is characterized in that: The robust model based on information gap decision theory described in step 6 is as follows: (1) Uncertainty description The relationship between the predicted and actual values ​​of PV output and natural water supply is as follows; Where: is the actual value of the photovoltaic power station s in time period t; e represents the photovoltaic output prediction deviation, e max and e min represents the upper and lower limits of the photovoltaic output forecast deviation; β is the photovoltaic output deviation coefficient; represents the water inflow forecast deviation, and represents the upper and lower limits of the water inflow forecast deviation; α is the deviation coefficient of the water inflow forecast; U S and U I are the uncertain sets of photovoltaic power output and water inflow respectively; I γ,i,t is the actual value of the natural water flow of cascade hydropower unit i during period t on a typical day γ; (2) Construct a robust model based on information gap decision theory; Write the original deterministic model into matrix form; min B(X,d0) stF(X,d0)=0,G(X,d0)≤0 Where: X represents the decision variable; d0 represents the uncertain parameters, including natural water supply and photovoltaic output; F and G represent equality and inequality constraints respectively; B represents the original objective function; In information gap decision theory, the uncertainty of uncertain parameters is regarded as the deviation between the predicted value and the actual value. The uncertainty set is as follows: d0∈U(d * ,D0) Where: d * represents the predicted value of the uncertain parameter, U(d * ,D0) is the uncertainty set indicating that the uncertain parameters deviate from the predicted value range; D0 is the maximum fluctuation range of the uncertain parameters; The relevant constraints after considering water-light uncertainty are modified as follows; V γ,i,t =V γ,i,t-1 +I γ,i,t +q γ,i-1,t -q γ,i,t Where: I γ,i,t and They belong to the uncertain water supply set U I and the PV output uncertainty set U S ; The robust model based on information gap decision theory is described as follows; maxα,β maxC=C(X,d0)≤(1+λ C )·C0 minW=W(X,d0)≥(1-λ W )·W0 g(x)≤0,f(x)=0 Where: C and W represent the residual load variance and total power generation; C0 and W0 represent the residual load variance and total power generation obtained under the deterministic model, which are called basic quantities; λ C and λ W represents the deviation coefficient of the remaining load and total power generation; g(x) and f(x) represent the remaining inequality and equality constraints.

8. The multi-objective medium-term scheduling method for cascade hydro-solar-storage considering the uncertainty of incoming water and photovoltaic power according to claim 7 is characterized in that: In step 7, the ε constraint method is used to solve the model and the fuzzy decision method is used to select the compromise solution as follows: Step 7.1: Solve using the ε-constraint method The uncertain constraints of water supply and photovoltaic power generation in the robust model are simplified as follows: The simplified information gap decision theory model is as follows: maxα,β maxC=C(X,d0)≤(1+λ C )·C0 minW=W(X,d0)≥(1-λ W )·W0 g(x)≤0,f(x)=0 The specific steps of solving the multi-objective model using the ε constraint method are as follows: 1) Calculate the optimal value of each single objective and obtain their optimal solutions, which represent the optimal solutions under the single objective; 2) Calculate the payoff matrix: Substitute the optimal solution of a single objective into other objectives to obtain the payoff matrix; 3) Select the primary goal: Select the primary goal based on the decision maker’s preference; 4) Determine the ε constraint: Taking another objective as a constraint, the original multi-objective model is transformed into the following model; maxα(X) g(x)≤0,f(x)=0 C=C(X,d0)≤(1+λ C )·C0 W=W(X,d0)≥(1-λ W )·W0 β≥ε i Where: n represents the number of equal parts; ε i is the lower limit of the target placed in the constraint during the i-th solution, β(X1) and β(X2) are the values ​​of β(X) when considering the targets α(X) and β(X) separately; 5) Solve the single-objective model n times to obtain the Pareto solution set; Step 7.2: Use fuzzy decision-making method to select Pareto compromise solution. The specific steps are as follows; 1) Determine the satisfaction of each Pareto solution: Where: A v represents the satisfaction of the Pareto solution v; represents the value of the Pareto solution; and Indicates upper and lower limits; 2) Determine the target weight Determine the weight of the target according to the decision maker's preference v ; 3) Determine overall satisfaction Where: A represents comprehensive satisfaction, V represents the number of Pareto solutions; According to the fuzzy decision-making method, the solution with the largest comprehensive satisfaction is selected as the compromise solution; the weights of the two objectives, total system power generation W and residual load variance C, are both set to 0.5; the comprehensive satisfaction is calculated by the following formula: Where: W max and W min is the upper and lower limits of the total power generation of the system; C max and C min are the upper and lower limits of the residual load variance.

9. The multi-objective medium-term scheduling method for cascade hydro-solar-storage considering the uncertainty of incoming water and photovoltaic power according to claim 1 is characterized in that: The equipment parameters of the cascade hydro-solar-storage complementary power generation system described in step 8 include the upper and lower limits of unit output, hydro-electricity conversion coefficient, maximum reservoir capacity, maximum climbing grade and upper and lower limits of vibration range, as well as load data, water inflow and photovoltaic output forecast data.

Images