Short-term peak regulation scheduling method for hydropower-dominated water wind-light system based on multi-energy shared power transmission channel

Abstract

The invention relates to the technical field of hydropower dispatching, and discloses a short-term peak regulation dispatching method for a hydropower main water-guiding wind-light system based on a multi-energy shared power transmission channel. The method comprises the following steps: firstly, respectively aggregating all wind power plants and all photovoltaic electric fields sharing the same power transmission channel into one wind power plant and one photovoltaic electric field, and quantifying output prediction error distribution of the aggregated wind power plants and photovoltaic electric fields by adopting kernel density estimation; secondly, in order to reduce the problem solving scale, a hydroelectric generating set recombination strategy based on the set theory is provided, generating sets sharing the same power transmission channel in the same hydropower station are combined into a virtual generating set, and the hydraulic power and electric power connection between the virtual generating sets is recoupled; finally, differentiated chance constraints based on power transmission channels are introduced, flexibility of water and electricity is fully utilized, and output prediction errors of the aggregated wind power plant and the photovoltaic electric field are accurately compensated. In addition, in order to further improve the solving efficiency, the short-term peak regulation scheduling model is constructed into a mixed integer linear programming model.

Description

Technical Field

[0001] The present invention relates to the technical field of hydropower dispatching, and in particular to a short-term peak-shaving dispatching method for a hydropower-dominated water-wind-solar system based on a multi-energy shared transmission channel. Background Art

[0002] The large-scale grid connection of renewable energy, mainly wind and solar, has not only increased the peak-to-valley difference in load, but also brought major challenges to the absorption of wind and solar due to its inherent anti-regulation, uncontrollability and uncertainty. As a flexible balancing power source, hydropower has the characteristics of low minimum load, fast start-up time and strong load tracking ability, which can provide a buffer for the large influx of wind and solar. In addition, hydropower stations are mostly built in mountainous areas with superior terrain and abundant water resources. These areas also have abundant wind and solar resources. In order to make full use of the transmission channels and maximize the absorption of wind and solar, a large number of hydropower-dominated water, wind and solar systems have emerged. However, the bundled transmission of hydropower, wind and solar is bound to cause the problem of blockage of transmission channels. In addition, affected by different transmission tasks and constrained by geographical factors and economic restrictions, hydropower-dominated water, wind and solar systems are still faced with multi-channel transmission of different units of the same hydropower station (such as Figure 1 For example, the power of units 1 to 14 of the Three Gorges Hydropower Station is sent to the Southern Grid through the Jiangcheng and Longzheng DC lines, while units 15 to 26 are sent to the State Grid through the Yihua, Genan, and Linfeng DC lines. In addition, short-term peak-shaving optimization scheduling is a typical high-dimensional, multi-stage, nonlinear problem. Therefore, in the short-term peak-shaving scheduling of hydropower-dominated hydro-wind-solar systems under multi-channel transmission, a huge set of constraints with highly coupled water and power space-time is formed, which greatly increases the difficulty of modeling and optimization solution.

[0003] In response to the above problems, the present invention takes the hydropower-dominated water, wind and solar system in a certain river basin in the southwest as the research object, deeply analyzes and estimates the wind and solar output prediction errors of the system, and proposes a short-term peak-shaving scheduling method for the hydropower-dominated water, wind and solar system based on a multi-energy shared transmission channel. The method of the present invention can make full use of the flexibility of hydropower, dynamically compensate for the uncertainty of wind and solar output, reduce the system power abandonment rate, and provide an important technical reference for the short-term peak-shaving optimization scheduling of other hydropower-dominated water, wind and solar systems in southwest my country. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a short-term peak-shaving scheduling method for a hydropower-dominated water-wind-solar system based on a multi-energy shared transmission channel. By making full use of the flexible adjustment capability of hydropower, the output prediction errors of wind and light are dynamically balanced, and the power abandonment rate is effectively reduced.

[0005] The technical solution of the present invention is:

[0006] A short-term peak-shaving scheduling method for a hydropower-dominated hydro-wind-solar system based on a multi-energy shared transmission channel. First, all wind farms and all photovoltaic farms sharing the same transmission channel are aggregated into one wind farm and one photovoltaic farm, respectively, and kernel density estimation is used to quantify the output forecast error distribution of the aggregated wind farms and photovoltaic farms. Secondly, in order to reduce the scale of the problem, a hydropower unit reorganization strategy based on set theory is proposed. The units sharing the same transmission channel in the same hydropower station are merged into a virtual unit, and the hydraulic and electrical connections between the virtual units are recoupled. Finally, differentiated opportunity constraints based on transmission channels are introduced to make full use of the flexibility of hydropower and accurately compensate for the output forecast errors of aggregated wind farms and photovoltaic farms. In addition, in order to further improve the solution efficiency, the short-term peak-shaving scheduling model is constructed as a mixed integer linear programming model.

[0007] The specific steps include:

[0008] (1) Aggregation of wind and photovoltaic fields based on transmission channels

[0009]

[0010] Where: is the output of the aggregated wind farm through transmission channel l in time period t (MW); The output (MW) of the aggregated PV field through transmission channel l in time period t. is the output (MW) of wind farm w through transmission channel l in time period t; w is the wind farm number; W l is the total number of wind farms in the transmission channel l; is the output (MW) of the photovoltaic field s through the transmission channel l in time period t; s is the photovoltaic field number; S l is the total number of photovoltaic electric fields in the transmission channel l.

[0011] (2) Description of the forecast error of wind and photovoltaic power output after aggregation

[0012]

[0013] Where: is the output forecast error of the aggregated wind farm through the transmission channel l in time period t (MW); is the output prediction error of the aggregated PV field through the transmission channel l in time period t (MW); is the predicted wind power output through transmission channel l in time period t (MW); is the predicted PV power output through transmission channel l in time period t (MW); is the actual wind power output through transmission channel l in time period t (MW); is the actual PV power output through transmission channel l in time period t (MW); is the relative output prediction error of the aggregated wind farm through the transmission channel l in time period t (MW); is the relative output prediction error (MW) of the aggregated PV field through the transmission channel l in time period t.

[0014] (3) Using kernel density estimation to quantify wind and photovoltaic power output prediction errors

[0015]

[0016] Where: For wind power Estimated probability density function value at the output; is the forecast error of the mth wind power output through the transmission channel l in time period t (MW); For photovoltaic Estimated probability density function value at the output; is the prediction error of the nth PV power output through transmission channel l in time period t (MW); ke(·) is the kernel function; h is the bandwidth; M is the total number of historical wind power output; N is the total number of historical PV power output.

[0017] (4) Based on steps (1)-(3), the probability density functions (PDFs) and cumulative distribution functions (CDFs) of the output forecast errors of the aggregated wind farm and the aggregated photovoltaic farm are estimated.

[0018] (5) Description of hydropower output based on hydropower unit reconstruction strategy

[0019]

[0020] Where: is the hydropower output through transmission channel l in time period t (MW); is the output (MW) of virtual unit v in hydropower station i through transmission channel l in time period t; i is the hydropower station number; I l is the number of hydropower stations in the transmission channel l; v is the virtual unit number; V i is the number of virtual machine groups contained in hydropower station i.

[0021] (6) Construct a short-term peak load regulation objective function based on the wind and solar power output aggregated in step (1) and the hydropower output in step (5)

[0022]

[0023] C t =L t -P t (11)

[0024]

[0025] Where: F is the peak-to-valley difference of residual load (MW); C t is the residual load in period t (MW); L t is the original load in period t (MW); P t is the total output of the hydropower-dominated water, wind and solar power system in time period t (MW); l is the transmission channel number; L is the total number of transmission channels; P l,t is the hydropower, wind and solar power output (MW) through transmission channel l during time period t.

[0026] (7) Based on steps (1), (5) and (6), describe the grid constraints of short-term peak-shaving dispatching of multi-channel power transmission of hydropower-dominated hydro-wind-solar system

[0027] a) Transmission channel capacity constraints

[0028]

[0029] Where: are the lower and upper limits of the output of transmission channel l (MW).

[0030] b) Positive compensation differentiation opportunity constraints based on transmission channels

[0031]

[0032] Where: is the upper limit of the output of the virtual unit v in the hydropower station i through the transmission channel l (MW); β is the positive compensation confidence interval.

[0033] c) Opportunity constraints for negative compensation differentiation based on transmission channels

[0034]

[0035] Where: is the lower limit of the output of the virtual unit v in the hydropower station i through the transmission channel l (MW); γ is the negative compensation confidence interval.

[0036] (8) Based on step (5), describe the hydropower operation constraints of short-term peak-shaving dispatching of multi-channel power transmission of hydropower-dominated hydro-wind-solar system

[0037] a) Water balance constraints

[0038]

[0039] Where: V i,t is the reservoir capacity of hydropower station i in time period t (1×10 4 m 3 );R i,t is the interval flow of hydropower station i in time period t (m 3 / s); is the outflow flow of virtual machine group v in hydropower station i in period t (m 3 / s); is the power generation flow of virtual unit v in hydropower station i in time period t (m 3 / s); is the water abandonment flow of virtual unit v in hydropower station i in period t (m 3 / s); is the hydropower station g in the period t-τ g,i Outbound flow (m 3 / s); g is the number of the power station directly upstream of hydropower station i; τ g,i is the lag time between hydropower station g and i (h); Ω i is the set of power stations directly upstream of hydropower station i; Δt is the scheduling step length (h).

[0040] b) Outbound flow constraints

[0041]

[0042] Where: are the lower and upper limits of the power generation flow of hydropower station i (m 3 / s).

[0043] c) Power generation flow constraints

[0044]

[0045] Where: are the lower and upper limits of the power generation flow of hydropower station i (m 3 / s).

[0046] d) Discarded water flow constraints

[0047]

[0048] Where: are the lower and upper limits of the water discharge of hydropower station i (m 3 / s).

[0049] e) Water level constraint

[0050]

[0051] Where: Z i,t is the water level above the dam of hydropower station i in time period t (m); is the lower and upper limits of water level of hydropower station i (m).

[0052] f) Water level and storage capacity curve constraints

[0053]

[0054] Where: is the water level and storage capacity curve of hydropower station i.

[0055] g) Tailwater level power generation flow curve constraints

[0056]

[0057] Where: is the tailwater level of hydropower station i in time period t (m); is the relationship curve between tailwater level and power generation flow of hydropower station i.

[0058] h) Head constraint

[0059]

[0060] Where: H i,t is the water head of hydropower station i in period t (m).

[0061] i) Output constraints

[0062]

[0063] Where: are the lower and upper output limits of virtual unit v in hydropower station i (MW).

[0064] j) Dynamic characteristic curve constraints

[0065]

[0066] Where: It is the relationship curve between the output of virtual unit v in hydropower station i and the power generation flow and water head.

[0067] k) Initial and final storage capacity constraints

[0068] z i,init =Z i,1 (28)

[0069] z i,end =Z i,T+1 (29)

[0070] Where: z i,init 、z i,end are the initial and final water levels of hydropower station i (m).

[0071] (9) Based on step (5), the objective function of the model is linearized by introducing corresponding auxiliary variables to construct a mixed integer phenomenon programming (MILP) model. The linearized objective function is:

[0072] minF=min( R MAX -RMIN ) (30)

[0073]

[0074] Where: R MAX , R MIN is an auxiliary variable.

[0075] (10) Based on steps (2), (3), (6), and (7), the differentiated opportunity constraint is converted into an equivalent linear constraint. The transformed differentiated opportunity constraint is:

[0076]

[0077] Where: and It is the inverse function of the cumulative distribution function CDF of the output forecast error of the aggregated wind farm and the aggregated photovoltaic farm. is the positive compensation capability of the transmission channel l, is the negative compensation capability of transmission channel l.

[0078] (11) Constraint aggregation technology is used to linearize the univariate nonlinear constraints (Formula (23) and Formula (24)) and the binary dynamic characteristic curve (Formula (27)) in the model.

[0079] (12) Combining the constraints of steps (5)-(8) (wherein step (7) only considers the transmission channel capacity constraint), the differentiated opportunity constraints transformed in step (10), and the linearized objective function of step (9), a MILP model is constructed and solved to obtain a scheduling solution.

[0080] Beneficial effects of the present invention: In view of the challenges brought to short-term peak-shaving optimization scheduling by the multi-channel transmission and multi-energy shared transmission channels of the hydropower-dominated hydro-wind-solar system, the present invention proposes a short-term peak-shaving scheduling method for the hydropower-dominated hydro-wind-solar system based on a multi-energy shared transmission channel. First, all wind farms and all photovoltaic farms sharing the same transmission channel are aggregated into one wind farm and one photovoltaic farm, respectively, and the kernel density estimation is used to quantify the distribution of output prediction errors of the aggregated wind farms and photovoltaic farms; secondly, in order to reduce the scale of the problem to be solved, a hydropower unit reorganization strategy based on set theory is proposed, which merges the units sharing the same transmission channel in the same hydropower station into a virtual unit, and recouples the hydraulic and electrical connections between the virtual units; finally, differentiated opportunity constraints based on transmission channels are introduced to make full use of the flexibility of hydropower and accurately compensate for the output prediction errors of the aggregated wind farms and photovoltaic farms. In addition, in order to further improve the solution efficiency, the short-term peak-shaving scheduling model is constructed as a mixed integer linear programming model. Compared with the existing technology, the present invention can accurately utilize the regulation capacity of hydropower, dynamically compensate for the output prediction errors of wind and light, and effectively reduce the power abandonment rate of hydropower-dominated water-wind-solar systems, and has significant practicality. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 This is a schematic diagram of the hydropower-dominated water, wind and solar system;

[0082] Figure 2 It is a schematic diagram of the overall solution framework of the present invention;

[0083] Figure 3 This is a geographical location map of a hydropower system in a river basin;

[0084] Figure 4 The predicted output, actual output and estimated range of actual output for wind and solar power;

[0085] Figure 5 The optimal peak-shaving results of different models;

[0086] Figure 6 The power output process of the shared transmission channel for different models;

[0087] Figure 7 This is the detailed output process of each power station in different transmission channels of Model 1. DETAILED DESCRIPTION

[0088] The specific implementation of the present invention is further described below in conjunction with the accompanying drawings and the invention content.

[0089] This embodiment mainly includes three parts: wind power and photovoltaic power output forecast error analysis, model construction based on hydropower unit reconstruction strategy, introduction of differentiated opportunity constraints and linearization processing. The detailed solution framework is as follows: Figure 2shown.

[0090] 1. Analysis of wind and solar output prediction errors:

[0091] Due to the frequent changes in meteorological factors, wind and solar outputs are volatile and uncontrollable. In order to reduce the impact of output prediction errors, the present invention combines the regional aggregation effect and aggregates the wind farms and photovoltaic farms that share the same transmission channel into one wind farm and one photovoltaic farm. The specific process is as follows:

[0092]

[0093] Where: is the output of the aggregated wind farm through transmission channel l in time period t (MW); The output (MW) of the aggregated PV field through transmission channel l in time period t. is the output (MW) of wind farm w through transmission channel l in time period t; w is the wind farm number; W l is the total number of wind farms in the transmission channel l; is the output (MW) of the photovoltaic field s through the transmission channel l in time period t; s is the photovoltaic field number; S l is the total number of photovoltaic electric fields in the transmission channel l.

[0094] Based on the aggregated wind farms and photovoltaic farms, kernel density estimation is further used to quantify the wind and solar output prediction errors. The specific process is as follows:

[0095]

[0096] Where: is the output forecast error of the aggregated wind farm through the transmission channel l in time period t (MW); is the output prediction error of the aggregated PV field through the transmission channel l in time period t (MW); is the predicted wind power output through transmission channel l in time period t (MW); is the predicted PV power output through transmission channel l in time period t (MW); is the actual wind power output through transmission channel l in time period t (MW); is the actual PV power output through transmission channel l in time period t (MW); is the relative output prediction error of the aggregated wind farm through the transmission channel l in time period t (MW); is the relative output prediction error of the aggregated PV field through the transmission channel l in time period t (MW); For wind power Estimated probability density function value at the output; is the forecast error of the mth wind power output through the transmission channel l in time period t (MW); For photovoltaic Estimated probability density function value at the output; is the prediction error of the nth PV power output through transmission channel l in time period t (MW); ke(·) is the kernel function; h is the bandwidth; M is the total number of historical wind power output; N is the total number of historical PV power output.

[0097] 2. Model construction based on hydropower unit reconstruction strategy

[0098] A. Objective Function

[0099]

[0100] C t =L t -P t (44)

[0101]

[0102] Where: F is the peak-to-valley difference of residual load (MW); C t is the residual load in period t (MW); L t is the original load in period t (MW); P t is the total output of the hydropower-dominated water, wind and solar power system in time period t (MW); l is the transmission channel number; L is the total number of transmission channels; P l,t is the hydropower, wind and solar power output (MW) through transmission channel l during time period t; is the hydropower output through transmission channel l in time period t (MW); is the output (MW) of virtual unit v in hydropower station i through transmission channel l in time period t; i is the hydropower station number; I l is the number of hydropower stations in the transmission channel l; v is the virtual unit number; V i is the number of virtual machine groups included in hydropower station i; is the predicted output of wind farm w through transmission channel l in time period t (MW); is the predicted output (MW) of the photovoltaic power field s through the transmission channel l in time period t; t is the scheduling period number; T is the scheduling cycle number.

[0103] B. Grid Constraints

[0104] a) Transmission channel capacity constraints

[0105]

[0106] Where: are the lower and upper limits of the output of transmission channel l (MW).

[0107] b) Positive compensation differentiation opportunity constraints based on transmission channels

[0108]

[0109] Where: is the upper limit of the output of the virtual unit v in the hydropower station i through the transmission channel l (MW); β is the positive compensation confidence interval.

[0110] c) Opportunity constraints for negative compensation differentiation based on transmission channels

[0111]

[0112] Where: is the lower limit of the output of the virtual unit v in the hydropower station i through the transmission channel l (MW); γ is the negative compensation confidence interval.

[0113] C. Operational constraints of hydropower

[0114] a) Water balance constraints

[0115]

[0116] Where: V i,t is the reservoir capacity of hydropower station i in time period t (1×10 4 m 3 );R i,t is the interval flow of hydropower station i in time period t (m 3 / s); is the outflow flow of virtual machine group v in hydropower station i in period t (m 3 / s); is the power generation flow of virtual unit v in hydropower station i in time period t (m 3 / s); is the water abandonment flow of virtual unit v in hydropower station i in period t (m 3 / s); is the hydropower station g in the period t-τ g,i Outbound flow (m 3 / s); g is the number of the power station directly upstream of hydropower station i; τ g,i is the lag time between hydropower station g and i (h); Ω i is the set of power stations directly upstream of hydropower station i; Δt is the scheduling step length (h).

[0117] b) Outbound flow constraints

[0118]

[0119] Where: are the lower and upper limits of the power generation flow of hydropower station i (m 3 / s).

[0120] c) Power generation flow constraints

[0121]

[0122] Where: are the lower and upper limits of the power generation flow of hydropower station i (m 3 / s).

[0123] d) Discarded water flow constraints

[0124]

[0125] Where: are the lower and upper limits of the water discharge of hydropower station i (m 3 / s).

[0126] e) Water level constraint

[0127]

[0128] Where: Z i,t is the water level above the dam of hydropower station i in time period t (m); is the lower and upper limits of water level of hydropower station i (m).

[0129] f) Water level and storage capacity curve constraints

[0130]

[0131] Where: is the water level and storage capacity curve of hydropower station i.

[0132] g) Tailwater level power generation flow curve constraints

[0133]

[0134] Where: is the tailwater level of hydropower station i in time period t (m); is the relationship curve between tailwater level and power generation flow of hydropower station i.

[0135] h) Head constraint

[0136]

[0137] Where: H i,t is the water head of hydropower station i in period t (m).

[0138] i) Output constraints

[0139]

[0140] Where: are the lower and upper output limits of virtual unit v in hydropower station i (MW).

[0141] j) Dynamic characteristic curve constraints

[0142]

[0143] Where: It is the relationship curve between the output of virtual unit v in hydropower station i and the power generation flow and water head.

[0144] k) Initial and final storage capacity constraints

[0145] z i,init =Z i,1 (64)

[0146] z i,end =Z i,T+1 (65)

[0147] Where: z i,init 、z i,end are the initial and final water levels of hydropower station i (m).

[0148] 3. Linearize the objective function of the model

[0149] The objective function (43) is the minimization objective of the peak-to-valley difference of residual load commonly used in power grid peak regulation engineering practice. However, since the objective function is nonlinear, it is impossible to directly construct a MILP model to solve it. Therefore, by introducing the auxiliary variable R MIN , R MAX It is linearized. The objective function expression after linearization is as follows:

[0150] minF=min(R MAX -R MIN ) (66)

[0151]

[0152] Where: R MAX , R MIN is an auxiliary variable.

[0153] In addition, there are still univariate nonlinear constraints in the model, such as the water level storage capacity curve constraint of formula (59), the tailwater level power generation flow curve constraint of formula (60), and the binary dynamic characteristic curve constraints such as formula (63). In order to solve these nonlinear constraint problems, constraint aggregation technology is further used to linearize them.

[0154] 4. Equivalent transformation of differentiated opportunity constraints

[0155] According to formulas (37) and (38), and Formulas (51) and (52) are transformed into equivalent linear forms. The specific transformation steps are as follows:

[0156]

[0157] when Can be further transformed into:

[0158]

[0159] when Will Substituting into formula (71), when Will Substitute into formula (72). Then formulas (51) and (52) are equivalently converted into the following deterministic form:

[0160]

[0161] Where: and is the inverse function of the CDF of the output forecast error of the aggregated wind farm and the aggregated photovoltaic farm. is the positive compensation capability of the transmission channel l, is the negative compensation capability of transmission channel l.

[0162] 5. Example Analysis

[0163] The actual data of a typical day in the flood season of a hydropower-dominated water-wind-solar system in a river basin in southwest my country is used to perform a 24-hour simulation calculation as a verification example of the method of the present invention. In this system, hydropower stations A, D, E, F, G, H and I are located in the main river, and power stations B and C are located in tributaries and merged into the main river between hydropower stations A and D. The topological structure is as follows: Figure 3As shown. Among them, the A and G power stations on the main rivers are annual regulation and above power stations, and the total installed capacity of the hydropower system is 8759MW. The hydropower-dominated water-wind-solar system compensates for the output prediction errors of wind and light by utilizing the regulation capacity of the hydropower station, alleviating the phenomenon of power abandonment and power restriction in the power grid. Due to the different power transmission tasks, there is not only multi-channel transmission between water, wind and light, but also extremely complex multi-channel transmission relationships between hydropower units. For example, #1 of hydropower station D is connected to the grid through transmission channel 1, while the power generation of #2, #3 and #4 of D is connected to the grid through transmission channel 2. These unique characteristics make the short-term peak-shaving optimization of the hydropower-dominated water-wind-solar system a major challenge. In addition, in order to reduce the impact of meteorological factors on the output prediction of a single wind farm and photovoltaic farm, the wind farm and photovoltaic farm in the study area are aggregated into one wind farm and one photovoltaic farm respectively. The aggregated wind power part is transmitted through transmission channel 1, and the other part is transmitted through transmission channel 4, while the aggregated photovoltaic power is transmitted through all channels through transmission channel 4. The specific parameters are shown in Table 1.

[0164] Table 1 Basic parameters of hydropower-dominated water, wind and solar systems

[0165]

[0166] To verify the effectiveness of the method of the present invention, the model proposed in the present invention is named Model 1, and the control models are named Model 2, Model 3 and Model 4. Model 2 is the same as Model 1, the only difference is that Model 2 does not consider the wind and light output prediction errors, Model 3 is the same as Model 2, but does not consider the transmission channel capacity constraint, and the scheduling process of Model 4 is based on the actual scheduling data of the power scheduling department, which is used to verify the effectiveness of Model 1. In addition, the confidence intervals of the wind power output prediction error and the photovoltaic output prediction error of the present invention are set to 95% respectively. For ease of understanding, the constraint settings of the above four models are shown in Table 2.

[0167] Table 2 Model settings

[0168]

[0169] Figure 4 The wind and solar power forecast, actual output, and actual output estimation range are displayed. Figure 4 It can be seen that in certain periods, the actual output of wind and solar power is much lower than the predicted output. If there is not enough backup power, power shortages are likely to occur during these periods.

[0170] The peak load regulation results of different models are shown in Table 3. Figure 5Table 4 shows that the peak-to-valley difference of the residual load of Model 1, 2, 3 and 4 is reduced by 45.92%, 55.52%, 65.21% and 53.34% respectively compared with the original load. This shows that all models have effectively alleviated the peak load regulation pressure of the power grid to varying degrees. Figure 5 It can be seen that Model 1 can effectively avoid power abandonment, while the hydropower output of Models 3 and 4 is fully loaded at peak, resulting in power abandonment rates of 4.41% and 4.44% respectively (see Table 3). In addition, although Model 2 did not experience power abandonment, when the actual output of wind and solar power is far lower than the predicted output, Model 2 may face the risk of power shortage.

[0171] Table 3 Peak shaving results of different models

[0172]

[0173] The power abandonment rate of the shared transmission channel of different models is shown in Table 4. Figure 6 As shown. In the case studied in this invention, only transmission channel 1 and transmission channel 4 involve water, wind and solar power sharing transmission channels. Figure 6 As shown in (a), the output of Model 3 and Model 4 exceeds the capacity limit of the transmission channel (see black boxes I and II), resulting in different degrees of power abandonment (see Table 4). Figure 6 As can be seen in (b), from period 1 to period 8, the hydropower output of Models 2, 3 and 4 is lower than the lower limit of hydropower output. When the actual output of wind and solar power is lower than the predicted output, this may lead to power shortage.

[0174] Table 4 Power curtailment rate of shared transmission channels in different models

[0175]

[0176] The detailed output process of each transmission channel is as follows: Figure 7 As shown in the figure, since Model 1 takes into account the forecast errors of wind and solar output, as long as the actual output process of hydropower remains within the allowable range (as shown by the black solid line), there is sufficient reserve power to compensate for the forecast errors of wind and solar output.

Claims

1. A short-term peak-shaving dispatching method for a hydropower-dominated water-wind-solar system based on a multi-energy shared transmission channel, characterized in that: The steps include: (1) Aggregation of wind and photovoltaic fields based on transmission channels Where: is the output of the aggregated wind farm through transmission channel l in time period t, MW; The output of the aggregated photovoltaic field through the transmission channel l in time period t, MW; is the output of wind farm w through transmission channel l in time period t, MW; w is the wind farm number; W l is the total number of wind farms in the transmission channel l; is the output of photovoltaic field s through transmission channel l in time period t, MW; s is the photovoltaic field number; S l is the total number of photovoltaic electric fields in the transmission channel l; (2) Description of the forecast error of wind and photovoltaic power output after aggregation Where: is the output forecast error of the aggregated wind farm through the transmission channel l in time period t, MW; is the output prediction error of the aggregated PV field through the transmission channel l in time period t, MW; is the predicted wind power output through transmission channel l in time period t, MW; is the predicted PV power output through transmission channel l in time period t, MW; is the actual wind power output through transmission channel l during period t, MW; is the actual PV power output through transmission channel l during time period t, MW; is the relative output prediction error of the aggregated wind farm through the transmission channel l in time period t, MW; is the relative output prediction error of the aggregated PV field through the transmission channel l in time period t, MW; (3) Using kernel density estimation to quantify wind and photovoltaic power output prediction errors Where: For wind power Estimated probability density function value at the output; is the forecast error of the mth wind power output through the transmission channel l in time period t, MW; For photovoltaic Estimated probability density function value at the output; is the prediction error of the nth PV power output through transmission channel l in time period t, MW; ke(·) is the kernel function; h is the bandwidth; M is the total number of historical wind power output; N is the total number of historical PV power output; (4) estimating the probability density function and cumulative distribution function of the output prediction error of the aggregated wind farm and the aggregated photovoltaic farm based on steps (1) to (3); (5) Description of hydropower output based on hydropower unit reconstruction strategy Where: is the hydropower output through transmission channel l in time period t, MW; is the output of virtual unit v in hydropower station i through transmission channel l in time period t, MW; i is the hydropower station number; I l is the number of hydropower stations in the transmission channel l; v is the virtual unit number; V i is the number of virtual machine groups included in hydropower station i; (6) Construct a short-term peak load regulation objective function based on the wind and solar power output aggregated in step (1) and the hydropower output in step (5) C t =L t -P t (11) Where: F is the peak-to-valley difference of residual load, MW; C t is the residual load in period t, MW; L t is the original load in period t, MW; P t is the total output of the hydropower-dominated water, wind and solar power system in time period t, MW; l is the transmission channel number; L is the total number of transmission channels; P l,t is the hydropower, wind and solar power output through transmission channel l in time period t, MW; (7) Based on steps (1), (5) and (6), describe the grid constraints of short-term peak-shaving dispatching of multi-channel power transmission of hydropower-dominated hydro-wind-solar system a) Transmission channel capacity constraints Where: are the lower and upper limits of the output of transmission channel l, MW; b) Positive compensation differentiation opportunity constraints based on transmission channels Where: is the upper limit of the output of virtual unit v in hydropower station i through transmission channel l, MW; β is the confidence interval for positive compensation; c) Opportunity constraints for negative compensation differentiation based on transmission channels Where: is the lower limit of the output of the virtual unit v in the hydropower station i through the transmission channel l, MW; γ is the negative compensation confidence interval; (8) Based on step (5), describe the hydropower operation constraints of short-term peak-shaving dispatching of multi-channel power transmission of hydropower-dominated hydro-wind-solar system a) Water balance constraints Where: V i,t is the reservoir capacity of hydropower station i in time period t, 1×10 4 m 3 ; R i,t is the interval flow of hydropower station i in time period t, m 3 / s; is the outflow flow of virtual unit v in hydropower station i in period t, m 3 / s; is the power generation flow of virtual unit v in hydropower station i in time period t, m 3 / s; is the water abandonment flow of virtual unit v in hydropower station i in period t, m 3 / s; is the hydropower station g in the period t-τ g,i Outbound flow, m 3 / s; g is the number of the power station directly upstream of hydropower station i; τ g,i is the lag time between hydropower station g and i, h; Ω i is the set of power stations directly upstream of hydropower station i; Δt is the scheduling step length, h; b) Outbound flow constraints Where: are the lower and upper limits of the power generation flow of hydropower station i, m 3 / s; c) Power generation flow constraints Where: are the lower and upper limits of the power generation flow of hydropower station i, m 3 / s; d) Discarded water flow constraints Where: are the lower and upper limits of the water discharge of hydropower station i, respectively, 3 / s; e) Water level constraint Where: Z i,t is the water level above the dam of hydropower station i in time period t, m; is the lower and upper limits of the water level of hydropower station i, m; f) Water level and storage capacity curve constraints With i,t =f i ZV (In i,t ) (23) Where: f i ZV (·) is the water level and storage capacity curve of hydropower station i; g) Tailwater level power generation flow curve constraints Where: is the tailwater level of hydropower station i in period t, m; f i ZQ (·) is the tailwater level power generation flow relationship curve of hydropower station i; h) Head constraint Where: H i,t is the water head of hydropower station i in time period t, m; i) Output constraints Where: is the lower and upper limits of the output of virtual unit v in hydropower station i, MW; j) Dynamic characteristic curve constraints Where: is the relationship curve between the output of virtual unit v in hydropower station i and the power generation flow and water head; k) Initial and final storage capacity constraints With i,init =Z i,1 (28) With i,end =Z i,T+1 (29) Where: z i,init 、z i,end are the initial and final water levels of hydropower station i, m; (9) Based on step (5), the objective function of the model is linearized by introducing corresponding auxiliary variables to construct a mixed integer phenomenon programming model MILP. The linearized objective function is: minF=min(R MAX -R MIN (30) Where: R MAX , R MIN is an auxiliary variable; (10) Based on steps (2), (3), (6), and (7), the differentiated opportunity constraint is converted into an equivalent linear constraint. The transformed differentiated opportunity constraint is: Where: and is the inverse function of the cumulative distribution function of the output forecast error of the aggregated wind farm and the aggregated photovoltaic farm; is the positive compensation capability of the transmission channel l, is the negative compensation capability of the transmission channel l; (11) linearizing the one-dimensional nonlinear constraints shown in formula (23) and formula (24) and the two-dimensional dynamic characteristic curve shown in formula (27) in the model by using constraint aggregation technology; (12) Combining the constraints of steps (5)-(8), where step (7) only considers the capacity constraint of the transmission channel, the differentiated opportunity constraint after transformation in step (10), and the linearized objective function in step (9), a MILP model is constructed and solved to obtain a scheduling solution.

Images